Ville Pulkki, Matti Karjalainen — Communication Acoustics- An Introduction to Speech, Audio and Psychoacoustics

Yield

846 of 2862 paragraphs kept (29.6%). Target for mining: 30%–50% — ⚠️ BELOW target.

Pre-Assessment

Ville Pulkki and Matti Karjalainen are both highly credible acoustics researchers; Pulkki is a professor at Aalto University known for spatial audio (Vector Base Amplitude Panning, VBAP) and Karjalainen was a pioneering professor of audio signal processing at Aalto (Helsinki University of Technology) before his passing in 2010, with decades of work in speech acoustics, loudspeaker design, and psychoacoustics. The book is a 2015 Wiley textbook, so it’s reasonably current for fundamentals, though research in some areas (e.g., immersive audio) has advanced since.

The likely yield given a mining stance for “applications of acoustic communication” is predominantly structured concepts/frameworks: the book is organized to link physical acoustics, speech production/perception, audio technology, and psychoacoustics into an applied framework. Expect fewer raw facts you’ll need to memorize and more usable conceptual maps connecting theory to application domains (speech coding, loudspeaker/room acoustics, hearing aids, spatial audio). The authors’ applied-research backgrounds make this a strong source for technology-oriented applications rather than speculation. Expect moderately dense technical content — mining will likely mean skimming chapter structures and targeted extraction of application-relevant sections (speech, psychoacoustics, audio systems) rather than reading linearly.

Reading Goal

Prioritizing: Causal Relationships (Understanding applications requires seeing how acoustic signals cause behaviors/outcomes), Factual Claims (Needs concrete facts about who uses acoustic communication and how), Connection Points (Mapping acoustics to diverse application domains (tech, biology, medicine) is exactly the goal), Definitional Content (Needs clear definitions of acoustic communication concepts to build understanding) De-prioritizing: Contrarian or Surprising Claims (Applications focus is pragmatic; surprises less central but notable), Inspiration and Speculation (Future possibilities useful but secondary to understanding existing applications)

Thesis

Sound quality judgments depend on listener context and expectations · ch. 68

Argument

1. Sound technology work needs acoustics, perception, and signal processing jointly. — ch. 0
2. Hearing serves both environmental orientation and communication since early animal evolution — ch. 0
3. Echolocating bats build image-like environmental representations from reflected chirps — ch. 0
4. Sound storage digitized fully but human ears force analog endpoints — ch. 0
5. Sound cannot persist in time because it leaves no physical trace — ch. 0
6. Digital audio’s rise made auditory perception knowledge essential to engineers — ch. 0
7. Noise is defined by annoyance, disturbance, or hearing harm rather than loudness alone — ch. 0
8. Vision dominates audition except where visual events are absent or ambushed — ch. 0
9. Voice has become a natural interface for human-machine communication — ch. 0
10. Voice-based computer interfaces required granting machines language processing and world knowledge — ch. 0
11. Noise is defined as sound that annoys, disturbs, or risks hearing damage — ch. 0
12. Sensory efficiency has driven animal evolution and survival — ch. 0
13. Natural and engineered systems organize as vertical control hierarchies — ch. 1
14. Engineering perspectives increasingly serve computational modelling of auditory perception. — ch. 1
15. Internal models of reality must trade completeness for tractability. — ch. 1
16. Research advances only through sustained iteration across abstraction levels — ch. 1
17. Machines lead humans in logic tasks but lag in pattern recognition. — ch. 1
18. Pythagoras established the inverse relation between pitch and string length — ch. 1
19. Communication acoustics joins basic research and engineering toward shared goals — ch. 1
20. Multidisciplinary problems cannot be mastered within a single knowledge domain — ch. 1
21. Computational modelling is merging experimentation and theory into one method — ch. 1
22. Computer simulation dramatically expanded modeling’s power over physical models — ch. 1
23. The human auditory system is the hardest link in the communication chain. — ch. 1
24. Very general theories guide thinking but resist direct application to specific cases — ch. 1
25. Abstracting away details and concretizing them frame scientific research as one cycle — ch. 1
26. Systems are analyzed through complementary structure and function concepts — ch. 1
27. Evolution produced language and engineering, but machines still lag in pattern recognition — ch. 1
28. Physical, biological, and engineered systems share describable general properties — ch. 1
29. Heterarchies distribute control for versatility while data become information through observation — ch. 1
30. A system’s output depends on its remembered history — ch. 1
31. Heterarchical, distributed control can outperform strict hierarchy under changing conditions. — ch. 1
32. Similar-looking phenomena across domains still need domain-specific scientific concepts. — ch. 1
33. The human auditory system is communication acoustics’ hardest link to model — ch. 1
34. Audio and speech technology design should be grounded in perceptual mechanisms — ch. 1
35. A mass on a spring forms the simplest vibrating system — ch. 2
36. Sound as physical event and sound as auditory perception are distinct — ch. 2
37. Most sound arises from vibrating mechanical objects in a medium — ch. 2
38. Sound’s physical force is negligible yet its informational effects matter greatly. — ch. 2
39. Room reflections shape how well speech communicates across a space. — ch. 5
40. Resonance can be engineered to strengthen sound or damped to prevent malfunction — ch. 5
41. Helmholtz resonance shapes violin bodies, speaker designs, and car cabin noise — ch. 5
42. Bending stiffness makes bar vibrations dispersive and inharmonic. — ch. 5
43. Impedance discontinuities split an arriving wave into reflected and transmitted parts — ch. 5
44. Solids support shear forces and thus transversal waves, unlike fluids. — ch. 5
45. Relative motion between a sound source and listener shifts perceived pitch — ch. 5
46. Noise barriers only effectively block high-frequency traffic sound, not low frequencies — ch. 5
47. Sound intensity is a vector describing the directional flow of sound energy. — ch. 5
48. Sound pressure is tiny pressure deviation, and human hearing spans a specific Pa range — ch. 5
49. Sound diffraction arises when a wavefront’s edge acts as a secondary source — ch. 5
50. Tube end conditions determine which harmonic partials an instrument produces — ch. 5
51. Rigid-wall tube terminations fully reflect waves while open ends reflect with phase inversion. — ch. 5
52. Any wave can be decomposed into two counter-travelling component waves. — ch. 5
53. Coherent sounds interfere by amplitude while incoherent sounds sum by power. — ch. 5
54. Opposing waves do not annihilate because their energy just shifts form. — ch. 5
55. Stiffness makes transversal waves dispersive and their modal tones inharmonic — ch. 5
56. Compressible air in a volume forms a mechanical spring for resonance — ch. 5
57. A vibrating string has infinitely many resonant frequencies — ch. 5
58. Resonance can amplify desired sound or cause harmful machine vibration — ch. 5
59. The Helmholtz resonator models confined air as a spring with an air-mass plug — ch. 5
60. Vibration generates acoustic waves while sound radiation drains vibration energy — ch. 5
61. Coupled masses vibrate either along or perpendicular to spring direction — ch. 5
62. Opposing traveling waves of the same frequency create standing waves with nodes. — ch. 5
63. Coupled oscillators produce complex resonances shared by instruments, voices, and rooms — ch. 5
64. Any discontinuity in a medium reflects an incident wavefront — ch. 5
65. Coherent sound addition yields double the amplitude gain of incoherent addition — ch. 5
66. The absorption coefficient measures the fraction of incident sound energy absorbed. — ch. 5
67. Any vibration decomposes into weighted sums of a system’s modal oscillations — ch. 5
68. Phase alignment determines whether sounds sum by amplitude or by power — ch. 5
69. Tube termination conditions determine which harmonics a wind instrument produces — ch. 5
70. Early reflections boost perceived loudness while loud late reverberation degrades intelligibility — ch. 6
71. In steady state, reverberant sound level is uniform throughout a room. — ch. 7
72. Sound from a source falls 6 dB per distance doubling until reverberation takes over — ch. 7
73. Room sound decay is exponential, appearing linear on a dB scale — ch. 7
74. Hard-walled rooms behave as three-dimensional resonators with many audible modal frequencies — ch. 8
75. Beyond the reverberation distance, loudness stays constant while reverberance increases. — ch. 8
76. A concert hall’s acoustic field has billions of vibrational degrees of freedom. — ch. 9
77. Room reflections can be modelled as sources in mirrored rooms — ch. 9
78. Geometric acoustics fails on diffraction at low frequencies in complex rooms — ch. 9
79. Element-based acoustics methods grow computationally costly at higher frequencies — ch. 9
80. Modal density scales with spatial dimension, from flat in 1D to squared in 3D. — ch. 9
81. Ray-tracing and image-source methods fail on low-frequency diffraction at corners — ch. 9
82. Room acoustics can be simulated by statistically tracing sound rays until their energy decays — ch. 9
83. Above a critical frequency, room acoustics become statistically tractable as diffuse fields. — ch. 9
84. Non-linear acoustics governs high-pressure waves like shocks and cavitation — ch. 9
85. Converting continuous sound to digital requires time sampling then numeric quantization — ch. 10
86. Time-frequency transforms make signal processing and interpretation easier — ch. 10
87. Loudspeakers at high power and low frequencies become highly nonlinear. — ch. 10
88. Any continuous signal decomposes into sinusoidal partials and convolution simplifies in frequency domain — ch. 10
89. The Fourier transform measures how a signal matches each sinusoid pair — ch. 10
90. The Fourier transform trades costly time-domain convolution for simple frequency-domain multiplication — ch. 10
91. Physically realizable systems cannot respond to future input values — ch. 10
92. A sound signal is the registered time-varying vibration of a medium — ch. 10
93. Computer-displayed continuous-looking curves are actually stored sample sequences — ch. 10
94. Non-linear and time-variant systems generate frequency components absent from the input — ch. 10
95. Human hearing resists linear time-invariant modeling except at the periphery — ch. 10
96. LTI systems cannot create new frequency components from their input signals — ch. 10
97. Frequency-domain analysis mirrors how the human ear processes sound — ch. 11
98. Signal cropping is equivalent to applying a rectangular window — ch. 11
99. The ear’s analysis resembles Fourier frequency decomposition of sound — ch. 11
100. Group delay of modulation is the phase representation most relevant to hearing — ch. 11
101. The auditory system largely ignores signal phase, driving spectral analysis conventions — ch. 11
102. Undersampled signals alias because high-frequency components mirror below the Nyquist frequency — ch. 13
103. Standard audio sampling rates range up from 8kHz speech to 192kHz ultra-high fidelity — ch. 13
104. Autocorrelation reveals a signal’s fundamental periodicity through periodic peaks. — ch. 13
105. Wavelet and constant-Q transforms trade time and frequency resolution non-uniformly — ch. 13
106. Filter banks split broadband signals into sub-bands much like cochlear hearing does — ch. 13
107. Short-time Fourier analysis converts signals into spectrograms via windowed transforms — ch. 13
108. Digital signal processing is inherently band-limited by reconstruction filters — ch. 13
109. Each added quantization bit yields 6dB SNR improvement — ch. 13
110. Filter banks drive equalization, audio coding, and spatial reproduction — ch. 13
111. LP inverse filtering flattens signals into spectrally whitened residuals — ch. 14
112. Any band-limited linear time-invariant system can be implemented as a digital filter — ch. 14
113. Autocorrelation is the dominant method for linear prediction analysis — ch. 14
114. Linear prediction models signals as all-pole IIR systems predicting samples from prior ones — ch. 14
115. Air absorption and room acoustics can be modeled as low-pass and arbitrary filters — ch. 14
116. Feedback (IIR) filters can become unstable, while feedforward (FIR) filters cannot — ch. 14
117. Speech recognition picks the model maximizing the probability of observed features — ch. 15
118. Adaptive filtering changes filter coefficients over time in response to signals. — ch. 15
119. The similarity between artificial and biological neural networks is only superficial. — ch. 15
120. Linear prediction order scales with sample rate to capture speech formants. — ch. 15
121. Real-world signals change over time, so fixed filters need adaptation — ch. 15
122. Electret microphones trap charge in material so no external voltage is needed — ch. 16
123. Condenser microphones dominate modern audio because they convert sound to electrical signals efficiently. — ch. 16
124. A microphone’s sensing principle determines whether its pickup pattern is omnidirectional, dipolar, or cardioid — ch. 16
125. Perfect sound reproduction tolerates arbitrary signal delay when reproducing recordings. — ch. 16
126. Room impulse response measurements cannot be made with true perfect impulses from practical sources. — ch. 16
127. Slight response deviations go unnoticed without an audible reference — ch. 16
128. Enclosures prevent acoustic short-circuiting that silences low frequencies — ch. 16
129. Most audio reproduction chain components are now perceptually flawless — ch. 16
130. Ideal audio reproduction permits gain change and pure delay without degrading quality — ch. 16
131. Electrostatic transduction, the condenser microphone’s inverse, delivers the highest headphone and speaker quality. — ch. 16
132. Loudspeakers and room acoustics remain the weak link in audio quality — ch. 16
133. A good microphone needs flat response from 20 Hz to 20 kHz. — ch. 16
134. Reverberant rooms make loudspeaker directivity important for sound quality — ch. 16
135. Sound reinforcement permits only milliseconds of delay while recorded reproduction tolerates seconds — ch. 16
136. The condenser microphone and its electret variant dominate modern microphone technology — ch. 16
137. Closed headphones gain isolation but become sensitive to leakage effects. — ch. 16
138. Headphones come in intra-aural, supraaural, and circumaural form factors. — ch. 16
139. Headphone cushion leakage degrades low-frequency response but spares mid and high frequencies. — ch. 16
140. Only part of loudspeaker input power becomes sound energy — ch. 16
141. The technically precise term for frequency response is magnitude response — ch. 16
142. Perfect sound reproduction requires an LTI system with unity transfer function — ch. 16
143. Loudspeaker-room interactions remain the weak link in audio reproduction chains — ch. 16
144. Real loudspeakers show flat magnitude response only below 800 Hz with a midrange dip — ch. 16
145. Headphone cushion leaks alter only the low-frequency transfer to the eardrum — ch. 16
146. Reverberant rooms favor directional loudspeakers to control sound colouration — ch. 16
147. A plus or minus 5 dB loudspeaker deviation goes unnoticed without a direct reference — ch. 16
148. Headphones originated as noise isolation for radio operators — ch. 16
149. Noise is defined by its disturbing effect relative to a desired signal — ch. 17
150. Objective measures based on hearing models substitute for tedious listening tests of distortion. — ch. 17
151. Signal processing can compensate for loudspeaker and headphone imperfections. — ch. 17
152. Equalizing a loudspeaker response can worsen off-axis sound quality. — ch. 17
153. Equalizers invert measured system impulse responses to flatten magnitude response — ch. 17
154. Even ideal equalization degrades off-axis response and can amplify distortion and noise — ch. 17
155. Total harmonic distortion measures all components a fundamental generates, including non-harmonic ones. — ch. 17
156. Nonlinear equalization methods remain underdeveloped in audio engineering — ch. 18
157. Linear audio distortion is easy to equalize while non-linear distortion resists compensation — ch. 18
158. Average speaking fundamental frequency is 120 Hz for men and 200 Hz for women — ch. 19
159. A full speech communication chain spans brain, vocal organs, channel, and hearing. — ch. 19
160. The glottal waveform can be estimated indirectly through inverse modeling of microphone recordings — ch. 19
161. Vowels keep the vocal tract open and vary little by context, unlike consonants — ch. 19
162. Phonetic information concentrates either in a phone’s steady core or its transition boundary — ch. 19
163. Unvoiced speech arises from turbulence at constrictions or sudden pressure releases — ch. 19
164. Speech segments are classified by abstract classes, concrete sounds, and transition units — ch. 19
165. All vowel distinctions reduce to tongue position, openness, and lip rounding — ch. 19
166. Consonants differ from vowels in intensity, context-dependence, and articulation constriction — ch. 19
167. Languages differ widely in vowel category counts, from two to over ten. — ch. 19
168. Phonation types differ by glottal airflow and spectrum shape — ch. 19
169. The International Phonetic Alphabet attempts a universal notation for every speech sound — ch. 19
170. Voiced speech originates from airflow through the vibrating vocal folds across the glottis — ch. 19
171. Speech is a continuous flow that still requires discrete units to be understood — ch. 19
172. Speech sound generation was mechanically modeled as early as the 18th century — ch. 19
173. Speakers shift glottal closure timing to control the spectral brightness of their voice. — ch. 19
174. Speech is uniquely human among species’ communication abilities. — ch. 19
175. Articulation shapes the voice by filtering glottal sound through the vocal tract — ch. 19
176. Speech is the original form of linguistic communication, predating writing. — ch. 19
177. Languages range from two to over ten vowel categories — ch. 19
178. Computerized speech handling required simplified phonetic notations tailored per language — ch. 19
179. Vocal tract length differs by sex and its resonances form speech formants. — ch. 19
180. The vocal apparatus functions as a pressure source driving a resonant filter system — ch. 20
181. Speech production maps onto a three-stage cascaded electrical circuit analogy — ch. 20
182. In tonal languages like Chinese, pitch changes alter the phoneme itself — ch. 20
183. Vocal tract acoustics reduce to a cross-sectional area function along one axis — ch. 20
184. Voiced and unvoiced speech differ in whether source is glottal oscillation or frication noise — ch. 20
185. Consonant classes are produced by distinct vocal tract closure configurations — ch. 20
186. Languages vary in whether timing or syllable quantity carries meaning — ch. 20
187. Acoustic reflections arise at any change in tube cross-sectional area — ch. 20
188. Speech production reduces to lungs as compressor driving a glottal oscillator through resonating tracts — ch. 20
189. Speakers vary pitch to signal emotion, questions, and focus. — ch. 20
190. Prosody denotes speech features not tied to segmental structure — ch. 20
191. Speakers use pitch variation to signal statements versus questions — ch. 20
192. Glottal closing generates most vocal-tract excitatory energy. — ch. 20
193. The glottal waveform’s closing phase is abrupt and carries most excitatory energy. — ch. 20
194. Plosive consonants begin with silence then a burst when the tract opens. — ch. 20
195. Mathematical source-filter models of speech enable practical synthesis and coding algorithms — ch. 20
196. The vocal tract can be simulated as a cascade of acoustic transmission line segments — ch. 21
197. High-pitched soprano singing is unintelligible because harmonics are too sparse for formants — ch. 21
198. Formant tracking to harmonics boosts loudness, not intelligibility, in soprano singing — ch. 21
199. Overtone singing selects harmonics with sharp vocal-tract resonances to create melodies — ch. 21
200. Sopranos retune formants to match harmonics to boost loudness, not intelligibility — ch. 21
201. Articulatory synthesis models speech dynamics but is too complex for practical synthesizers — ch. 21
202. A unified source-filter model describes human and synthetic voice alike — ch. 21
203. Singing and speech differ systematically in pitch control, spectrum, vibrato, and range — ch. 21
204. Any phone of a language can be synthesized with 4-5 parallel formant filters — ch. 21
205. Instruments spanning multiple sound-generation classes force contested classification — ch. 22
206. Music and speech are utility sounds distinguished by their primary role — ch. 22
207. Instrument sounds are endlessly varied yet fall into families sharing generation principles — ch. 22
208. A single amplitude envelope misses that harmonics decay and modulate at different rates — ch. 24
209. Musical instrument sounds divide into attack, decay, sustain, and release phases — ch. 24
210. Hearing is highly sensitive to temporal changes in sound — ch. 24
211. Resonators color spectral and temporal content of transmitted sound. — ch. 24
212. Most real instruments are nonlinear and resist simple filter-cascade modelling — ch. 24
213. Resonance frequencies produce longer build-up and decay than off-resonance — ch. 24
214. Fretted instruments set pitch by vibrating string length — ch. 24
215. Piano hammers rebound and strike strings multiple times per note — ch. 24
216. Idiophone and most membranophone pitch cannot be controlled by the player — ch. 24
217. Plucked strings vibrate autonomously after excitation, simplifying their modelling. — ch. 24
218. Bowed string motion is a stick-slip effect producing a sawtooth wave — ch. 24
219. Speech tracts lack coupling that brass instruments synchronize with. — ch. 24
220. Musical instruments couple excitation to resonances; human speech does not — ch. 24
221. Aerophone design tunes resonance modes so the instrument becomes musically usable — ch. 24
222. FM synthesis generates harmonic spectra when the modulating frequency equals the carrier frequency — ch. 25
223. Placing all string damping at terminations yields efficient digital synthesis — ch. 25
224. Natural background sounds are synthesized statistically for virtual environments — ch. 25
225. Continuous-excitation instruments like flutes attack slowly while impulse-excited ones do not — ch. 25
226. A guitar body’s effect can be simulated by convolution with its measured impulse response — ch. 25
227. Computational models can synthesize natural ambient sounds like footsteps and thunder — ch. 25
228. Hearing is a staged chain from air capture through mechanical conduction to neural processing — ch. 26
229. The human auditory system evolved specifically for spoken-language analysis — ch. 26
230. Human hearing diverged from other mammals mainly via speech analysis — ch. 26
231. Auditory nerve responses can be experimentally recorded by microelectrodes. — ch. 27
232. The ear emits measurable echoes requiring active neural involvement — ch. 27
233. Loud sounds spread cochlear activation broadly and respond faster than soft sounds — ch. 27
234. Stronger tones in the cochlea actively mask neighboring weaker tones via two-tone suppression. — ch. 27
235. Cochlear pitch mapping exploits the mass of vibrating liquid varying with route length — ch. 27
236. Two tones presented together generate an audible combination (difference) tone not physically present. — ch. 27
237. Pinna asymmetry creates front-back and elevation cues for directional hearing. — ch. 27
238. Single auditory nerve fibres encode only a narrow 20-40 dB dynamic range. — ch. 27
239. Head shading creates interaural level differences used for sound localization — ch. 27
240. The cochlea’s sharp frequency selectivity needs an unexplained second filter. — ch. 27
241. The Eustachian tube pressure-balances the middle ear against altitude changes — ch. 27
242. The middle ear evolved as an impedance transformer to bridge air-water sound energy loss — ch. 27
243. Middle-ear impedance matching boosts pressure transfer roughly 30-fold — ch. 27
244. The ear canal acts as a quarter-wavelength resonator amplifying 3-4 kHz by 10 dB — ch. 27
245. Single auditory neurons encode only a narrow 20-40 dB dynamic range — ch. 27
246. A competing tone at another frequency suppresses a cochlea channel’s response — ch. 27
247. Hair cells actively amplify cochlear vibration rather than only sensing it — ch. 27
248. Cubic difference tones reveal nonlinear cochlear mechanics even at low sound levels — ch. 27
249. The auditory periphery is non-linear and cannot be captured by one transfer function. — ch. 27
250. Ion concentration differences across the basilar membrane create an electric potential. — ch. 27
251. The cochlear nucleus functions like a preamplifier and initial spectral analyzer — ch. 28
252. Acoustic engineering usually needs only cochlear models, not higher nuclei — ch. 28
253. Population coding across many neurons achieves precise and robust auditory processing — ch. 28
254. Auditory models link acoustic stimulus features to perceived attributes computationally. — ch. 28
255. Auditory nerve phase locking degrades above 1500 Hz and is weak above 4 kHz — ch. 28
256. Loudness above 60 dB is encoded by cross-frequency spread, not single-neuron rate — ch. 28
257. fMRI identifies active brain regions well but is noisy and time-slow for auditory study. — ch. 28
258. Single neuron auditory responses are mapped through controlled stimulus repetition — ch. 28
259. Computational auditory models serve knowledge, audio-quality measurement, and diagnostics — ch. 28
260. Bottom-up neuron-by-neuron auditory modelling is computationally and knowledge-wise infeasible. — ch. 28
261. The superior colliculus is the only mammalian brain nucleus with a topographical auditory map — ch. 28
262. Single-neuron auditory responses are mapped invasively in animals only — ch. 28
263. Speech remains understandable at high levels because timing carries formant info — ch. 28
264. Humans cannot reliably report auditory attributes on absolute internal scales. — ch. 29
265. Constant stimuli remains gold standard but is too slow for routine use — ch. 29
266. Simple stimuli fail to represent how hearing processes real complex sounds — ch. 29
267. Tracking audiometry maps hearing thresholds but ignores response bias — ch. 29
268. Trained listeners improve test sensitivity but misrepresent general audience opinion — ch. 29
269. Psychophysical scales are built by adjusting sounds to match reference ratios — ch. 29
270. Listener personality biases reported auditory thresholds across subjects — ch. 29
271. Valid psychoacoustic listening tests require blind conditions and carefully worded tasks. — ch. 29
272. Game-like testing yields less biased hearing-aid results in children than formal tasks. — ch. 29
273. Introspection alone cannot serve as accepted scientific auditory observation — ch. 29
274. Thresholds and auditory scales link physical sound attributes to perceived auditory events — ch. 29
275. Psychoacoustics developed largely through communication technology challenges — ch. 29
276. Thresholds can be measured by tracing stimulus change until perception flips — ch. 29
277. Psychoacoustics studies higher auditory functions where physiology is unknown — ch. 29
278. Relative magnitude judgments can be anchored into absolute psychophysical scales — ch. 29
279. Psychophysical functions map continuous physical sound scales onto varied perceptual scales. — ch. 29
280. Perceived-loudness matching often needs separate listening tests to achieve — ch. 29
281. Constant-stims method maps full psychophysical functions rather than single thresholds — ch. 29
282. Signal detection theory originated in WWII radar plane detection — ch. 29
283. Trained listener panels trade real-world representativeness for analytic test resolution — ch. 29
284. Vision dominates hearing in localization, identification, and size judgment — ch. 29
285. Forced-choice guessing floors push threshold criterion above 50 percent detection — ch. 29
286. Paired comparison against anchors reduces bias in absolute judgments — ch. 29
287. Uncontrolled loudness differences confound audio quality listening-test results — ch. 29
288. Perceptual thresholds blur because listener internal noise varies by personality — ch. 29
289. Anechoic chambers misrepresent how loudspeakers sound in normal rooms — ch. 29
290. Test method means the logic selecting sound attributes across successive tasks — ch. 29
291. Auditory perception must be studied indirectly through measurable external reactions. — ch. 29
292. Auditory experimentation resembles measuring an unreliable device against an unpredictable system — ch. 29
293. Ascending and descending stimulus runs are averaged to derive perceptual thresholds — ch. 29
294. Mean Opinion Score scales standardize audio-quality evaluation on a 1-to-5 range — ch. 29
295. Perceived loudness depends on frequency content and duration, not just sound pressure level — ch. 29
296. Adjustment steps sized near the JND prevent subjects overestimating imperceptible changes — ch. 29
297. The method of limits is inherently prone to response bias — ch. 29
298. Analysis of variance tests whether any independent variable affects a dependent variable. — ch. 30
299. Perceptual structure analysis indirectly recovers individual feature-usage in auditory perception. — ch. 30
300. Repertory grid technique lets each subject generate their own descriptive dimensions via triadic comparisons. — ch. 30
301. Audio codec testing requires methods that reveal unknown perceptual attributes — ch. 30
302. Multidimensional scaling maps perceived sound dissimilarities without labeling the perceptual dimensions — ch. 30
303. Consensus-based sensory panels risk leader bias but training can counter it — ch. 30
304. Free sorting adds post-hoc verbal labels that enable interpreting perceptual dimensions. — ch. 30
305. Adaptive staircase procedures trade efficiency and bias avoidance against convergence risk — ch. 30
306. Verbal elicitation dominates sensory science via consensus or individual vocabularies — ch. 30
307. Descriptive sensory analysis maps perceived sound differences onto an N-dimensional attribute space — ch. 30
308. Careful listening tests validly reveal how acoustic attributes become auditory attributes. — ch. 30
309. Sound attributes are independent variables manipulated to elicit dependent responses. — ch. 30
310. Sensory scales can be calibrated for equal intensity across different attributes — ch. 30
311. Paired comparison lets flash profiling skip time-consuming attribute training. — ch. 30
312. Subjective listening tests map manipulated acoustic variables to measured responses. — ch. 30
313. Listening-test statistics must inform experimental design from the outset — ch. 30
314. Generalized procrustes analysis aligns different subjects’ attribute ratings into common dimensions. — ch. 30
315. Assessors can use their own vocabulary instead of agreeing on descriptors — ch. 30
316. Pointing gestures can localize sound sensations where words fall short — ch. 30
317. The cochlea splits sound into narrowband signals that the brain processes individually — ch. 31
318. The hearing threshold is measured from sounds reaching the subject’s binaural position. — ch. 31
319. Sound audibility depends on both its level and its frequency — ch. 31
320. Practical speech communication uses only a small slice of hearing’s full range — ch. 31
321. Human hearing is most sensitive to sounds around 3-4 kHz. — ch. 31
322. Typical one-metre speech at 60-70 dB is optimally tuned for auditory communication — ch. 31
323. Uniform masking noise must drop 3 dB per octave above 500 Hz. — ch. 31
324. Hearing’s lower limit is the minimum audible field; upper is pain — ch. 31
325. The phon scale defines loudness relative to equal-loudness matches at 1 kHz — ch. 31
326. Loudness is measured subjectively against 1 kHz reference tones — ch. 31
327. One sound can mask another even when their spectra do not overlap. — ch. 31
328. A-weighted decibels are a rough practical approximation of perceived loudness — ch. 31
329. Spectral masking by loud sounds is exploited to design lossy audio codecs. — ch. 31
330. Tone-on-tone masking can be detected through beating and combination tones alone — ch. 32
331. Coherent amplitude modulation across frequency bands may drive auditory source grouping. — ch. 32
332. Backward masking is a weak and unreliable effect appearing mainly in novice listeners. — ch. 32
333. Forward masking suppresses sounds for 150-200 ms after masker offset even at high levels. — ch. 32
334. Hearing groups frequency bands by shared envelope modulation, reducing masking — ch. 32
335. Complex real-world maskers are less effective than simple energetic ones predict — ch. 32
336. Sounds mask both preceding and following sounds but forward masking dominates. — ch. 32
337. Forward masking decays within roughly 200 ms after sound offset — ch. 34
338. Each inner hair cell responds best to one frequency plus a widening critical band — ch. 34
339. Loudness stays constant below critical bandwidth then grows beyond it — ch. 34
340. Psychoacoustic tuning curves directly reflect cochlear mechanical processing. — ch. 34
341. Hearing segregates sounds by frequency unless they fall within a critical band — ch. 34
342. Longer masker sounds produce stronger forward masking — ch. 34
343. Wider bandwidth increases perceived loudness at equal sound pressure — ch. 35
344. The Bark scale is linear below 500 Hz and logarithmic above it. — ch. 36
345. Human spectral analysis resolution runs about 10–15 percent of center frequency — ch. 36
346. A 10-15 percent bandwidth window serves spectral analysis for audio tech — ch. 36
347. Notched-masker tests prevent off-frequency listening to measure filter width — ch. 36
348. Auditory bands from the same source are summed to compute total loudness — ch. 36
349. The ear’s 120-dB and 20 Hz-20 kHz working range exceeds man-made devices — ch. 37
350. Pitch discrimination is best near 500 Hz and degrades at both extremes — ch. 38
351. Pitch is the perceptual property ordering sounds from low to high — ch. 38
352. Pitch cues convey identity, intonation, and emotion beyond mere speech content — ch. 38
353. Pitch is not uniquely tied to a single physical frequency — ch. 38
354. Narrowband noise can carry clearer pitch than some periodic signals — ch. 38
355. Frequency JND is best around 500 Hz at about 0.3 percent — ch. 38
356. Musical pitch perception spans roughly 30 Hz to 4 kHz. — ch. 38
357. Pitch perception requires a tone lasting at least two signal periods — ch. 39
358. Listeners perceive the fundamental frequency even when its lowest partials are absent — ch. 40
359. Western orchestral tuning pitch has risen historically for a brighter sound — ch. 40
360. A set of harmonically related partials shifted together still changes perceived pitch. — ch. 40
361. Inharmonic sounds like bells can produce multiple simultaneous pitches — ch. 40
362. The mel scale was derived by asking listeners to double perceived pitch — ch. 40
363. Outdoor wall and ground reflections produce repetition pitch in real settings. — ch. 40
364. Place theory cannot fully explain how pitch is extracted from components — ch. 40
365. Louder signals need disproportionately more dB to double perceived loudness — ch. 40
366. The phon defines loudness level equal to dB SPL at 1 kHz — ch. 40
367. Masking noise raises the loudness threshold and fades audible tones — ch. 40
368. Pitch perception involves both spectral and temporal representations of sound — ch. 40
369. Summing a signal with its delayed copy creates perceived pitch. — ch. 40
370. Loudness perception sums auditory bands and signals distance by comparison — ch. 40
371. Absolute pitch is roughly ten times more precise than trained relative pitch. — ch. 40
372. Humans perceive a fundamental frequency even when it is absent from the signal — ch. 40
373. Doubling perceived loudness requires a 10 dB increase. — ch. 40
374. Human hearing excels at comparative judgments but fails at absolute ones. — ch. 40
375. Amplitude modulation of noise creates pitch with a flat spectrum — ch. 40
376. Loudness is perceptual while intensity is physical and must be distinguished — ch. 40
377. Frequency modulation detection depends on both carrier and modulation frequency. — ch. 40
378. Place and timing theories fail to explain some complex pitch perception phenomena — ch. 40
379. Noise with a sharp spectral peak is perceived as pitched. — ch. 40
380. Doubling perceived pitch of a 2-kHz tone requires about a 15-kHz frequency jump — ch. 40
381. Perceived loudness can be measured by iterative ratio-matching across subjects — ch. 40
382. Shifting harmonic partials upward raises perceived pitch despite unchanged envelope modulation — ch. 40
383. Auditory loudness doubles with every 10 dB increase in sinusoid level — ch. 41
384. Frequency spread in the cochlea is asymmetric with steeper low-frequency slopes — ch. 41
385. The basilar membrane’s fixed-spread frequency resolution shrinks relative bandwidth as frequency rises — ch. 41
386. Perceived loudness of concurrent tones hinges on frequency separation via three distinct mechanisms — ch. 41
387. Humans detect level changes of about 1 dB at moderate volumes — ch. 41
388. AM tones change perceived frequency spread, giving extra detection cues that noise lacks. — ch. 41
389. Amplitude-modulation detection improves with tone level but is level-independent for noise. — ch. 41
390. Amplitude-modulation detection becomes easier once modulation spreads the spectrum beyond one critical band — ch. 42
391. Hearing forms loudness by integrating sound energy over about 100-200 ms — ch. 42
392. Short sounds decay in perceived loudness faster than long sounds. — ch. 43
393. Time-varying sound loudness is set by integrated loudness peaks, not moment-to-moment level — ch. 43
394. Human hearing can in principle distinguish astronomically many timbres. — ch. 43
395. Humans can match durations of silence to sounds of similar length. — ch. 43
396. Temporal masking makes silence following a sound seem shorter — ch. 43
397. Timbre distinguishes sounds sharing pitch, loudness, and duration. — ch. 43
398. A sound’s onset transients largely determine instrument identity — ch. 43
399. Brief silences are perceptually longer than their physical duration. — ch. 43
400. Perceived sharpness is stable across subjects and rises with spectral centre of gravity — ch. 44
401. Hearing repackages one-dimensional sound signals into multidimensional time-frequency objects — ch. 44
402. Sound level has only a weak effect on perceived sharpness — ch. 44
403. Fluctuation strength is standardized against a 4 Hz amplitude-modulated reference tone — ch. 45
404. Modulation frequency determines whether sound is heard as fluctuation, roughness, or multiple tones — ch. 45
405. Slow sound-level changes are heard as fluctuation and fast ones as roughness — ch. 45
406. Amplitude-modulated loudness directly increases perceived fluctuation strength — ch. 47
407. Fast amplitude modulations within a critical band produce roughness perception — ch. 47
408. Just intonation leaves many intervals audibly mistuned like the D-A fifth — ch. 47
409. Reversing a signal’s polarity changes its perceived bass level — ch. 47
410. Auditory cortex specifically attenuates responses to frequently repeated sounds — ch. 47
411. Arbitrary-sound roughness is estimable by summing per-critical-band contributions — ch. 47
412. Western major scales consist of five whole tones and two semitones in fixed arrangement — ch. 47
413. The brain organizes simultaneous sound events into separate auditory streams via auditory scene analysis. — ch. 47
414. Impulsiveness is estimated from level statistics rather than directly measured — ch. 47
415. Individual notes are assigned to exactly one auditory stream, altering perceived rhythm. — ch. 47
416. Interval dissonance increases with position higher in the harmonic series — ch. 47
417. Gestalt grouping principles apply to hearing as well as vision — ch. 47
418. Psychoacoustic tonality measures tone-character, distinct from musical tonality terms — ch. 47
419. Rhythm derives from natural body repetitions like heartbeat and walking. — ch. 47
420. Noise is masked more easily by tonal parts than tones are masked by noise — ch. 47
421. Auditory cortex adaptation decreases response to frequent sounds, shaping perception — ch. 47
422. No single standard method measures perceived tonality accurately — ch. 47
423. Auditory spatial information alone allows orientation in one’s surroundings. — ch. 47
424. Simple integer frequency ratios between notes produce the greatest measured consonance — ch. 47
425. Machine hearing mirrors human perception by comparing feature vectors against known patterns — ch. 47
426. Synchronized phase harmonics produce a ‘buzzy’ sharpness perceptible without room reflections — ch. 47
427. Headphone group delay alters audibility of phase-spectrum effects — ch. 47
428. Reversing the phase of a single harmonic can drastically change its perceived loudness — ch. 47
429. Hearing perceives synchronized neural firing across critical bands as buzzy timbre — ch. 47
430. No tuning system can make all diatonic intervals have simple ratios — ch. 47
431. Room reverberation can render phase-sensitive signals perceptually phase-insensitive — ch. 47
432. Musical eras are distinguished by their systematic handling of sensory dissonance — ch. 47
433. One second of exposure triggers neural spectral adaptation in hearing — ch. 47
434. Peak dissonance between two tones occurs at a quarter critical band separation — ch. 47
435. Impulsive transients grab attention by signaling state changes in physical objects — ch. 47
436. Group delay affects perceived quality only when it varies with frequency — ch. 47
437. Two close sine tones produce amplitude beats at their frequency difference — ch. 47
438. Perception is a brain interpretation that need not use all sensory detail — ch. 47
439. Humans can isolate a single speaker amid competing voices and noise — ch. 47
440. No tuning system can make all diatonic scale intervals perfectly consonant — ch. 47
441. Whether spectral adaptation dynamics are purely peripheral remains experimentally unresolved — ch. 47
442. Reverberation destroys the phase information that could encode meaning — ch. 47
443. Simultaneous cochlear neuron firing across frequencies produces buzzy timbre — ch. 47
444. Group delay perception thresholds depend strongly on signal content and context — ch. 47
445. Auditory perception runs parallel bottom-up and top-down processes — ch. 47
446. A single note can sound rough only below roughly 100 Hz fundamental — ch. 47
447. Perceptual hearing properties map onto objective signal analysis quantities — ch. 47
448. Musical notation is a lossy code requiring stylistic knowledge to realize — ch. 47
449. Humans are perceptually tuned to amplitude and frequency modulations across specific rate bands — ch. 48
450. Roughness between note partials determines perceived musical dissonance — ch. 48
451. Perceived musical dissonance reduces to roughness from beating partials. — ch. 48
452. HRTF definition normalizes ear responses by the head’s absence — ch. 49
453. Mannequin heads with ear microphones capture realistic spatial audio for recording — ch. 49
454. The outer ear’s resonant cavities enable elevation perception where other cues fail — ch. 49
455. Head, torso, and shoulder acoustics shape the localization cues reaching each ear. — ch. 49
456. Echoes and reverberation convey room geometry and surface properties to listeners — ch. 49
457. Sound localization evolved in nocturnal mammals and remains vital for prey and predator. — ch. 49
458. Spatial hearing is calibrated through learning to each listener’s head and pinna geometry — ch. 49
459. Spatial hearing is described in three distinct coordinate systems with different strengths. — ch. 49
460. Head-related transfer functions are measured with in-ear microphones in anechoic chambers — ch. 49
461. A dedicated cone-of-confusion coordinate system describes hard-to-discriminate sound directions. — ch. 49
462. The ear sits 8 cm off-center, so source-ear distance changes with azimuth — ch. 50
463. ITD spans 0-700 microseconds across azimuth independent of frequency — ch. 52
464. Interaural level differences and time shifts trade off to control lateralization — ch. 52
465. The head shadows and reflects sound to create level differences between the ears. — ch. 52
466. Human sound localization is remarkably accurate given it processes only two ear signals. — ch. 52
467. Auditory spatial perception emerges from weighted combination of localization cues — ch. 52
468. Humans can detect interaural timing differences as small as ten to twenty microseconds — ch. 52
469. Shoulder reflections create a distinguishable 1-ms delay in head-related impulse responses — ch. 52
470. Shoulder reflection delays versus direct sound vary inversely with source elevation — ch. 52
471. Low interaural coherence makes sound events seem diffuse and surrounding — ch. 53
472. A stationary tone’s frequency alone can shift perceived elevation in the median plane. — ch. 53
473. Low-frequency interaural level differences cue source distance and signal coherence. — ch. 53
474. Humans perceive interaural level differences at frequencies where plane waves produce none — ch. 53
475. Cone of confusion makes time and level cues ambiguous for elevation — ch. 53
476. The auditory system continually relearns to decode pinna-imposed spectral cues. — ch. 53
477. The ear achieves spatial hearing through signal analysis rather than directional receptors — ch. 54
478. Spectral-cue adaptation in hearing requires visible sound sources as a reference — ch. 54
479. Synchronized visual motion captures perceived sound direction within 30 degrees — ch. 54
480. Single-ear listeners can still localize sound sources but with degraded accuracy — ch. 54
481. Sound width perception depends on signal type, not physical source size — ch. 54
482. Humans adapt to a room’s spatial cues through learning over hundreds of trials — ch. 54
483. Two identical sounds within 30–40 ms are fused into one located auditory event. — ch. 54
484. Delayed identical sounds merge spatially until delays exceed 30-40 ms — ch. 54
485. Dynamic auditory cues can override clear visual evidence of source location — ch. 54
486. Evolution shaped localization to favor broadband transient sounds like movement — ch. 54
487. Neural refractory periods after strong onsets disable directional hearing for delayed sounds — ch. 54
488. The precedence effect suppresses spatial identity of reflections, not their audibility — ch. 54
489. Sound localization blur is smallest in front and largest at the sides — ch. 54
490. Sound localization fuses independent sensory evidence into one coherent percept — ch. 54
491. Headphone audio internalizes because static binaural cues signal an inside-the-head source — ch. 54
492. Dynamic binaural head-movement cues override conflicting visual evidence of source location — ch. 54
493. Binaural listening yields a 5-7 dB intelligibility advantage over monaural — ch. 54
494. Speech elements near masked threshold carry intelligibility only under binaural listening. — ch. 54
495. Binaural listening yields only 5–7 dB advantage over monaural spatially-separated sources — ch. 54
496. Separate sensory modalities fuse into one coherent perception unless cues conflict — ch. 54
497. Anti-phase signals vs diotic noise yield 18 dB binaural detection advantage around 200-300 Hz — ch. 55
498. Hearing evolved as an omnidirectional alarm system complementing directional vision — ch. 55
499. Instantaneous interaural level differences improve intelligibility beyond better-ear listening alone — ch. 55
500. The auditory system cancels room coloration to recover source signals — ch. 55
501. Perceived auditory distance has an acoustic horizon under 10 meters — ch. 55
502. Binaural listening gives 0 to 8 dB advantage in multi-talker conditions — ch. 55
503. Listeners perceptually subtract room coloration from source signals, called binaural decolouration. — ch. 55
504. More reverberation relative to direct sound makes listeners judge a source as farther away. — ch. 55
505. Perceived auditory distance follows a compressive power function of actual distance — ch. 56
506. Auditory models that extract distance cues from ear signals remain an open research direction — ch. 56
507. Reverberator pre-delay controls perceived source distance independently of loudness ratio — ch. 56
508. The auditory system compensates for room acoustics to recover the original source signal — ch. 56
509. Exaggerated interaural level differences signal that a sound source is near — ch. 56
510. Auditory distance localization error varies dramatically across studies. — ch. 56
511. DFT-based psychoacoustic models derive an auditory excitation spectrum on the ERB scale — ch. 57
512. Frequency masking is modelled as convolution with a spreading function — ch. 57
513. Real-time auditory models enable speech recognition, sound reproduction, and hearing aids — ch. 57
514. A complete mathematical model of the entire auditory system will never exist — ch. 57
515. Auditory spectra smooth harmonic detail while preserving formant structure — ch. 57
516. Hearing sensitivity to frequency can be emulated by filtering and weighting short-time power spectra — ch. 57
517. Computational auditory modelling began only in the 1960s — ch. 57
518. Psychoacoustic models explain experimental results without addressing physiology — ch. 57
519. Hearing acts as a spectral analyser extracting quantifiable perceptual metrics from sound — ch. 57
520. Gammatone filter banks model human auditory filtering using measured nerve responses — ch. 60
521. Combining DRNL paths produces a double impulse that loses temporal fine structure — ch. 60
522. Middle ear acts as a band-pass filter attenuating 6 dB per octave — ch. 60
523. Gammatone filters fail on level-dependence and fast-onset sounds — ch. 60
524. Filter bank models outperform DFT in mimicking auditory temporal resolution — ch. 60
525. Pitch perception can be modeled by autocorrelation across auditory filter bands — ch. 61
526. Adaptation, temporal integration, and post-masking can be simulated by one combined feedback element. — ch. 61
527. Hair-cell adaptation to sound arises from neurotransmitter pool depletion and recycling dynamics — ch. 61
528. Neurotransmitter loss in the synaptic cleft produces auditory adaptation. — ch. 61
529. The cochlear amplifier requires nonlinear, level-dependent elements to model. — ch. 61
530. Different hearing tasks exploit different model time constants. — ch. 61
531. A single auditory nerve fibre’s firing is unpredictable and noisy — ch. 61
532. Auditory nerve adaptation operates like automatic gain control. — ch. 61
533. Functional and physiological modeling approaches divide along intended use. — ch. 61
534. Clean auditory sensation emerges only from pooling many noisy nerve fibres — ch. 61
535. A single feedback process yields complementary onset-emphasis and loudness-integration outputs. — ch. 61
536. Cochlear modelling faces a tradeoff between spatial accuracy and computational cost — ch. 61
537. Loudness models need 100-200 ms time constants to capture temporal integration and post-masking. — ch. 61
538. Auditory adaptation acts like high-pass filtering that emphasizes stimulus onsets. — ch. 61
539. Pitch perception is encoded by the timing of neural firing in the cochlea — ch. 61
540. Cascade AGC loops compress steady signals logarithmically while amplifying onsets. — ch. 61
541. Robust spatial hearing despite reverberation has driven decades of binaural modeling research. — ch. 62
542. Jeffress coincidence detection explains how neurons encode interaural time differences. — ch. 62
543. Autocorrelation models explain pitch and can separate concurrent vowel sounds differing in fundamental frequency — ch. 62
544. Two competing loudness models predict time-variant loudness by peak tracking or frame averaging — ch. 62
545. Interaural time differences are computed by cross-correlating the two ear signals — ch. 62
546. The IACC maximum value estimates interaural coherence between the two ear signals — ch. 62
547. Count-comparison and coincidence-detection models imply different brain receptive-field architectures for direction coding. — ch. 63
548. Neural networks trained on single-subject data can accurately model human elevation localization — ch. 63
549. Equalization-cancellation subtracts masker signals to reveal hidden targets — ch. 63
550. Cross-correlation sidelobes near 1 kHz stem from interaural phase ambiguities — ch. 63
551. High-frequency auditory analysis preserves only the temporal envelope of signals — ch. 63
552. Binaural unmasking can be modelled as equalizing then subtracting two ear signals — ch. 63
553. Auditory modelling accuracy trades off sharply against computational cost — ch. 63
554. Two brain models differ on whether localization uses narrow or wide receptive fields — ch. 63
555. Close-miked mono reproduction stays clear because the room barely affects the signal — ch. 64
556. Visual color affects perceived loudness of the same sound — ch. 64
557. Spaced mics sound warm and ambient because incoherent signals exceed loudspeaker angle. — ch. 64
558. Audio quality improvements mask video defects, but video quality does not mask audio defects. — ch. 64
559. The ORTF technique blends coincident and spaced microphone characteristics — ch. 64
560. Required audio fidelity scales with the application’s communicative demands — ch. 64
561. First-order Ambisonics reproduction is restricted to three loudspeakers — ch. 64
562. Authentic sound reproduction is judged by perceptual identity, not signal fidelity — ch. 64
563. The listening experience intended by engineers is what gets conveyed, not raw sound — ch. 64
564. Human hearing adapts to different room acoustics to preserve timbre perception — ch. 64
565. Cinema has historically pioneered new audio reproduction formats before home use. — ch. 64
566. Three-dimensional VBAP extends pair panning to position sources on loudspeaker triangles — ch. 64
567. Sound reproduction’s unifying trait is its endpoint in the human listener — ch. 64
568. Sound within one critical band is perceived as a single source — ch. 64
569. Stereophonic two-channel reproduction became widespread after the two-channel record format arrived. — ch. 64
570. Cinema 5.1 surround sound originated in 1976 and reached homes in the late 1990s. — ch. 64
571. Audio engineering is the production of meaningful sound signals across media. — ch. 64
572. Loudspeaker signals can be computed as virtual microphone signals with first-order directional patterns. — ch. 64
573. Panning between non-symmetric loudspeaker pairs distorts perceived spatial width. — ch. 64
574. Human audio-visual synchronicity tolerance is asymmetric because sound naturally lags light. — ch. 64
575. Mono reproduction exaggerates recording-room coloration that stereo cancels — ch. 64
576. Object-based audio decouples soundtracks from specific loudspeaker layouts — ch. 64
577. Tactile vibration raises perceived loudness and lowers preferred bass levels — ch. 64
578. Tactile frequency discrimination is crude, often off by an octave — ch. 64
579. Quadraphonic sound failed due to delivery problems and too few speakers. — ch. 64
580. Directional coincident microphones suppress room reverberation, killing sense of space — ch. 64
581. Recorded audio is engineered for plausibility, not faithful reproduction of reality — ch. 64
582. Listening room specifications exist to make psychoacoustic test results comparable across sites — ch. 64
583. Spherical harmonics act as basis functions for building any microphone pattern — ch. 64
584. Audio production divides into recording, mixing, mastering, and live sound stages — ch. 64
585. Instrument directionality shapes the captured spectrum depending on mic placement — ch. 64
586. Auditory-tactile systems improve music reproduction quality and enable tactile feedback. — ch. 64
587. Spot miking captures each instrument with its own near microphone for independent channels — ch. 64
588. Standard Ambisonics microphones use four capsules to capture omnidirectional and three-dipole signals. — ch. 64
589. VBAP predicts perceived source azimuth accurately but not personal elevation — ch. 64
590. Physical accuracy of a sound field does not predict perceived accuracy — ch. 64
591. Surround recordings pair spaced mics one-to-one with each loudspeaker — ch. 64
592. Gain ratios alone cause loudness shifts; sum-of-squares normalization prevents them — ch. 64
593. Tactile-audio desynchrony is as critical as audiovisual desynchrony — ch. 64
594. The Blumlein XY pair captures reverberance fully but causes phase artefacts from side sound. — ch. 64
595. Room reverberation suppresses the comb-filter colouring of amplitude panning — ch. 64
596. Amplitude panning creates auditory objects with no real physical source — ch. 64
597. Room acoustics smear directional perception and mask recorded reverberation — ch. 64
598. No single ideal listening room acoustic exists because humans adapt to room differences — ch. 64
599. DirAC divides captured sound into diffuse and non-diffuse streams — ch. 64
600. DirAC reproduction quality depends on one dominant source per frequency band — ch. 64
601. FFT block convolution trades speed for double-block latency in reverbs — ch. 65
602. Headphone and loudspeaker playback of stereo differ yet yield similar spatial order. — ch. 65
603. Stereo loudspeakers create directional illusion through summed phase shifts mimicking time delays — ch. 65
604. Nonlinear spatial analysis distributes measured impulse responses across loudspeaker arrangements — ch. 65
605. Uncorrelated noise from surround speakers creates enveloping source perception. — ch. 65
606. Crosstalk-cancelled playback yields externalized sound only in the frontal hemisphere — ch. 65
607. Each person’s unique pinna and head shape undermines binaural recordings for other listeners — ch. 65
608. Regularization trades inversion accuracy against loudspeaker dynamic-range demands — ch. 65
609. Sound field control can deliver different audio to different zones of a room. — ch. 65
610. Computed room impulse responses often sound unnatural in the late reverberant tail — ch. 65
611. Microphone directivity shapes how reverberant a convolved signal sounds — ch. 65
612. Shoulder reflections imprint identifiable arcs on measured head-related impulse responses. — ch. 65
613. Headphone binaural reproduction fails without dynamic head-movement cues — ch. 65
614. Binaural audio reproduces ear canal signals from real or virtual scenarios. — ch. 65
615. Interaural level differences only emerge at frequencies above roughly 400 Hz. — ch. 65
616. Wave field synthesis is nearly always generated synthetically, not from real recordings. — ch. 65
617. DIY in-ear binaural recording yields convincing spatial reproduction — ch. 65
618. Sound source location is encoded in waveform distortions between the ears — ch. 65
619. HRTFs can virtualize multi-channel loudspeaker layouts on headphones. — ch. 65
620. Head-tracked HRTF filtering restores realistic external sound localization — ch. 65
621. Inverted-polarity (antiphasic) stereo signals cancel lows and smear image — ch. 65
622. Compression, limiting, and expansion attenuate or boost audio differently by amplitude. — ch. 65
623. Phase-based localization sensing has a hard frequency ceiling near 1.6 kHz — ch. 65
624. Wave field synthesis needs loudspeakers spaced at most half a wavelength apart. — ch. 65
625. Time-frequency modulation and amplitude modulation are distinct operations — ch. 66
626. MDCT exploits inter-frame aliasing cancellation to achieve critical sampling — ch. 66
627. Frequency-domain processing enables efficient audio coding for monophonic and multichannel signals — ch. 66
628. Non-redundant time-frequency transforms introduce aliasing artifacts when signals are modified in the frequency domain. — ch. 66
629. MDCT excels for coding but lacks robustness to time-frequency domain edits — ch. 66
630. Stereo signals can be processed for headphones to recreate natural binaural sound fields — ch. 66
631. Each STFT sample represents a circularly continuous tone — ch. 66
632. Modifying STFT bins creates edge transients that synthesis windows imperfectly suppress — ch. 66
633. Modifying frame gains in overlap-add processing reveals previously cancelled aliasing artifacts as audible distortion. — ch. 66
634. The ear’s 1-2 ms time resolution makes monaural temporal cues unusable for localization. — ch. 66
635. Decorrelators degrade audio quality on transient-rich signals like applause. — ch. 66
636. Time–frequency audio processing exploits the ear’s own spectral analysis structure — ch. 66
637. Decorrelation enables creating independent channels while preserving perceptual quality. — ch. 66
638. Audio analysis bands can follow perceptual scales while processing stays on full-resolution QMF bands. — ch. 66
639. Head and pinna scattering imprint direction-dependent spectral notches onto the ear signal. — ch. 66
640. Human spatial resolution limits make stereo-to-multichannel upmixing perceptually plausible. — ch. 66
641. Spatial audio is synthesized by controlling band energies and inter-channel interdependencies. — ch. 66
642. Downsampled filter banks enable efficient DSP implementations like FFT for DFT. — ch. 66
643. Fixed-width filter bands mismatch hearing’s finer low-frequency resolution — ch. 66
644. Audio coders prefer critically sampled filter banks for minimal data transmission. — ch. 66
645. Modern synthetic voices learn from single-speaker speech databases rather than hand-crafted phonetic rules. — ch. 67
646. Synthetic speech dates to late-18th-century acoustic-mechanical machines — ch. 67
647. Speech coding is the most mature speech technology because it needs no databases. — ch. 67
648. Auditory elevation calibration requires visual confirmation of sound sources — ch. 67
649. Data-based speech synthesis splits into sample-reassembly and parameter-model approaches. — ch. 67
650. Logarithmic quantization suits speech but distorts music beyond tolerance — ch. 67
651. Rule-based speech synthesis survives only where computational power is scarce. — ch. 67
652. The hardest part of speech-to-speech translation is language processing, not signal processing. — ch. 67
653. Wideband speech coding spans 50 Hz to 7000 Hz for brighter fuller quality — ch. 67
654. Bandwidth extension and error concealment are speech-enhancement techniques for missing or narrow data. — ch. 67
655. Source-filter parametric synthesis enables voice modification that unit-selection cannot. — ch. 67
656. Mobile phone codecs adapt their bit rate across eight levels to optimize call quality — ch. 67
657. The 1930s vocoder pioneered reducing speech to slowly-varying parameters to cut bandwidth — ch. 67
658. Language knowledge beyond acoustics is required for automatic speech recognition. — ch. 67
659. The telephone band of 300-3400 Hz remains the standard compromise for speech intelligibility. — ch. 67
660. Head movements resolve front-back ambiguity by creating direction-specific binaural cue changes — ch. 67
661. Spatial hearing is worst near 1.7 kHz where head wavelength matches. — ch. 67
662. Speech synthesis selects units by minimizing target and concatenation costs. — ch. 67
663. Humans localize single sources well but fail at source distributions — ch. 67
664. Speech coding is fundamentally a quality-versus-bitrate optimization tradeoff. — ch. 67
665. Human speech evolved as a robust channel tolerant to high background noise — ch. 67
666. High-quality synthesized speech requires tens of hours of manually-segmented recorded speech. — ch. 67
667. Speech recognition shifted from time warping and shallow nets to HMMs then deep networks — ch. 67
668. Synthesis and recognition demand language knowledge databases that coding does not. — ch. 67
669. Reverberation low-pass filters speech modulation while noise raises its floor — ch. 68
670. PAQM estimates audible quality degradation by comparing specific loudness across critical bands. — ch. 68
671. Squaring the pressure signal makes noise and reverberation measurable on a common scale — ch. 68
672. Spatial audio synthesis quality is measurable against a real-source reference. — ch. 68
673. Auditory-model quality metrics fail on acoustically captured signals — ch. 68
674. Binaural auditory models cannot yet explain human perception of complex spatial audio — ch. 68
675. Perceptual auditory models will eventually replace traditional audio metrics — ch. 68
676. Musical instrument design evolved as early product sound quality experimentation — ch. 68
677. Hearing damage begins after long exposure above 85 dB — ch. 68
678. Noise engineering now targets annoyance and quality, not just hearing loss — ch. 68
679. The transmission channel changes from acoustic path to electrical system in device-mediated speech. — ch. 68
680. Quality and impairment are measured with distinct MOS variants. — ch. 68
681. Sound quality means judged preferability, not category membership — ch. 68
682. The public’s subjective opinion is the final truth on sound quality — ch. 68
683. Unlike the eye, the ear has no primary directional sensitivity — ch. 68
684. Timbral quality dominates perceived sound quality over spatial reproduction — ch. 68
685. Digital audio coding made perceptual modeling central to audio engineering — ch. 68
686. Commercial telephone bandwidth standards derive from interwar human listening tests — ch. 68
687. MUSHRA only yields reliable results when sample differences are easily perceptible — ch. 68
688. Human auditory memory cannot compare sounds separated by more than a few seconds — ch. 68
689. Audio quality bottlenecks shifted from storage to transducers — ch. 68
690. Computational sound quality models are trained on subjective listening data — ch. 68
691. Binaural intelligibility level difference measures speech unmasking benefit — ch. 68
692. Comparative MOS ratings allow a test sound to score above the reference. — ch. 68
693. Human auditory adaptation neutralizes room acoustics but not device imperfections — ch. 68
694. PEAQ uses a trained neural-network cognitive model to predict human MOS ratings. — ch. 68
695. Objective sound quality models have limited domains of validity and cannot replace subjective testing — ch. 68
696. Sound quality cannot be assessed except through psychophysical experiments with human subjects — ch. 68
697. Multi-stimulus quality tests fail when samples vary in several perceptual dimensions at once — ch. 68
698. Speech transmission quality hinges on undistorted information transfer — ch. 68
699. Speech intelligibility measures the proportion of correctly identified transmitted messages — ch. 68
700. Beyond intelligibility, telephony must preserve speaker recognizability — ch. 68
701. Speech intelligibility in public spaces is a regulated engineering requirement — ch. 68
702. MTF measurement uses amplitude-modulated noise bands across 0.63 to 12.5 Hz — ch. 69
703. Speech intelligibility metrics can be predicted from impulse response and noise — ch. 69
704. Speech intelligibility estimates compress 98 MTF values into one index — ch. 69
705. Reverberation time maps directly onto equivalent signal-to-noise ratio — ch. 70
706. All sound quality judgments derive from properties of the human auditory system. — ch. 71
707. Telecom device sound quality depends on network, room, and feedback conditions jointly — ch. 71
708. Concert hall acoustics shape the music composed for them — ch. 71
709. PESQ works as a signal-difference device rather than a true auditory model — ch. 71
710. Annoyance and disturbance refer to distinct degrees of noise harm — ch. 71
711. STI reliability is judged against its 0.15 category step size — ch. 71
712. Every sounding product is continuously evaluated for sound quality — ch. 71
713. Consumer devices are engineered to sound masculine and powerful, not merely quiet. — ch. 71
714. Concert hall quality is codified into standardized subjective attribute vocabularies — ch. 71
715. Industry relies on objective methods because exhaustive listening tests are impractical — ch. 71
716. EQUEST injects temporal masking knowledge into echo annoyance estimation — ch. 71
717. Echo delay thresholds produce distinct perceptual and speaking effects — ch. 71
718. Noise suppression algorithms can degrade speech quality in non-stationary noise — ch. 71
719. The first objective speech quality measure failed on mobile-network timing variations — ch. 71
720. Standardized concert hall acoustics metrics have known technical limitations compared to human perception — ch. 71
721. Entering a car generates designed auditory-tactile impressions of finishing quality — ch. 71
722. Sound perception is shaped by the listener’s expectations and context of use — ch. 71
723. Computational simulation is the best method to model the auditory system’s function — ch. 71
724. Speech intelligibility harms open-plan office quality, unlike in auditoria — ch. 71
725. Speech redundancy masks up to 30 percent consonant loss in auditoriums — ch. 71
726. MFCCs remain the dominant feature representation for speech recognition systems — ch. 71
727. Any sound can function as noise depending on the listener — ch. 71
728. Objective models of subjective annoyance remain limited to specific noise types — ch. 71
729. Auditory spectrum computation begins with FFT then emulates hearing sensitivity via filtering and weighting. — ch. 71
730. VoIP call delays exceed those of regular mobile calls — ch. 71
731. Speech quality can be predicted algorithmically using signals that mimic human auditory processing. — ch. 71
732. Sharpness, roughness, and narrowband temporal components increase noise annoyance — ch. 71
733. Excitation spread across frequency bands is modelled as convolution with a level-dependent spreading function. — ch. 71
734. Reverberation time is the key design variable separating concert hall types — ch. 71
735. Speech intelligibility alone cannot capture loss of speaker identity in codecs — ch. 71
736. Spatially broad sound sources should alter their perceived extent with listener distance — ch. 72
737. MIR metadata is largely assigned by expert listeners, motivating automation as music creation accelerates — ch. 72
738. Computational auditory scene analysis targets human-level performance without imitating the ear — ch. 72
739. Music databases need automatic high-level content querying beyond raw audio signals. — ch. 72
740. Sound design in interfaces is an extension of product sound quality — ch. 72
741. Interactive game audio requires dynamic rendering systems beyond traditional linear production. — ch. 72
742. Auditory displays deliver computer information to humans through sonification and earcons — ch. 72
743. Virtual sound source distance is limited by reproduction technology, not perception — ch. 72
744. Music information retrieval serves both casual listeners and music professionals needing structured database queries — ch. 72
745. Sonic interaction design studies sound perception in active embodied contexts rather than passive listening — ch. 72
746. Image-source room acoustics modelling becomes computationally costly in interactive VR — ch. 72
747. CASA aims to extract all perceivable information from ear canal signals. — ch. 72
748. Virtual audio engines render sounds based on avatar position, not just a fixed mix. — ch. 72
749. Blind source separation becomes very difficult when sources exceed microphones — ch. 72
750. Interactive rendering differs from reproduction because user actions alter the sound. — ch. 72
751. Source directivity is usually omitted from virtual audio for computational reasons — ch. 72
752. Speech intelligibility is the standard criterion for hearing function — ch. 73
753. Sensation level expresses tone level relative to one individual’s own threshold — ch. 73
754. Untreated hearing loss in children was once misdiagnosed as intellectual disability — ch. 73
755. Substantial hearing loss reduces ability to react to sonic warnings — ch. 73
756. PLP coefficients describe speech spectral features compactly and speaker-independently — ch. 73
757. Microprocessor electronics made cochlear implants possible — ch. 73
758. Hearing impairment degrades sensitivity, discrimination, and can cause hyperacusis or tinnitus — ch. 73
759. The 0 dB HL audiometric zero varies across frequency by design — ch. 73
760. Excess noise damages cochlear hair cells through several distinct mechanisms — ch. 73
761. Deaf, hard-of-hearing, and deafened denote distinct functional categories — ch. 73
762. Outer hair cell damage compresses dynamic range, causing recruitment and distorted loudness perception — ch. 73
763. Hearing’s temporal resolution varies from 1-2 ms depending on frequency — ch. 74
764. Recruitment makes loudness grow faster in impaired ears than normal ears. — ch. 74
765. Individual HRTF measurements give the most accurate external ear emulation — ch. 74
766. Cochlear hydrodynamics can be simulated by analog electrical equivalent circuits — ch. 75
767. Noise damage creates a 4-kHz notch while aging affects high frequencies — ch. 75
768. Higher-level hearing models rely on assumptions tested only against psychoacoustic data — ch. 75
769. Outer hair cell damage destroys the cocktail party effect — ch. 75
770. Hearing loss varies independently across different frequency ranges — ch. 75
771. Adaptation, temporal integration, and post-masking can be simulated in a single auditory element. — ch. 75
772. Basilar membrane models range from transmission-line equivalent circuits to finite element methods — ch. 75
773. Brief unprotected noise exposure during otherwise protected periods raises hearing loss risk — ch. 76
774. Even moderate 70 dB sound causes temporary hearing threshold shift with hours of exposure — ch. 76
775. Hearing aids are miniature signal-processing sound systems that must secure adequate signal-to-noise ratio — ch. 76
776. Outer hair cell damage widens critical bands and worsens speech-in-noise — ch. 76
777. Tinnitus can arise from unstable outer hair cell oscillation in the cochlea — ch. 76
778. Estimated ITDs match theoretical values especially at low frequencies — ch. 76
779. Recovery from temporary hearing shift takes roughly double the exposure time — ch. 76
780. Eardrum and middle ear status can be measured objectively via acoustic impedance — ch. 76
781. Adult cochlear implant success depends on prior experience with sound — ch. 76
782. Binaural hearing aid processing preserves spatial cues that monaural processing distorts — ch. 76
783. Outer hair cells provide active frequency-selective amplification in the cochlea — ch. 76
784. Tinnitus is treated by masking or habituation, not cured — ch. 76
785. Otoacoustic emissions objectively reveal auditory system functioning and defect location — ch. 76
786. Automatic gain control compresses sound to fit impaired hearing’s reduced dynamic range. — ch. 76
787. Pressure differences across the eardrum attenuate low-frequency sound and cause pain — ch. 76
788. Speech audiometry tests intelligibility by varying signal level and speech pace — ch. 76
789. Noise control should prioritize source reduction then path attenuation before personal protection — ch. 76
790. Combining compression with a limiter reduces distortion better than a limiter alone. — ch. 76
791. Hearing-aid users and small children require loudspeaker-based audiometry for accurate coupling — ch. 76
792. Bone conduction through the skull shapes how one hears one’s own voice — ch. 76
793. A healthy cochlea emits new frequencies when stimulated by two tones — ch. 76
794. Adaptive filters cancel hearing aid whistling caused by earphone-microphone feedback — ch. 76
795. Post-noise tinnitus and distortion signal auditory overload and hearing-loss risk — ch. 76
796. Low-frequency attenuation depends critically on proper seal and construction — ch. 76
797. Most modern cochlear implant patients achieve telephone conversation ability — ch. 76
798. Single loud impulses can instantly tear the eardrum or break ossicles — ch. 76
799. Speech processing can fail independently of general hearing ability — ch. 76
800. Subtractive microphone designs produce cardioid and dipole patterns via port spacing and delay — ch. 76
801. Cochlear implants bypass the conductive auditory path to stimulate nerve fibers directly — ch. 76
802. Pure-tone audiometry measures hearing thresholds at six standard frequencies — ch. 76
803. Beamforming combines multiple microphone signals into directive beams — ch. 76
804. Single-channel noise reduction exploits speech’s temporal, spectral, and statistical properties — ch. 76
805. Hearing-loss risk threshold should drop from 85 to 75 dB for full protection — ch. 76
806. Hearing-loss risk dominates noise research because it is predictable from sound properties — ch. 76
807. Tinnitus can originate from blood pressure changes in the ear — ch. 76
808. Noise is the most common cause of work-related impairments — ch. 76
809. Comparing bone- and air-conduction thresholds distinguishes conductive from sensorineural hearing loss — ch. 76
810. Hearing impairment degrades signal-to-noise tolerance more than quiet speech perception. — ch. 76

Annotation Roundup

Question

• The human auditory system is the hardest link in the communication chain. — Worth chasing: why did formal/instrumental knowledge of auditory functions only recently emerge? What technical barriers blocked measurement of the peripheral auditory system?
• Engineering perspectives increasingly serve computational modelling of auditory perception. — Worth chasing: which specific computational models of auditory perception and communication situations are the field converging on, and what are their practical applications?
• Helmholtz resonance shapes violin bodies, speaker designs, and car cabin noise — The car-cabin window infrasound amplification is a striking everyday safety phenomenon worth chasing — what are its measured dB levels and effects on drivers?
• Above a critical frequency, room acoustics become statistically tractable as diffuse fields. — Critical frequency depends on room volume and reverberation time — worth chasing the exact K-range formula and its engineering limits for room acoustics simulation.
• Human hearing resists linear time-invariant modeling except at the periphery — Which peripheral parts of the auditory system are LTI-modelable, and where does the non-linearity begin critically?
• The auditory system largely ignores signal phase, driving spectral analysis conventions — The caveat that the ear ‘deviates from simple Fourier analysis’ in detailed behavior — worth chasing for where phase does matter perceptually.
• Wavelet and constant-Q transforms trade time and frequency resolution non-uniformly — Why does the perceptual frequency scale justify a constant-Q (constant relative bandwidth) analysis rather than a linear one? The text asserts the motivation but does not spell out the psychoacoustic backing here.
• Any band-limited linear time-invariant system can be implemented as a digital filter — This claims total representability of LTI systems as digital filters under band-limiting — worth chasing the exact conditions and limits on sampling rate.
• Air absorption and room acoustics can be modeled as low-pass and arbitrary filters — Which specific rooms map to which filter responses — a question to chase in later modeling chapters.
• A microphone’s sensing principle determines whether its pickup pattern is omnidirectional, dipolar, or cardioid — Worth chasing: exactly how do microphone arrays make directional patterns electronically controllable, and what mathematics underlies beamforming?
• Equalizing a loudspeaker response can worsen off-axis sound quality. — Interesting that flattening magnitude response along one axis makes off-axis response worse — worth understanding why and whether it’s ever worth doing.
• Objective measures based on hearing models substitute for tedious listening tests of distortion. — Which objective auditory-perception models are actually used in industry for distortion, and how well do they correlate with listening-test judgments?
• Nonlinear equalization methods remain underdeveloped in audio engineering — Why have nonlinear EQ methods not advanced far? Worth chasing the Klippel 2006 reference to understand the technical barrier.
• Speakers use pitch variation to signal statements versus questions — The excerpt trails off into tone languages at the end but the rubric requires a complete stopping point; the core intonation point stands on its own.
• Glottal closing generates most vocal-tract excitatory energy. — The glottis is formally non-LTI, yet ‘relatively weak’ interaction justifies linearity — how weak is weak enough, quantitatively?
• Formant tracking to harmonics boosts loudness, not intelligibility, in soprano singing — Why is loudness worth more than intelligibility in opera specifically — is it stage convention, orchestral masking, or tradition? Worth chasing.
• Musical instruments couple excitation to resonances; human speech does not — Worth chasing: does this uncoupled design constrain what timbres speech can produce versus a fully coupled system like a trumpet?
• A single amplitude envelope misses that harmonics decay and modulate at different rates — Why do different harmonics of the kantele modulate at different frequencies — does this trace to specific sympathetic resonances or string nonlinearity?
• Speech tracts lack coupling that brass instruments synchronize with. — Why did speech evolution avoid the coupling that instruments exploit? Worth chasing — an evolutionary/cognitive angle on why vocal folds don’t lock to tract resonances.
• Cochlear pitch mapping exploits the mass of vibrating liquid varying with route length — This explanation frames the traveling-wave route-selection physics — worth checking how this mass-to-resonance argument reconciles with the alternative phase-response account in para 753.
• The cochlea’s sharp frequency selectivity needs an unexplained second filter. — The second filter is ‘hypothetical’ — worth chasing what it was eventually identified as (active cochlear mechanics/outer hair cells).
• Single auditory neurons encode only a narrow 20-40 dB dynamic range — How does the ear achieve its full ~120 dB dynamic range if single fibers only span 20-40 dB? Points to a rate-place/recruitment mechanism worth chasing.
• Loudness above 60 dB is encoded by cross-frequency spread, not single-neuron rate — Curious how broadband loudness summation is modeled computationally — the multi-band excitation model starting point.
• fMRI identifies active brain regions well but is noisy and time-slow for auditory study. — The noise of fMRI scanners as a confound for auditory studies is worth chasing — what workarounds exist, e.g. sparse sampling or acoustic dampening?
• Single neuron auditory responses are mapped through controlled stimulus repetition — How many repetitions are typically needed, and what stimulus controls matter most for reliable neural maps?
• Psychoacoustics developed largely through communication technology challenges — Which specific communication-technological challenges drove psychoacoustic method development over the past century?
• Uncontrolled loudness differences confound audio quality listening-test results — What calibration methods and loudness metrics are actually used to equalize levels across test items?
• Simple stimuli fail to represent how hearing processes real complex sounds — Worth chasing: which complex sounds (speech, music, environmental) best expose hearing functions that simple stimuli mask?
• Free sorting adds post-hoc verbal labels that enable interpreting perceptual dimensions. — Grounded theory as a qualitative-methods transplant into audio perception research is a methodological cross-pollination worth chasing.
• Generalized procrustes analysis aligns different subjects’ attribute ratings into common dimensions. — How exactly does generalized procrustes analysis rotate/scale individual subject spaces to align them? Worth chasing the Gower and Hand reference.
• Coherent amplitude modulation across frequency bands may drive auditory source grouping. — If modulation filter banks are unproven and controversial, what competing explanations exist for modulation perception and sound grouping?
• Narrowband noise can carry clearer pitch than some periodic signals — Why would an aperiodic noise outrank periodic tones in pitch strength? Worth chasing the Fastl-Stoll methodology to understand the auditory mechanism.
• Frequency JND is best around 500 Hz at about 0.3 percent — Why does relative JND rise again below 500 Hz if sensitivity is still strong there — what exact mechanism of decreasing sensitivity at low frequencies produces this?
• A set of harmonically related partials shifted together still changes perceived pitch. — Why does shifting harmonic partials jointly still shift pitch if modulation frequency is preserved? Worth chasing how modern models reconcile this with timing theory.
• Place theory cannot fully explain how pitch is extracted from components — What is the ‘other process’ that fixes perceived pitch as stimulation level broadens basilar-membrane activation? Worth chasing in the following pages.
• Shifting harmonic partials upward raises perceived pitch despite unchanged envelope modulation — What mechanism explains the pitch rise from inharmonic transposition when temporal envelope modulation is unchanged? Place theory seems implicated but the chapter admits models fall short.
• Loudness perception sums auditory bands and signals distance by comparison — The claim that loudness dependence on bandwidth carries no evolutionary deficiency — worth interrogating whether broadband-specific loudness could aid predator/prey distance assessment in ways not explored here.
• Amplitude-modulation detection improves with tone level but is level-independent for noise. — Why is the greatest AM sensitivity at exactly 4 Hz modulation frequency? Worth chasing the physiological basis.
• Human hearing can in principle distinguish astronomically many timbres. — Does the masking-effect reduction bring this down to a number that’s meaningful for musical composition or synthesis design? Worth chasing.
• Temporal masking makes silence following a sound seem shorter — This asymmetry — perceived duration of burst vs. its following silence — seems relevant to speech rhythm and prosody perception. Worth chasing how masking theory extends to conversational turn-taking.
• Brief silences are perceptually longer than their physical duration. — How much of the difficulty perceiving short speech pauses is masking vs. the perceptual length-inflation shown separately for bursts? Wants pinning down.
• Impulsiveness is estimated from level statistics rather than directly measured — The 10 dB/s onset-slope criterion suggests a concrete threshold worth probing for applied transient detectors.
• No single standard method measures perceived tonality accurately — The failure of T/N with multiple tones in one critical band suggests a gap worth investigating — what perceptual model would handle multi-tone critical bands correctly?
• One second of exposure triggers neural spectral adaptation in hearing — The enhancement effect (contrast enhancement after adaptation) seems behaviorally similar to visual afterimages - worth checking whether a shared neural gain-normalization mechanism operates across modalities.
• Headphone group delay alters audibility of phase-spectrum effects — Which headphone designs minimize group delay spread, and does this explain why high-end headphones advertise phase performance?
• Reversing the phase of a single harmonic can drastically change its perceived loudness — Which sounds count as ‘buzzy’ with this sensitivity, and do real signals like speech ever have such phase structure?
• Room reverberation can render phase-sensitive signals perceptually phase-insensitive — Implication worth chasing: if reverberation randomizes phase and erases group-delay percepts, then phase information is largely irrelevant for speech/audio in real rooms — yet audio coding expends bits on phase? How is that reconciled?
• No tuning system can make all diatonic intervals have simple ratios — Worth chasing: how does equal temperament’s slightly mistuned major third actually affect perceived consonance versus just intonation in practice?
• Rhythm derives from natural body repetitions like heartbeat and walking. — Why is systematic description of rhythm harder than melody? Worth chasing the structural reason.
• Auditory cortex adaptation decreases response to frequent sounds, shaping perception — The idea that adaptation is ‘a process of forgetting’ is nearly the inverse of standard models (adaptation as fatigue). Worth chasing the McKeown-Wellsted argument.
• Reversing a signal’s polarity changes its perceived bass level — How does mere time-inversion of an identical spectrum change perceived bass? A purely phase-based phenomenon with perceptual consequences worth exploring for sound design.
• Sound localization evolved in nocturnal mammals and remains vital for prey and predator. — Why did mammals rely on hearing over vision for spatial orientation as nocturnal animals? Worth chasing the sensory-tradeoff literature.
• Shoulder reflection delays versus direct sound vary inversely with source elevation — Could exploit shoulder-reflection delay as a robust elevation cue for spatial audio rendering, or as a bio-inspired trick for head-related transfer function synthesis?
• ITD spans 0-700 microseconds across azimuth independent of frequency — The spherical-head model accuracy claim and its frequency independence are useful quantitative facts for any audio engineering application, potentially worth cross-checking exact 600 vs 700 microsecond figures.
• Low-frequency interaural level differences cue source distance and signal coherence. — The claim that ILD serves as a near-field distance cue at low frequencies is worth verifying against head-related transfer function literature for measured near-field ILD values.
• Humans perceive interaural level differences at frequencies where plane waves produce none — The chapter then explains two functions for low-frequency ILD sensing (distance and coherence) - worth understanding which of these dominates in real listening conditions.
• Low interaural coherence makes sound events seem diffuse and surrounding — Whether the auditory system computes separate IC cues or derives the effect from fluctuating ITD/ILD remains unresolved — worth chasing.
• Spectral-cue adaptation in hearing requires visible sound sources as a reference — If adaptation requires visible sources, what happens to blind listeners’ elevation accuracy, and can touch substitute?
• Sound width perception depends on signal type, not physical source size — Why do equal-frequency sinusoids collapse perceived width? Is this an auditory grouping/capture phenomenon worth tracing further?
• Humans adapt to a room’s spatial cues through learning over hundreds of trials — The ‘template’ adaptation mechanism is unnamed and unexplained — worth chasing what perceptual-learning studies reveal about this process.
• Sound localization blur is smallest in front and largest at the sides — Why would side localization be worst despite strong binaural cues there — does the front/back ambiguity dominate? Worth clarifying.
• Perceived auditory distance follows a compressive power function of actual distance — The paragraph cuts off mid-sentence at the power function definition — worth chasing the actual exponent value Zahorik reported across conditions.
• The cochlear amplifier requires nonlinear, level-dependent elements to model. — What exactly does negative damping represent physically, and how does it map to the cochlear amplifier’s active gain?
• Cross-correlation sidelobes near 1 kHz stem from interaural phase ambiguities — The 2π phase-advance sidelobe phenomenon may explain why some listeners localize better than others near 1 kHz — worth checking whether there are individual differences in detecting these ambiguous lags.
• The Blumlein XY pair captures reverberance fully but causes phase artefacts from side sound. — Would be worth chasing: how severe are the colouration/directional artefacts in practice, and does it limit Blumlein pair use?
• Physical accuracy of a sound field does not predict perceived accuracy — What determines which wave-field deviations matter perceptually? This seems underexplored, a gap between physical acoustics and psychoacoustics.
• VBAP predicts perceived source azimuth accurately but not personal elevation — Why is perceived elevation personal while azimuth is uniform across subjects? Worth chasing whether this reflects individual HRTF differences.
• Audio quality improvements mask video defects, but video quality does not mask audio defects. — Why is the audio-to-video masking asymmetric? Is there a modality hierarchy in perception worth pursuing?
• Tactile-audio desynchrony is as critical as audiovisual desynchrony — Why do musicians have lower haptic-audio thresholds — is it training or a general precision of temporal integration? Worth chasing.
• Wave field synthesis is nearly always generated synthetically, not from real recordings. — Why has no one built the hundreds-of-microphones rig required for actual WFS recordings? The cost/engineering barrier seems ripe for eventual technological disruption.
• Headphone and loudspeaker playback of stereo differ yet yield similar spatial order. — Why does stereo spatial order survive the ILD/ITD cue transformation between formats? Worth tracing the perceptual mechanism.
• Regularization trades inversion accuracy against loudspeaker dynamic-range demands — Worth chasing: how is the regularization factor β chosen in practice — what on-set preferred values have been found across listening tests?
• Crosstalk-cancelled playback yields externalized sound only in the frontal hemisphere — Why does a front stereo dipole only render the frontal hemisphere of the scene — is this a hard physical constraint of the two-loudspeaker geometry?
• Shoulder reflections imprint identifiable arcs on measured head-related impulse responses. — Why must the source rotate with ear canal at center for the direct-path peak to remain at fixed latency?
• Interaural level differences only emerge at frequencies above roughly 400 Hz. — Passage cuts off mid-sentence, obscuring the stated consequence about low-frequency localization.
• Modifying frame gains in overlap-add processing reveals previously cancelled aliasing artifacts as audible distortion. — What error-masking or gain-adaptation strategies in real codecs (like MP3 or AAC) account for these hidden aliasing components during loudness or EQ adjustments?
• Fixed-width filter bands mismatch hearing’s finer low-frequency resolution — Cascaded filter banks at low bands trade off latency and complexity — worth understanding the cost/benefit tradeoff in spatial audio codecs.
• Language knowledge beyond acoustics is required for automatic speech recognition. — What specific statistical language-modeling techniques handle acoustically similar words in large vocabularies?
• Spatial hearing is worst near 1.7 kHz where head wavelength matches. — The 1.7 kHz blind spot in horizontal localization — could have design implications for spatial audio or wearable alarm placement worth chasing.
• Humans localize single sources well but fail at source distributions — Why does the auditory system sum overlapping signals in the ear canals rather than separate them? This seems like a fundamental architectural constraint worth chasing.
• Hearing damage begins after long exposure above 85 dB — The 85 dB A-weighted daily equivalent threshold — worth checking whether this aligns with OSHA/WHO standards on occupational noise limits.
• PEAQ uses a trained neural-network cognitive model to predict human MOS ratings. — How does the PEAQ training data (real devices, listener panels) get collected, and what ceiling does it hit vs. blind listening?
• Timbral quality dominates perceived sound quality over spatial reproduction — The finding that spatial advantage vanishes under timbral degradation is surprising — worth checking whether it holds across content types and other codec artefacts.
• Binaural auditory models cannot yet explain human perception of complex spatial audio — Why did the PEAQ extension to multichannel audio fail? What specific limitations of binaural models surfaced in the 2010 Liebetrau work?
• The first objective speech quality measure failed on mobile-network timing variations — Why did subtracting time-varying feature vectors cause the model to misconstrue syllable-length jitter? Worth chasing the PSQM algorithm details.
• Echo delay thresholds produce distinct perceptual and speaking effects — The neural-speech-production confusion mechanism for large-delay echoes is presented tentatively — worth chasing the Appel and Beerends source for the experimental evidence.
• Standardized concert hall acoustics metrics have known technical limitations compared to human perception — Worth chasing: what alternative methods are being proposed to overcome the omnidirectional-source and frequency-range limits of ISO 3382-1?
• Speech intelligibility harms open-plan office quality, unlike in auditoria — The STI threshold of 0.2 is a concrete design rule — worth exploring what acoustic treatments achieve it and how it interacts with audience design.
• Objective models of subjective annoyance remain limited to specific noise types — The stated goal of objective measurement of subjective annoyance — the failure mode worth understanding: why don’t these models generalize?
• Source directivity is usually omitted from virtual audio for computational reasons — Worth chasing: when does omitting source directivity actually break plausibility — and would simpler directional models like dipole sun be a cheap middle ground?
• Music databases need automatic high-level content querying beyond raw audio signals. — Two decades later, how close are music streaming services to answering such natural-language queries? Worth investigating current state.
• MIR metadata is largely assigned by expert listeners, motivating automation as music creation accelerates — Who employs these expert listeners, and what do they actually do across a typical work week?
• PLP coefficients describe speech spectral features compactly and speaker-independently — Why does transforming loudness spectra back into autocorrelation functions improve speaker-independence? Worth chasing the underlying mechanism.
• Tinnitus can originate from blood pressure changes in the ear — Pulsatile tinnitus tracking heartbeats as a vascular, not cochlear, mechanism — worth chasing for a fuller account of tinnitus causes.
• Adult cochlear implant success depends on prior experience with sound — The paragraph is cut off mid-sentence at the contrast (children) — worth chasing the pediatric outcome data. Xref
• Noise is defined by annoyance, disturbance, or hearing harm rather than loudness alone — Connects to One Square Inch of Silence — noise as subjective/‘unwanted sound’ rather than acoustic measurement.
• Voice-based computer interfaces required granting machines language processing and world knowledge — Connects to the broader history of conversational interfaces and the Turing test lineage.
• Echolocating bats build image-like environmental representations from reflected chirps — Connects to underwater sonar design, which engineers built by mimicking bat echolocation — an acoustic communication applied to navigation.
• Sound storage digitized fully but human ears force analog endpoints — Connects to the reader’s interest in acoustic applications and the physical transduction problem — why speakers and mics remain the hard interface.
• Computational modelling is merging experimentation and theory into one method — Connects to the ‘mature scientific instruments’ idea in Kuhn and to how Deep Listening / sound studies method merges modes of knowledge production.
• Systems are analyzed through complementary structure and function concepts — These systems concepts map directly onto control theory and feedback loops the reader knows from Milgrom’s game theory work and from signal processing.
• Natural and engineered systems organize as vertical control hierarchies — Connects to the layered signal-processing architecture common in communication systems and to game-theoretic hierarchy in Milgrom.
• Evolution produced language and engineering, but machines still lag in pattern recognition — The logic-vs-pattern-recognition split echoes the diamond-industry skill-over-machine theme in Opting out of the Legal System.
• The human auditory system is the hardest link in the communication chain. — Parallels James Gibson’s and Shannon-Weaver communication models where the receiver end resists formalization; connects to the reader’s noise/silence reading on why perceptual receivers are neglected.
• Pythagoras established the inverse relation between pitch and string length — Pythagorean harmony-as-universe-law connects to the reader’s knowledge of music theory and ancient Greek science history.
• Heterarchical, distributed control can outperform strict hierarchy under changing conditions. — Connects to the reader’s institutions work on decentralized reputation systems and the fact that control can be distributed yet effective.
• Internal models of reality must trade completeness for tractability. — Echoes the reader’s notes on representation metaphors failing on large debatable claims, and the principle that useful models need not be complete.
• The human auditory system is communication acoustics’ hardest link to model — Connects to prior reading on soundscape and silence work, which likewise privileges the listener/perceiver over physical signal.
• Sound as physical event and sound as auditory perception are distinct — Connects to the soundscape studies theme — Schafer’s work also separates acoustic events from their perceptual significance.
• Sound’s physical force is negligible yet its informational effects matter greatly. — Connects to the reader’s broader acoustics interest: the informational/informational-signal framing reappears across bioacoustics and communication technology.
• Helmholtz resonance shapes violin bodies, speaker designs, and car cabin noise — Violin body air-hole resonance ties directly to instrument-construction material physics from the trees-infrastructure reading (spruce soundboards, cavity tuning).
• Any vibration decomposes into weighted sums of a system’s modal oscillations — The linear-combination basis idea parallels Fourier decomposition and eigen-basis expansions familiar from linear algebra.
• Any discontinuity in a medium reflects an incident wavefront — Underpins echo and reverberation in rooms, and impedance matching in the ear — discontinuities at the eardrum and cochlear boundaries matter for hearing.
• Coupled oscillators produce complex resonances shared by instruments, voices, and rooms — Connects to why the author’s later focus on room acoustics and vocal tract acoustics builds on this coupling foundation.
• Any wave can be decomposed into two counter-travelling component waves. — The traveling-wave superposition principle resonates with Fourier decomposition used throughout signal processing — a foundational result reappearing across the book.
• Coherent sound addition yields double the amplitude gain of incoherent addition — The 6 dB vs 3 dB coherent/incoherent addition rule is the foundation behind speaker array design and comb-filtering artifacts in audio engineering.
• Rigid-wall tube terminations fully reflect waves while open ends reflect with phase inversion. — Underpins how organ pipes, flutes, and brass instruments form standing waves; connects to wind instrument acoustics.
• Bending stiffness makes bar vibrations dispersive and inharmonic. — Explains the characteristic inharmonic timbre of marimbas, xylophones, and gamelan metallophones; connects to the physics underlying these sounds.
• Tube end conditions determine which harmonic partials an instrument produces — Explains why flute and clarinet timbres differ — connects to any instrument acoustics knowledge and to wind instrument sound synthesis.
• Room reflections shape how well speech communicates across a space. — Connects to the reader’s running interest in how room acoustics aid or hinder intelligibility, extending soundscape and silence themes.
• Noise barriers only effectively block high-frequency traffic sound, not low frequencies — Explains why highway barriers near residences fail against rumbling truck noise — connects to environmental-noise and urban acoustic design applications.
• The Helmholtz resonator models confined air as a spring with an air-mass plug — The bottle-blowing demo connects to everyday acoustic phenomena and to the blowing-across-bottle sound familiar in sound design.
• Resonance can be engineered to strengthen sound or damped to prevent malfunction — This application of resonance to instrument soundboards connects to the ‘trees as infrastructure’ essay’s material on musical instrument wood and violin sound.
• Tube termination conditions determine which harmonics a wind instrument produces — This explains why clarinet and flute timbres differ audibly — ties to perception of instrument families.
• Stiffness makes transversal waves dispersive and their modal tones inharmonic — Explains why percussion and struck-bar instruments like marimbas and church bells sound inharmonic — connects to the science of instrument design and why metal bars are tuned by carving.
• Early reflections boost perceived loudness while loud late reverberation degrades intelligibility — Connects to room acoustics literature and audio engineering practice on reverberation thresholds for speech vs music intelligibility.
• Room sound decay is exponential, appearing linear on a dB scale — Connects to the general physics of resonator decay in other domains — spring damping, capacitor discharge share the same exponential form.
• Room reflections can be modelled as sources in mirrored rooms — Connects to the reader’s work on room acoustics simulation and methods used in audio/VR rendering.
• Element-based acoustics methods grow computationally costly at higher frequencies — Relates to the reader’s interest in applications: explains why real-time acoustic simulation is frequency-limited.
• Geometric acoustics fails on diffraction at low frequencies in complex rooms — Connects to why low-frequency room modes need different treatment than ray-based high-frequency modeling, relevant to loudspeaker and architectural acoustics design.
• Physically realizable systems cannot respond to future input values — Causality constraint echoes related physical constraints seen in silence/acoustic-communication reading.
• Human hearing resists linear time-invariant modeling except at the periphery — Ties directly to the reader’s stated goal of communication acoustics applications in hearing and biology.
• The Fourier transform trades costly time-domain convolution for simple frequency-domain multiplication — Connects to how convolution reverbs and filtering are implemented in audio software; also the reason FFT is everywhere in acoustic analysis.
• Loudspeakers at high power and low frequencies become highly nonlinear. — Connects to the reader’s audio/soundscape work and home audio reproduction interests.
• Frequency-domain analysis mirrors how the human ear processes sound — Connects to the reader’s existing soundscape/audio work where frequency-based signal decomposition maps onto human perception.
• The ear’s analysis resembles Fourier frequency decomposition of sound — Connects to the reader’s broader interest in acoustic communication — the ear-as-Fourier-analyzer framing recurs across hearing science and audio engineering contexts.
• Wavelet and constant-Q transforms trade time and frequency resolution non-uniformly — Constant-Q transform links to how the auditory filter bank behaves — matching tonal perception — which connects to the cochlear/bark-scale content likely later in this same book.
• Autocorrelation reveals a signal’s fundamental periodicity through periodic peaks. — Autocorrelation is the core of pitch detection in hearing research and music information retrieval (e.g., YIN algorithm).
• Each added quantization bit yields 6dB SNR improvement — Pairs with the book’s earlier discussion of the auditory system’s ~130 dB dynamic range as the benchmark for digital fidelity requirements.
• Air absorption and room acoustics can be modeled as low-pass and arbitrary filters — This reframes room acoustics and atmospheric absorption as filter responses — connects directly to the reader’s interest in acoustic communication applications and to prior material on room/ambient acoustics.
• The similarity between artificial and biological neural networks is only superficial. — Relevant to the reader’s acoustic-communication application focus: ANN-based speech recognition figures heavily in the book’s audio chapters.
• Loudspeakers and room acoustics remain the weak link in audio quality — Connects to the broader soundscape/silence reading: the ‘practically perfect’ electronics vs. imperfect transducer-room coupling echoes the theme that the physical interface with the world is where engineering ideals break down.
• Headphone cushion leaks alter only the low-frequency transfer to the eardrum — Directly relevant to the reader’s aim of applying acoustics — headphone seal quality matters for research applications like binaural reproduction.
• Closed headphones gain isolation but become sensitive to leakage effects. — The isolation-vs-leakage tradeoff is a design principle that maps onto many communication-audio system choices.
• Headphones originated as noise isolation for radio operators — Connects to the broader theme of individualizing acoustic control for research (binaural/HRTF audio) — relevant to the reader’s interest in acoustic communication applications.
• A plus or minus 5 dB loudspeaker deviation goes unnoticed without a direct reference — Connects to the reader’s noise/silence corpus: room acoustics interacting with loudspeakers echoes the soundscape work on how listening environment shapes perceived signal.
• Loudspeaker-room interactions remain the weak link in audio reproduction chains — Connects to the reader’s steady interest in acoustic engineering as applied communications, and the ‘one square inch of silence’ listening-work on how reproduction systems interact with physical rooms.
• Closed headphones gain isolation but become sensitive to leakage effects. — Directly relevant to consumer hi-fi headphone choice and binaural research setups that require controlled per-ear signals.
• Perfect sound reproduction tolerates arbitrary signal delay when reproducing recordings. — The 10 ms live-reinforcement threshold connects to the earlier soundscape discussion of timing and acoustic feedback.
• Noise is defined by its disturbing effect relative to a desired signal — Connects to the reader’s broader reading on noise/silence (The Soundscape, One Square Inch of Silence) — here noise is defined operationally via SNR rather than aesthetically.
• Even ideal equalization degrades off-axis response and can amplify distortion and noise — Connects to the reader’s standing interest in how idealized signal-processing fixes carry real-world side effects, echoing audio engineering tradeoffs in loudspeaker design literature.
• Speech sound generation was mechanically modeled as early as the 18th century — Connects to von Kempelen’s speaking machine, which later inspired other synthetic voice projects in the acoustic communication literature.
• Vocal tract length differs by sex and its resonances form speech formants. — The source-filter framing (glottal source + moving-tract filter) is the exact transfer-function architecture used in speech synthesis and analysis (LPC, vocoders).
• Phonation types differ by glottal airflow and spectrum shape — Connects to the reader’s ongoing interests in the mechanics of voice production and to source-filter theory used elsewhere in acoustics.
• The glottal waveform can be estimated indirectly through inverse modeling of microphone recordings — Connects to the reader’s interest in acoustic communication applications; inverse modeling of glottal waveform is used in speech synthesis and voice pathology detection.
• Speech is a continuous flow that still requires discrete units to be understood — Connects to orthographic transparency debates in reading science and to the continuous-vs-discrete modeling choice in speech synthesis.
• Computerized speech handling required simplified phonetic notations tailored per language — Connects directly to speech synthesis (TTS) systems, an application domain of communication acoustics.
• All vowel distinctions reduce to tongue position, openness, and lip rounding — These three articulatory dimensions map directly onto the vowel space used in formant-frequency plots and speech synthesis parameterization.
• Languages differ widely in vowel category counts, from two to over ten. — Connects to WALS (World Atlas of Language Structures) language typology research.
• A full speech communication chain spans brain, vocal organs, channel, and hearing. — The 1948 Shannon-Weaver linear communication model maps onto this source-channel-receiver speech chain.
• The glottal waveform can be estimated indirectly through inverse modeling of microphone recordings — Could connect to the reader’s interest in acoustic signal processing applications — non-invasive vocal diagnostics.
• Phonetic information concentrates either in a phone’s steady core or its transition boundary — Connects to the reader’s acoustic communication focus — this segmentation trade-off directly bears on how ASR and TTS systems represent speech.
• In tonal languages like Chinese, pitch changes alter the phoneme itself — Connects to how pitch is a primary acoustic carrier — relevant to acoustic communication applications in tonal vs. non-tonal contexts.
• Mathematical source-filter models of speech enable practical synthesis and coding algorithms — Links Fant’s source-filter theory to the speech synthesis/coding pipeline — relevant to anyone working in voice tech or acoustic modeling.
• Speech production maps onto a three-stage cascaded electrical circuit analogy — The cascaded circuit analogy parallels signal-processing treatments in other domains of communication acoustics.
• Glottal closing generates most vocal-tract excitatory energy. — Perceptually motivated: abrupt glottal closure is what gives vowels their formant excitation — links to why voice quality (breathy vs pressed) changes acoustics.
• The glottal waveform’s closing phase is abrupt and carries most excitatory energy. — Connects to source-filter speech synthesis theory — same glottal source model used in parametric synthesizers.
• Acoustic reflections arise at any change in tube cross-sectional area — The impedance-mismatch reflection principle that governs vocal-tract acoustics also drives cochlear impedance matching, transmission-line theory, and architectural acoustics.
• High-pitched soprano singing is unintelligible because harmonics are too sparse for formants — Connects to the intelligibility tradeoffs in speech synthesis and communication design that this reader explores across the book.
• Overtone singing selects harmonics with sharp vocal-tract resonances to create melodies — Echoes the formant structure work earlier in the book and links to the soprano formant-tuning passage at 631 as the same principle pushed to extremes.
• Formant tracking to harmonics boosts loudness, not intelligibility, in soprano singing — Connects with the ADDENDUM article on opera/music themes and generally to formant theory from earlier in the book’s voice chapters.
• Music and speech are utility sounds distinguished by their primary role — Connects to the soundscape literature (Schafer’s the tuning of the world) where all listened-to sound, even noise, can be re-framed as music.
• Musical instruments couple excitation to resonances; human speech does not — Connects to source-filter theory of speech production, where the glottal source and vocal-tract filter are conventionally treated as independent.
• A single amplitude envelope misses that harmonics decay and modulate at different rates — Plucked-string partial behavior parallels what was covered on plucked strings generally; connects to instrument synthesis questions.
• Musical instrument sounds divide into attack, decay, sustain, and release phases — Connects to sound-synthesis and sampling applications in acoustic communication technology.
• Speech tracts lack coupling that brass instruments synchronize with. — This contrasts the source-filter model for speech (glottal source uncoupled from tract) with instruments where excitation and resonator are strongly coupled — a distinction central to acoustic communication theory.
• Resonators color spectral and temporal content of transmitted sound. — The build-up/decay differing between resonance and non-resonance frequencies, perceived as reverberation, ties directly to the body’s role as an acoustic filter shaping both timbre and time-domain behavior of plucked/instrument sounds.
• Most real instruments are nonlinear and resist simple filter-cascade modelling — The linear-vs-nonlinear instrument modelling split connects to physical modeling synthesis literature (Smith, Bilbao) that underlies modern software instruments.
• FM synthesis generates harmonic spectra when the modulating frequency equals the carrier frequency — FM synthesis is the classic sound-generation method behind 80s synthesizers and early video game music; connects to sound design practice.
• Placing all string damping at terminations yields efficient digital synthesis — Connects to the earlier chapter discussion of resonators and Chapter 2’s air column and string physics.
• A guitar body’s effect can be simulated by convolution with its measured impulse response — Relates to convolution and impulse response concepts that would appear elsewhere in the reader’s acoustics and DSP knowledge.
• Human hearing diverged from other mammals mainly via speech analysis — Connects to the broader acoustic communication theme of the book; the mammal-to-human speech specialization frame echoes comparative biology discussions of auditory specializations.
• The human auditory system evolved specifically for spoken-language analysis — Connects to the reader’s interest in communication acoustics as applied to speech technology and spatial hearing.
• The ear canal acts as a quarter-wavelength resonator amplifying 3-4 kHz by 10 dB — Connects to why hearing is most sensitive around 3-4 kHz and why it aligns with speech emphasis and hearing-aid design bands.
• The middle ear evolved as an impedance transformer to bridge air-water sound energy loss — Connects to the reader’s sonar/bioacoustics interests — impedance matching is why aquatic insects and fish face different coupling problems than terrestrial hearing.
• Stronger tones in the cochlea actively mask neighboring weaker tones via two-tone suppression. — Relates to psychoacoustic masking effects relevant to audio compression codecs like MP3 that exploit masking thresholds.
• Two tones presented together generate an audible combination (difference) tone not physically present. — The ‘Tartini tone’ or missing fundamental phenomenon — ties to pitch perception literature that audio designers rely on.
• Loud sounds spread cochlear activation broadly and respond faster than soft sounds — Wide dynamic-range compression scheme in auditory nerve firing pays off in understanding why loud environments mask quiet detail.
• A competing tone at another frequency suppresses a cochlea channel’s response — Explains perceptual masking and how the auditory system enhances spectral peaks over broadband noise.
• Single auditory nerve fibres encode only a narrow 20-40 dB dynamic range. — Connects to puzzle of how the ear spans ~120 dB dynamic range via population coding across fibres — relevant to audio compression and loudness perception.
• Auditory nerve phase locking degrades above 1500 Hz and is weak above 4 kHz — Explains the perceptual basis for why pitch and temporal fine structure matter mainly below ~1.5 kHz, relevant to audio coding (e.g., MP3 high-frequency processing) and psychoacoustics.
• Population coding across many neurons achieves precise and robust auditory processing — Parallels to neural-network population coding ideas in machine learning and brain-computer-interface design.
• Speech remains understandable at high levels because timing carries formant info — Connects to the reader’s ongoing theme that signal vs. noise and redundancy sustain communication — the temporal encoding is an information-theoretic redundancy mechanism alongside the auditory soundscape work.
• The superior colliculus is the only mammalian brain nucleus with a topographical auditory map — Connects to spatial audio and cross-modal perception work; relevant to binaural hearing and sound localization systems.
• fMRI identifies active brain regions well but is noisy and time-slow for auditory study. — Connects to the reader’s interest in applications of acoustic communication — the scanner noise itself is an acoustic stimulus problem researchers must engineer around.
• Auditory models link acoustic stimulus features to perceived attributes computationally. — Connects to the reader’s work on Deep Listening and soundscape — a model that predicts subjective auditory attributes from acoustic input is the computational backbone for soundscape design applications.
• Acoustic engineering usually needs only cochlear models, not higher nuclei — Speaks directly to the reader’s goal of mapping acoustic science to engineering applications, and to the book’s own spatial audio theme.
• Psychophysical functions map continuous physical sound scales onto varied perceptual scales. — The psychometric function’s yes/no probability mapping connects to signal detection theory and threshold experiments.
• Anechoic chambers misrepresent how loudspeakers sound in normal rooms — Connects to audio engineering practice where room acoustics and speaker response are inseparable; relates to the Soundscape book in this library on room/ambient acoustics.
• Perceptual thresholds blur because listener internal noise varies by personality — Connects to signal detection theory’s distinction between sensitivity (d’) and response bias (criterion).
• Signal detection theory originated in WWII radar plane detection — Connects to the reader’s work on acoustic sensing applications — SDT’s military radar origin mirrors how acoustics often finds application in surveillance and detection contexts.
• Constant stimuli remains gold standard but is too slow for routine use — The speed-versus-accuracy tradeoff of threshold methods mirrors similar methodological tradeoffs in other measurement sciences.
• Psychoacoustics developed largely through communication technology challenges — Connects to the reader’s stated goal of understanding acoustic communication applications — the engineering/telecom origins of psychoacoustics are the historical root of applied communication acoustics.
• Anechoic chambers misrepresent how loudspeakers sound in normal rooms — Connects to standard audio-engineering practice and listener-test methodology used in loudspeaker reviews; also echoes the addendum article on how testing environments change perceived quality.
• Vision dominates hearing in localization, identification, and size judgment — Connects to ventriloquism effect and broader audiovisual integration literature; also relevant to spatial-audio and VR design where visual anchors dominate perceived sound location.
• Trained listener panels trade real-world representativeness for analytic test resolution — Connects to sensory-panel methodology debates in food and wine science, where trained tasters likewise diverge from consumer preference.
• Signal detection theory originated in WWII radar plane detection — Signal detection theory bridges radar engineering into psychoacoustics — a nice cross-domain case study of how a military engineering framework became the standard methodology for all threshold measurement.
• Listener personality biases reported auditory thresholds across subjects — Connects to signal detection theory and the response-bias problem familiar from psychophysics (e.g. Gescheider) and from yes/no testing methodology in hearing science.
• Adaptive staircase procedures trade efficiency and bias avoidance against convergence risk — Connects to psychophysical methodology discussions in hearing science and to the staircase method family from Levitt 1971 and Garcia-Perez 2011.
• Consensus-based sensory panels risk leader bias but training can counter it — Group consensus vs. averaged scores parallels the reputation systems and coordination problems from the law-merchant readings.
• Listening-test statistics must inform experimental design from the outset — Echoes standard research-design guidance across experimental psychology; connects to general knowledge about power analysis and stimulus selection.
• Repertory grid technique lets each subject generate their own descriptive dimensions via triadic comparisons. — Triadic elicitation (two similar, one different) resembles the method of sorted triads used in personal construct psychology and in some creativity research.
• Practical speech communication uses only a small slice of hearing’s full range — Directly relevant to understanding why communication systems target narrow bands; connects to Milgrom trade-range and bandwidth work and general info-theoretic design.
• Spectral masking by loud sounds is exploited to design lossy audio codecs. — Directly serves the reader’s application-focused goal: connects hearing science to a concrete engineering use (MP3/AAC perceptual coding).
• Uniform masking noise must drop 3 dB per octave above 500 Hz. — Practical detail for engineering uniform masking noise in test and calibration rigs.
• Complex real-world maskers are less effective than simple energetic ones predict — Informational masking is why real speech is easier to hear out of a babble than lab tones predict — directly relevant to cocktail party effect and hearing-aid design applications.
• Hearing groups frequency bands by shared envelope modulation, reducing masking — Explains how hearing groups simultaneous sounds by common modulation — a mechanism behind auditory scene analysis, connecting to cocktail-party-effect literature.
• Each inner hair cell responds best to one frequency plus a widening critical band — Critical band concept underpins psychoacoustic audio coding (MP3 masking) and speech intelligibility models — directly relevant to the reader’s acoustic-communication applications goal.
• Forward masking decays within roughly 200 ms after sound offset — This decay time sets limits on temporal masking in audio coding and on how quickly listeners recover sensitivity after loud sounds — relevant to acoustic communication systems.
• Longer masker sounds produce stronger forward masking — Shorter concurrent sounds that cause less forward masking matter for time-domain audio coding and transient sound design.
• Psychoacoustic tuning curves directly reflect cochlear mechanical processing. — Connects physical basilar-membrane mechanics to the perceptual phenomenon of masking — the bridge between acoustics and psychoacoustics relevant to applications like audio coding.
• Psychoacoustic tuning curves directly reflect cochlear mechanical processing. — Reinforces the mechanical-to-perceptual link in auditory modeling that runs through the earlier filterbank chapters.
• Human spectral analysis resolution runs about 10–15 percent of center frequency — Directly relevant to acoustic communication engineering: this is the usable resolution limit for designing spectral features in sound. Connects to the reader’s stated goal of acoustic communication applications.
• A 10-15 percent bandwidth window serves spectral analysis for audio tech — This rule of thumb underlies perceptual audio codec filterbank design and psychoacoustic models — an applied takeaway for the reader.
• Pitch cues convey identity, intonation, and emotion beyond mere speech content — Speaker-distinction, prosody, and object-size via pitch is a compact example for acoustic-communication application writing.
• Frequency JND is best around 500 Hz at about 0.3 percent — Relates to neural phase-locking limits, which connect to the broader theme of how auditory nerve firing encodes frequency, relevant to cochlear implant and hearing-aid signal processing design.
• Musical pitch perception spans roughly 30 Hz to 4 kHz. — Relates to the Mel-scale and auditory filterbank material from earlier Psychoacoustic chapters in this same book.
• Pitch perception requires a tone lasting at least two signal periods — Connects to the reader’s broader interest in how acoustic signals carry information — the minimum integration window for pitch is the temporal resolution limit of auditory feature extraction.
• Summing a signal with its delayed copy creates perceived pitch. — Connects to audio engineering knowledge of comb filtering and flanger/phaser effects, and to echo perception.
• Outdoor wall and ground reflections produce repetition pitch in real settings. — Connects to the reader’s ongoing soundscape work — real-world acoustic reflection phenomena in outdoor environments.
• Humans perceive a fundamental frequency even when it is absent from the signal — Relates to psychoacoustics and could map onto telephone/transducer design where low frequencies are absent but still perceived.
• Place and timing theories fail to explain some complex pitch perception phenomena — Connects to the chapter’s broader comparison between place and temporal theories of pitch perception, and to auditory perception topics in psychology.
• Absolute pitch is roughly ten times more precise than trained relative pitch. — Connects to debates about whether absolute pitch reflects perception or long-term memory — relates to general literature on expert calibration and reference standards.
• The mel scale was derived by asking listeners to double perceived pitch — The mel scale’s derivation-by-pitch-doubling connects to the earlier JND/psychoacoustic materials and is the perceptual basis of MFCC features in audio ML.
• Western orchestral tuning pitch has risen historically for a brighter sound — Pairs with the reader’s Complaints About Immodesty piece on shifting cultural norms, and general material on how standards drift over time.
• Place theory cannot fully explain how pitch is extracted from components — Place theory limitations echo timbre/loudness independence themes in the reader’s earlier excerpt on perceived-loudness scales.
• Loudness perception sums auditory bands and signals distance by comparison — Connects to earlier chapters on loudness models (ch 10 loudness formulas) and to the distance-estimation function of hearing mentioned in other psychoacoustics sources.
• Frequency spread in the cochlea is asymmetric with steeper low-frequency slopes — Connects to masking curve shapes described earlier in the book (Figure 9.6) and to ERB-rate scale concepts — the asymmetry explains why low tones mask high tones more than the reverse.
• AM tones change perceived frequency spread, giving extra detection cues that noise lacks. — Connects to the loudness/level-discrimination mechanism discussions in earlier chapters of this book on the basilar membrane excitation patterns.
• Amplitude-modulation detection becomes easier once modulation spreads the spectrum beyond one critical band — Connects to critical-band theory and spectral vs. temporal cue integration, relevant to cochlear implant and compression-algorithm design.
• Time-varying sound loudness is set by integrated loudness peaks, not moment-to-moment level — Connects to soundscape and noise-perception reading — this explains why intermittent sounds like passing aircraft or phone chirps are judged by peak loudness, relevant to the perceptual weighting behind noise annoyance.
• Brief silences are perceptually longer than their physical duration. — Directly relevant to the reader’s curated note in “Troubling Transparency” era: time-masking causing short silences in speech to be misperceived connects to the soundscape literature where silence duration perception underpins pause/beat work in Deep Listening and The Soundscape.
• Impulsive transients grab attention by signaling state changes in physical objects — Connects to the reader’s broader acoustic-communication interest: transient detection as an alarm/awareness mechanism parallels human alarm-signal evolution and modern alert-design.
• Auditory perception runs parallel bottom-up and top-down processes — Parallels visual attention models (e.g. Treisman’s feature integration theory) and cocktail-party effect literature.
• Humans can isolate a single speaker amid competing voices and noise — The cocktail party effect is the canonical example motivating computational auditory scene analysis and blind source separation algorithms.
• Machine hearing mirrors human perception by comparing feature vectors against known patterns — Connects to modern deep-learning audio classification and speech recognition pipelines that rely on feature extraction.
• The brain organizes simultaneous sound events into separate auditory streams via auditory scene analysis. — Connects to Bregman’s Auditory Scene Analysis, foundational for computational auditory scene analysis and hearing-aid design.
• Auditory spatial information alone allows orientation in one’s surroundings. — Connects to Deep Listening and soundscape work on how audition grounds spatial orientation.
• Individual notes are assigned to exactly one auditory stream, altering perceived rhythm. — Relates to Gestalt grouping by belongingness; ties to the reader’s interest in how perception organizes acoustic information.
• Psychoacoustic tonality measures tone-character, distinct from musical tonality terms — Connects to the book’s earlier foundational work on pitch salience, amplitude spectra, and noise versus tonal components developed across the previous hearing chapters.
• Hearing perceives synchronized neural firing across critical bands as buzzy timbre — Connects to the reader’s earlier reading on soundscapes and how signal structure is processed; group delay and phase perception tie to Deep Listening’s concerns about timbre perception.
• Synchronized phase harmonics produce a ‘buzzy’ sharpness perceptible without room reflections — Relates to headphone-versus-room listening differences relevant to binaural audio and loudspeaker design work.
• Group delay affects perceived quality only when it varies with frequency — Directly useful for two-way communication latency engineering, matching the reader’s goal of acoustic communication applications.
• Room reverberation can render phase-sensitive signals perceptually phase-insensitive — Relevant to perceptual audio coding design and convolutive blind source separation, where phase randomness in reverberant mixtures is a known obstacle.
• Peak dissonance between two tones occurs at a quarter critical band separation — Connects to critical band / auditory filter concepts from earlier chapters on the auditory periphery and to how masking and partial masking shape perception.
• Simple integer frequency ratios between notes produce the greatest measured consonance — This is the quantitative core beneath Pythagorean tuning and the historical ‘music of the spheres’ claims about integer ratios.
• Rhythm derives from natural body repetitions like heartbeat and walking. — Bodily grounding of rhythm connects to the notion of keynote/ambient sounds shaping human tempo in The Soundscape and to circadian/biological rhythms.
• Auditory perception runs parallel bottom-up and top-down processes — Connects to broad listening/soundscape concepts and to signal-vs-keynote frameworks from earlier reading on the soundscape.
• Gestalt grouping principles apply to hearing as well as vision — Connects to the reader’s soundscape work: audio-stream segregation is foundational for how keynotes, signals, and soundmarks are perceptually organized into coherent listening.
• Perceptual hearing properties map onto objective signal analysis quantities — Frames the bridge between psychoacoustic perception (Ch 1-2) and spectrogram/time-frequency objective analysis (Ch 3), which underlies audio quality evaluation (Ch 17) — architecture relevant to an applications-focused reading.
• Group delay perception thresholds depend strongly on signal content and context — Connects to everyday audio engineering constraints on loudspeaker element placement; a clean design rule (elements within 20-30 cm path difference) an engineer could apply directly.
• No tuning system can make all diatonic scale intervals perfectly consonant — The irreducible tuning tradeoff parallels the mathematical impossibility results in other domains—no perfect compromise satisfying all constraints simultaneously, like Arrow-style impossibility theorems.
• Humans are perceptually tuned to amplitude and frequency modulations across specific rate bands — Connects to the reader’s standing work on soundscape/auditory perception — modulation-rate bands map to how low-frequency ambient noise and vibration are felt vs. heard.
• Perceived musical dissonance reduces to roughness from beating partials. — Links to the reader’s deep-listening and soundscape material — roughness as a carrier of both musical meaning and environmental displeasure.
• Sound localization evolved in nocturnal mammals and remains vital for prey and predator. — Connects to the broader acoustic communication theme across the book — evolutionarily, spatial hearing was likely a precursor to any communicative use of sound.
• The outer ear’s resonant cavities enable elevation perception where other cues fail — Connects to acoustic ‘keynote/figure’ soundscape concepts; the pinna acts as a direction-dependent spectral filter akin to sonic ‘figure’ shaping.
• Mannequin heads with ear microphones capture realistic spatial audio for recording — Binaural microphone technique directly enables the immersive spatial audio applications the reader is studying.
• Spatial hearing is calibrated through learning to each listener’s head and pinna geometry — Connects to HRTF personalization and headphone rendering challenges in spatial audio engineering.
• Echoes and reverberation convey room geometry and surface properties to listeners — Connects to the earlier One Square Inch of Silence material about how soundscape reveals environmental structure — auditory spatial awareness as environmental sensing, relevant to acoustic ecology.
• Low interaural coherence makes sound events seem diffuse and surrounding — Links to the reader’s interest in acoustic spatialization applications (surround sound, binaural reproduction, room acoustics).
• Dynamic binaural head-movement cues override conflicting visual evidence of source location — Cross-modal cue weighting — dynamic auditory cues beating vision parallels known ventriloquism-reversal results where non-visual cues win.
• Synchronized visual motion captures perceived sound direction within 30 degrees — Directly relevant to binaural/augmented-reality audio applications where loudspeaker placement and visual content must align within ~30 degrees.
• Sound width perception depends on signal type, not physical source size — Connects to my soundscape reading — the perceptual gap between physical sound fields and auditory events echoes how lo-fi/hi-fi environments shape what listeners detect.
• The precedence effect suppresses spatial identity of reflections, not their audibility — Connects to Haas effect and echo suppression in audio engineering; also to cocktail-party effect literature.
• Neural refractory periods after strong onsets disable directional hearing for delayed sounds — Links to general neural adaptation/refractory mechanisms in sensory systems; also explains why transient-rich sounds localize best.
• Binaural listening yields a 5-7 dB intelligibility advantage over monaural — Directly supports claims about spatial release from masking; concrete quantitative value useful for hearing-aid design discussions.
• Speech elements near masked threshold carry intelligibility only under binaural listening. — Relates directly to the cocktail party effect and spatial release from masking, a core acoustic-communication application.
• The ear achieves spatial hearing through signal analysis rather than directional receptors — Connects to cognitive-science discussions of the eye-ear asymmetry in scene analysis (e.g. Bregman’s auditory scene analysis).
• Headphone audio internalizes because static binaural cues signal an inside-the-head source — Directly explains a well-known phenomenon in audio engineering and VR audio — connects to spatial audio and head-tracking technology applications.
• Evolution shaped localization to favor broadband transient sounds like movement — A functional-evolutionary reading of hearing that parallels adaptationist explanations for signal design in acoustic communication — connects to the acoustic-communication applications.
• Sound localization fuses independent sensory evidence into one coherent percept — Connects to the Deep Listening material already read about holistic vs. focused perception of the soundscape.
• Sound width perception depends on signal type, not physical source size — Directly relevant to the reader’s acoustic communication goal — explains why spatialized sound fields (e.g., a brook, crowd, wind) read as wide only when decorrelated noise; useful for psychoacoustic soundscape design.
• Two identical sounds within 30–40 ms are fused into one located auditory event. — Connects to listener envelopment and precedence-effect material in earlier sound-perception chapters of this same book.
• Anti-phase signals vs diotic noise yield 18 dB binaural detection advantage around 200-300 Hz — Directly serves the acoustic communication reading goal: this quantifies the upper bound of binaural processing advantage, relevant to hearing technology and sound source separation applications.
• Binaural listening gives 0 to 8 dB advantage in multi-talker conditions — Connects to the cocktail party effect discussion in Section 11.7.1 referenced here; also relevant to the acoustic communication reading goal about how listeners separate sound sources.
• The auditory system cancels room coloration to recover source signals — Connects to the broader arc in this book on how binaural hearing separates source from environment; also relates to how room correction and equalization systems (Toole 2006) replicate what the ear does naturally.
• Listeners perceptually subtract room coloration from source signals, called binaural decolouration. — Connects to Toole’s sound reproduction literature and to loudspeaker/room equalization practice — a listener-side complement to DSP-based room correction.
• Reverberator pre-delay controls perceived source distance independently of loudness ratio — Connects directly to the reader’s stated purpose on acoustic communication applications; a concrete tool audio engineers use to shape how listeners perceive spatial relations — parallels the pre-delay discussion in sound design for VR and film.
• The auditory system compensates for room acoustics to recover the original source signal — Connects to the cocktail party effect and to psychoacoustic work on binaural unmasking of sources in reverberation.
• Real-time auditory models enable speech recognition, sound reproduction, and hearing aids — Directly serves the reader’s purpose of understanding applications of acoustic communication: connects auditory modelling to engineering applications like hearing aids, spatial audio, and cochlear implants.
• DFT-based psychoacoustic models derive an auditory excitation spectrum on the ERB scale — Relates to signal processing familiarity with the DFT and to the ERB frequency scale from prior chapters on cochlear tuning and auditory filters.
• Hearing sensitivity to frequency can be emulated by filtering and weighting short-time power spectra — Connects to Chapter 9’s equal loudness contours and Chapter 10’s excitation patterns from earlier in this book; also to psychoacoustic modeling in audio engineering.
• Auditory spectra smooth harmonic detail while preserving formant structure — Connects to the broader soundscape/noise chapters read earlier: frequency warping here parallels how the ear’s nonlinear sensitivity shapes what signals matter, echoing how lo-fi environments erase distinctions.
• Gammatone filter banks model human auditory filtering using measured nerve responses — Directly relevant to models of speech and hearing used across the reader’s acoustic-communication corpus; addresses the mechanism question of how auditory nerve data becomes a computational filter.
• Gammatone filters fail on level-dependence and fast-onset sounds — The precedence effect is a key auditory phenomenon relevant to echo suppression in room acoustics and hearing-aid design; the slow-onset limitation constrains applications modeling transient sounds.
• Different hearing tasks exploit different model time constants. — Connects to application-specific modeling choices: loudness vs pitch need different temporal constants, relevant to the reader’s interest in acoustic communication applications.
• Hair-cell adaptation to sound arises from neurotransmitter pool depletion and recycling dynamics — This is the Meddis hair-cell model, widely invoked in auditory modeling literature; connects to the autocorrelation pitch model that follows in 13.11.
• Auditory adaptation acts like high-pass filtering that emphasizes stimulus onsets. — Onset emphasis links to soundscape work on how the auditory system flags new/changing events over steady states.
• Cochlear modelling faces a tradeoff between spatial accuracy and computational cost — Transmission-line electrical-equivalent modelling echoes Milgrom’s and other accounts where a physical system gets mapped onto an analogous formal structure.
• Two competing loudness models predict time-variant loudness by peak tracking or frame averaging — The temporal-integration and time-varying-loudness mechanisms parallel the reader’s prior work on how the ear averages across frames to handle transient sounds.
• Autocorrelation models explain pitch and can separate concurrent vowel sounds differing in fundamental frequency — Connects to the reader’s soundscape/psychoacoustics reading — autocorrelation as the mechanism behind both pitch and source segregation (cocktail-party effect).
• Auditory modelling accuracy trades off sharply against computational cost — Connects to perceptual audio-coding (MP3/AAC psychoacoustic models) which use DFT-based processing — bridges hearing science to consumer audio technology.
• Binaural unmasking can be modelled as equalizing then subtracting two ear signals — The equalization-cancellation concept is essentially a reference-channel subtraction — the same principle behind noise-cancelling headphones and adaptive beamformers; applies directly to hearing-aid algorithm design.
• Cross-correlation sidelobes near 1 kHz stem from interaural phase ambiguities — Phase ambiguity causing spatial localization errors near 1 kHz connects to the classic ‘cone of confusion’ and why low-frequency ITD cues break down — relevant to HRTF-based audio rendering.
• Required audio fidelity scales with the application’s communicative demands — Connects to information-theoretic ideas of necessary vs. redundant signal features in a communication channel.
• Authentic sound reproduction is judged by perceptual identity, not signal fidelity — Connects to the reader’s work on silence/audiology and the reproduction of natural soundscapes for hearing aid testing.
• Spaced mics sound warm and ambient because incoherent signals exceed loudspeaker angle. — The warmth claim (Lipshitz 1986) connects to earlier chapters on interaural cues and diffuse-field perception in this same book.
• Spherical harmonics act as basis functions for building any microphone pattern — Connects to Fourier-style decomposition thinking - spherical harmonics are the ‘Fourier series’ of sound field directionality.
• Sound within one critical band is perceived as a single source — The principle that perceptual limits allow simpler reproduction connects to auditory perception work already read, and to the general theme that perceptual ceilings license engineering shortcuts.
• DirAC reproduction quality depends on one dominant source per frequency band — The DirAC model constraint (one dominant source per band) parallels beamforming/localization assumptions in speech separation and other time-frequency masking work.
• Room reverberation suppresses the comb-filter colouring of amplitude panning — Connects to my noise/silence reading: reverberation here functions as a protective mask, echoing how keynote ambient sound can wash out signal details.
• Recorded audio is engineered for plausibility, not faithful reproduction of reality — Connects to the book’s earlier theme of perceptual rather than physical sound — audio targets listener expectations, not acoustic truth.
• Object-based audio decouples soundtracks from specific loudspeaker layouts — Connects to the reader’s interest in acoustic communication applications — this is the audio equivalent of format-agnostic information design, applicable in tech and display contexts.
• Human hearing adapts to different room acoustics to preserve timbre perception — Connects to the reader’s broader theme of silence and sound perception — the ear canal signal being radically different yet perception constant is a form of perceptual constancy like the figure-ground/adaptation ideas in The Soundscape.
• Human audio-visual synchronicity tolerance is asymmetric because sound naturally lags light. — Connects to the reader’s soundscape/silence work: the same physical fact about sound and light propagation speeds that shapes perception of distance also shapes content-production timing standards.
• Auditory-tactile systems improve music reproduction quality and enable tactile feedback. — Connects to chapter 12’s discussion of multimodal processing and to haptic interfaces in human-computer interaction literature, e.g. perceiving identity through texture/vibration on touchscreens.
• Directional coincident microphones suppress room reverberation, killing sense of space — Connects to the reader’s existing knowledge of room acoustics/reverberant field levels from the telephony discussion at location 1833.
• Stereo loudspeakers create directional illusion through summed phase shifts mimicking time delays — The frequency-dependent directional ambiguity mirrors how the brain resolves sound location generally — connects to work on auditory scene analysis and the duplex theory of localization.
• Sound field control can deliver different audio to different zones of a room. — This is the acoustic version of ‘personal audio zones’ that consumer tech has long pursued; connects to discussions of directional speakers and sound beams like the Audio Spotlight.
• DIY in-ear binaural recording yields convincing spatial reproduction — Connects to ASMR and binaural video communities that trade on exactly this diy in-ear recording effect.
• Each person’s unique pinna and head shape undermines binaural recordings for other listeners — Connects to the individuality/bespoke measurement problem that also plagues the diamond-industry and reputation-system literature on bespoke verification.
• HRTFs can virtualize multi-channel loudspeaker layouts on headphones. — Applies directly to the acoustic-communication applications lens: spatial audio rendering of existing multichannel content.
• FFT block convolution trades speed for double-block latency in reverbs — Links to standard DSP textbooks (Oppenheim and Schafer cited) and the general FIR-filter latency problem.
• Phase-based localization sensing has a hard frequency ceiling near 1.6 kHz — The head-size-to-wavelength ambiguity threshold (~800 Hz where half-wavelength meets head diameter) echoes other engineering tradeoffs where a physical dimension and signal wavelength interact — connects to wave mechanics, antenna theory, and any physical sensing designs that must stay below a resolution cutoff.
• Time–frequency audio processing exploits the ear’s own spectral analysis structure — Connects to Chapter 3’s discussion of the ear canal signal time–frequency analysis (Section 3.2.6); echoes how cochlear mechanics perform spectral decomposition discussed in earlier chapters.
• Downsampled filter banks enable efficient DSP implementations like FFT for DFT. — The FFT-as-efficient-DFT point is a foundational signal-processing axiom connecting to the reader’s broader acoustics knowledge base.
• MDCT excels for coding but lacks robustness to time-frequency domain edits — Directly connects to why spectral editing/denoising tools prefer STFT — the standard practical choice across audio restoration literature.
• Modifying STFT bins creates edge transients that synthesis windows imperfectly suppress — Explains the classic STFT flickering/aliasing tradeoff against time-frequency resolution that connects to phase vocoder and audio effect literature.
• Fixed-width filter bands mismatch hearing’s finer low-frequency resolution — Connects filter bank design to ERB/Bark auditory frequency-resolution models from Section 9.4.3 of this same book.
• Audio analysis bands can follow perceptual scales while processing stays on full-resolution QMF bands. — Connects to the Bark scale and perceptual audio coding literature the reader knows from psychoacoustics.
• Decorrelators degrade audio quality on transient-rich signals like applause. — Links to spatial audio/spatialization applications where decorrelation is used to widen sound; the applause failure case is the classic test signal in psychoacoustics.
• Human spatial resolution limits make stereo-to-multichannel upmixing perceptually plausible. — Connects to psychoacoustics of spatial resolution limits familiar from soundscape and binaural literature.
• Low interaural coherence makes sound events seem diffuse and surrounding — Directly relevant to spatial audio and loudspeaker reproduction design — connects to how the reader’s stated goal of acoustic-communication applications would encode spatial breadth.
• Mobile phone codecs adapt their bit rate across eight levels to optimize call quality — Connects to the reader’s acoustic engineering background on codec standards (G.729 CS-ACELP, AMR-NB).
• Logarithmic quantization suits speech but distorts music beyond tolerance — Ties to the earlier chapters on the matching of coding (mu-law companding) to signal statistics — a classic signal-class-properties point.
• Synthetic speech dates to late-18th-century acoustic-mechanical machines — Ties to the historical-thread interest; synthetic voice machines as an early acoustic communication technology.
• Rule-based speech synthesis survives only where computational power is scarce. — Connects to the broader pattern in computing where data-driven approaches displace rule-based heuristics once compute becomes cheap — parallels the shift seen in machine translation and image recognition.
• Human speech evolved as a robust channel tolerant to high background noise — Connects to the reader’s noise/soundscape reading: the acoustic robustness of speech against a noisy background is what noise pollution degrades.
• Source-filter parametric synthesis enables voice modification that unit-selection cannot. — The source-filter model of speech connects to the reader’s broader acoustics knowledge of vocal tract filtering.
• Speech recognition shifted from time warping and shallow nets to HMMs then deep networks — Connects to the broader arc of neural-network history (the ‘AI winters’ and deep-learning resurgence) also covered in histories like The AI Con.
• The hardest part of speech-to-speech translation is language processing, not signal processing. — Connects to modern LLM-based real-time translation apps and voice-cloning preservation of speaker characteristics.
• Speech coding is the most mature speech technology because it needs no databases. — Relates to the book’s broader theme that model-based approaches beat data-hungry ones when a physical mechanism is understood.
• Spatial hearing is worst near 1.7 kHz where head wavelength matches. — Connects to head-related transfer function design in spatial audio rendering.
• Synchronized visual motion captures perceived sound direction within 30 degrees — Connects to design of audiovisual displays and VR/AR systems where vision and audio must be aligned.
• Humans localize single sources well but fail at source distributions — Relates to cocktail-party problem and auditory scene analysis literature.
• Sound quality judgments depend on listener context and expectations — Connects to Chapter 8 psychoacoustics stripping out cognitive factors, and to the reader’s broader interest in acoustic communication applications — the same signal (intelligible speech) flips valence by context.
• Commercial telephone bandwidth standards derive from interwar human listening tests — Connects to the reader’s concern with practical applications of acoustic communication — this is the canonical origin of psychoacoustics as an engineering field.
• Noise engineering now targets annoyance and quality, not just hearing loss — Directly connects to the reader’s existing knowledge base on noise, silence, and soundscape — annoyance-based noise metrics expand on A-weighted exposure measures.
• Digital audio coding made perceptual modeling central to audio engineering — Links to psychoacoustic masking models used in MP3/AAC codecs — a concrete application of the acoustic-perception models central to the reader’s stated goal.
• Beyond intelligibility, telephony must preserve speaker recognizability — Connects to synthetic-voice quality debates and TTS system design; the point that synthetic speech can be measured with the same methods as natural speech is directly applicable.
• Reverberation low-pass filters speech modulation while noise raises its floor — Connects to how the modulation transfer function decomposes a room’s degradation of intelligibility into these two separable mechanisms.
• Hearing damage begins after long exposure above 85 dB — Connects to Hempton’s One Square Inch of Silence coverage of noise exposure thresholds and hearing risk.
• Computational sound quality models are trained on subjective listening data — Connects to the book’s overall theme of computational auditory models (e.g., auditory periphery modeling earlier in Pulkki) predicting perceptual outcomes.
• Multi-stimulus quality tests fail when samples vary in several perceptual dimensions at once — Parallels a known limitation in multidimensional scaling and survey design where confounded dimensions produce uninterpretable data.
• MUSHRA only yields reliable results when sample differences are easily perceptible — Connects to psychometric limitations of rating scales — links loosely to how subjective scales saturate or discriminate poorly at small effect sizes.
• Human auditory adaptation neutralizes room acoustics but not device imperfections — Connects to psychoacoustic adaptation/auditory scene analysis literature and to the book’s earlier section on human adaptation to acoustic conditions.
• Audio quality bottlenecks shifted from storage to transducers — Connects to the history of high fidelity audio and digital audio codec development covered elsewhere in the book.
• PEAQ uses a trained neural-network cognitive model to predict human MOS ratings. — Resonates with how trained pattern-recognition substitutes for human panels, paralleling trust/reputation systems that formalize subjective judgments.
• Human auditory memory cannot compare sounds separated by more than a few seconds — This limits of auditory memory constraint matters for A/B testing methodology generally, also relevant to the Milgrom institutional work on information constraints.
• Auditory-model quality metrics fail on acoustically captured signals — Limits of instrument-based quality measures parallel the Milgrom point that full-information institutions are too costly — bounded applicability of tools.
• Binaural auditory models cannot yet explain human perception of complex spatial audio — Connects to PEAQ objective quality measurement standard and psychoacoustic modelling from earlier audio engineering reading.
• Spatial audio synthesis quality is measurable against a real-source reference. — Connects to PEAQ audio quality standards and psychoacoustic test methodology from prior audio engineering knowledge.
• Speech intelligibility metrics can be predicted from impulse response and noise — A computational acoustic analogue of the chapter’s modulation-reduction framework, bearing on how intelligibility prediction serves PA system and hall design.
• Reverberation time maps directly onto equivalent signal-to-noise ratio — Connects to the noise and acoustics reading everywhere in this library — reverberation degrading intelligibility is equivalent to additive noise in STI terms.
• Reverberator pre-delay controls perceived source distance independently of loudness ratio — This is a practical audio-engineering tool relevant to the reader’s interest in applied acoustic communication: controlling perceived distance of voice in mix and VR.
• Speech intelligibility alone cannot capture loss of speaker identity in codecs — Connects to modern telephony/VoIP codec analysis and the broader limitation of single-metric audio quality assessment.
• The first objective speech quality measure failed on mobile-network timing variations — Connects to the broader story of why objective metrics fail in real-world conditions, echoing the FAHQMT debate and listening-test methodology threads.
• Telecom device sound quality depends on network, room, and feedback conditions jointly — Reinforces the systems view: quality is not a device property but an emergent joint outcome, parallel to how the book treats room acoustics and source-environment coupling.
• Industry relies on objective methods because exhaustive listening tests are impractical — Connects to the recurring theme of perceptual models substituting for human listeners across audio quality measurement.
• Sound perception is shaped by the listener’s expectations and context of use — Connects to the reader’s work on silence/noise and how context gives meaning to sounds
• Entering a car generates designed auditory-tactile impressions of finishing quality — The idea that small physical gestures and their sounds carry deliberate style parallels the reader’s article on trees as infrastructure and on dress codes — all about designed surfaces communicating meaning
• Noise suppression algorithms can degrade speech quality in non-stationary noise — Connects to the reader’s noise-and-silence corpus: noise-canceling and suppression technologies as a form of sonic intervention that introduces its own failures, echoing Hempton’s anti-noise-device critique.
• Echo delay thresholds produce distinct perceptual and speaking effects — Connects to auditory feedback control of speech production, which runs parallel to the book’s earlier discussion (in Soundscape material) of how ambient sound shapes production.
• VoIP call delays exceed those of regular mobile calls — Connects to the practical telecom engineering dimension of the reader’s acoustic communication goals.
• Concert hall acoustics shape the music composed for them — Reversing the usual causality — architecture influencing art rather than art demanding architecture — parallels themes in the reader’s trees-as-infrastructure piece.
• Reverberation time is the key design variable separating concert hall types — Connects to the reader’s silence/acoustic-composition work and to how performance spaces shape the sound experience.
• Standardized concert hall acoustics metrics have known technical limitations compared to human perception — Connects to Chapter 16/17 themes of auditory models versus physical measurements, and to the book’s running tension between objective metrics and subjective listening.
• Speech intelligibility harms open-plan office quality, unlike in auditoria — Connects to reading purpose of acoustic applications — this is a direct design specification arising from communication acoustics.
• Objective models of subjective annoyance remain limited to specific noise types — Directly relevant to reader’s application goal — annoyance models as an acoustic-communication application domain.
• Sharpness, roughness, and narrowband temporal components increase noise annoyance — Connects to the reader’s deep work on the soundscape (Schaffer) and loudness/annoyance research in One Square Inch of Silence.
• Computational simulation is the best method to model the auditory system’s function — Explains why simulation rather than analytic math dominates auditory modelling, a methodological point relevant to any acoustic-communication application reading.
• Interactive game audio requires dynamic rendering systems beyond traditional linear production. — Connects to procedural/adaptive systems thinking in software design and to generative media debates.
• Image-source room acoustics modelling becomes computationally costly in interactive VR — Echoes the audiovisual VR discussions in Ch. 17 and the real-time constraints framing throughout the book.
• Virtual sound source distance is limited by reproduction technology, not perception — Connects to binaural cues and direct-to-reverberant ratio from Chapter 12 discussed in a prior reading of this book.
• Sonic interaction design studies sound perception in active embodied contexts rather than passive listening — Parallels the shift in deep listening / soundscape work toward embodied, active listening.
• Untreated hearing loss in children was once misdiagnosed as intellectual disability — Connects to the general history of misdiagnosis and to how sensory deficits shape early development, elsewhere in this book and in disability studies.
• Excess noise damages cochlear hair cells through several distinct mechanisms — Connects directly to the reader’s prior work on noise exposure, soundscapes, and silence research where noise-induced hearing loss was a recurring theme.
• Outer hair cell damage compresses dynamic range, causing recruitment and distorted loudness perception — Connects to complaints about hearing aids overcompensating for loud sounds — the loudness discomfort level shrinks precisely because recruitment overemphasizes dynamic range.
• Outer hair cell damage destroys the cocktail party effect — Connects to noise research from One Square Inch of Silence and The Soundscape — the physical basis for why the noise-battered world degrades comprehension.
• Outer hair cell damage widens critical bands and worsens speech-in-noise — Connects to the reader’s wider interest in acoustic communication applications — a concrete mechanism for why hearing-impaired listeners struggle specifically with speech in noisy environments, relevant to hearing aid and cochlear implant design.
• Annoyance and disturbance refer to distinct degrees of noise harm — The harm-vs-annoyance split parallels the objective/subjective distinction threaded through earlier soundscape reading.
• Noise is the most common cause of work-related impairments — Connects to the noise-as-public-health-crisis theme in the earlier silence/noise books.
• Hearing-loss risk dominates noise research because it is predictable from sound properties — The same loud music as pleasure-or-terror depending on context echoes the subjective/objective noise tension from the soundscape literature.
• Tinnitus is treated by masking or habituation, not cured — Sound-masking and habituation therapy connect to soundscape/active-listening ideas from earlier reading on silence and acoustic ecology.
• Single-channel noise reduction exploits speech’s temporal, spectral, and statistical properties — Connects to statistical speech enhancement and spectral subtraction methods used in telephony and voice assistants.
• Binaural hearing aid processing preserves spatial cues that monaural processing distorts — Connects to the reader’s broader interest in acoustics applications in medicine/assistive technology; relates to interaural level difference concepts.
• Most modern cochlear implant patients achieve telephone conversation ability — Connects to the reader’s broader interest in acoustic communication applications — outcomes data for a major assistive acoustic technology.
• Adult cochlear implant success depends on prior experience with sound — Connects to critical-period / neural plasticity literature familiar to the reader.
• Hearing-loss risk threshold should drop from 85 to 75 dB for full protection — Connects to noise research in One Square Inch of Silence and the conflict between regulatory floors and actual protection.
• Even moderate 70 dB sound causes temporary hearing threshold shift with hours of exposure — Connects to earlier soundscape work on chronic urban noise levels and their health effects.
• Noise control should prioritize source reduction then path attenuation before personal protection — Mirrors the hierarchy of controls in occupational safety and the source-first ethos of the earlier soundscape books.
• Brief unprotected noise exposure during otherwise protected periods raises hearing loss risk — Relates to the noise-exposure / hearing-loss causality chain from chapters on the auditory system and hearing damage.
• Bone conduction through the skull shapes how one hears one’s own voice — Explains why recordings of one’s own voice sound alien — a well-known phenomenon; connects to general audio-perception knowledge.
• Hearing-aid users and small children require loudspeaker-based audiometry for accurate coupling — The uncontrolled coupling problem parallels the calibration and placement concerns in loudspeaker measurement and room acoustics discussed elsewhere; also connects to hearing-aid engineering design constraints.
• Pressure differences across the eardrum attenuate low-frequency sound and cause pain — Fits the reader’s interest in acoustic applications — mechanism explaining flight-related hearing loss and ear pain.
• Hearing aids are miniature signal-processing sound systems that must secure adequate signal-to-noise ratio — Connects to the reader’s existing work on acoustics and soundscape — SNR as design target parallels the emphasis on signal-versus-noise throughout.
• Hearing impairment degrades signal-to-noise tolerance more than quiet speech perception. — Directly bears on the SNR / cocktail-party hearing literature; also connects to spatial hearing and binaural unmasking topics in earlier chapters of this book. Seed
• Helmholtz resonance shapes violin bodies, speaker designs, and car cabin noise — A single physical principle — Helmholtz resonance — explaining a violin’s voice, a subwoofer’s cabinet, and a dangerous car-window drone is a clean pop-science essay structure.
• Early reflections boost perceived loudness while loud late reverberation degrades intelligibility — Possible article angle: how the 60ms/100ms reflection window is exploited in concert hall design and hearing aid algorithms.
• Linear prediction order scales with sample rate to capture speech formants. — Good rule-of-thumb detail for a practical article on speech analysis or audio DSP tooling.
• Computational models can synthesize natural ambient sounds like footsteps and thunder — An angle on procedural audio for virtual environments — how sound design in games leans on physics-based synthesis rather than sample playback.
• Single-neuron auditory responses are mapped invasively in animals only — The methodological constraint — invasive single-neuron recording restricted to animals — is worth explaining in any accessible piece on how we know what we know about the auditory pathway.
• fMRI identifies active brain regions well but is noisy and time-slow for auditory study. — An essay angle: the irony that the loudest research instrument in cognitive science is used to study hearing — how researchers engineer silence into noisy machines.
• Signal detection theory originated in WWII radar plane detection — Angle: how WWII radar detection problems became the backbone of modern psychological measurement of perception — methodological diffusion from engineering to experimental psychology.
• Pitch cues convey identity, intonation, and emotion beyond mere speech content — Pitch as a multi-function acoustic cue — identity, emotion, and physical size — could anchor an essay on how one signal parameter serves many communication channels.
• Louder signals need disproportionately more dB to double perceived loudness — Angle: why audio mixing/mastering and hearing-aid gain use phon-like scales — practical engineering consequence of the sones-vs-phons nonlinearity.
• A sound’s onset transients largely determine instrument identity — Strong essay angle: audio forensics and synthetic instrument design lean on onset cues — the whole field of tactile transducers and excitation vocoders exploits this.
• A single note can sound rough only below roughly 100 Hz fundamental — Angle for an article: synthetic bass sounds (electronic music sub-bass, dubstep) inherit roughness that acoustic instruments escape because their harmonics naturally decay — a perceptual argument about synthesized vs. acoustic timbre.
• Reversing a signal’s polarity changes its perceived bass level — An article angle on how identical frequency spectra can nonetheless produce perceptibly different timbres purely through phase — implications for audio engineers and equipment reviewers.
• Headphone audio internalizes because static binaural cues signal an inside-the-head source — Article angle: why ordinary headphones always sound ‘inside your head’ and how head-tracking binaural solves it — a concrete explainer of a phenomenon everyone has experienced.
• Headphone audio internalizes because static binaural cues signal an inside-the-head source — Angles for an applied-audio piece: why head-tracked audio (binaural rendering that responds to head movement) is the missing ingredient for convincing externalized AR/VR sound.
• Evolution shaped localization to favor broadband transient sounds like movement — Angle: why warning and alarm sounds (in tech, urban design, nature) all converge on broadband transients — evolution has pre-tuned our ears to a specific acoustic signature.
• Hearing evolved as an omnidirectional alarm system complementing directional vision — The counterfactual — engineered ears with acoustic lenses — is a fun speculative-design angle for art or explainer writing.
• Reverberator pre-delay controls perceived source distance independently of loudness ratio — A technical explainer angle: the reason a long gap between direct and reflected sound must mean the source was close — how reverbs fool the auditory system via physics.
• Auditory models that extract distance cues from ear signals remain an open research direction — The open problem of extracting distance perception from room responses is a natural angle for an essay on the unsolved mechanics of spatial hearing and its engineering implications.
• Human audio-visual synchronicity tolerance is asymmetric because sound naturally lags light. — An essay angle on how a natural physical asymmetry (the speed of sound) quietly hard-codes itself into broadcast standards like the ITU recommendation.
• Visual color affects perceived loudness of the same sound — Potential essay: how multisensory integration means sound design for media must account for visual emotion cues, not just acoustics.
• Sound field control can deliver different audio to different zones of a room. — Angle: sound field control as the future of personal audio — silent private zones in public spaces without headphones.
• DIY in-ear binaural recording yields convincing spatial reproduction — A cheap practical how-to: replacing headphone drivers with microphones to make binaural recordings — an engaging hands-on application angle.
• Logarithmic quantization suits speech but distorts music beyond tolerance — The difference in how speech vs music tolerates quantization distortion is a neat angle for a piece on why codec design is media-specific.
• The hardest part of speech-to-speech translation is language processing, not signal processing. — Article angle: why speaking-to-speaking translation remains unsolved decades later — the bottleneck is linguistic, not acoustic.
• Auditory elevation calibration requires visual confirmation of sound sources — Angle: why our spatial hearing is trained by what we see — implications for blind users of spatial audio and for audiovisual VR calibration.
• Musical instrument design evolved as early product sound quality experimentation — Angle: luthiers as proto-audio-engineers — gradual experimentation centuries before formal sound quality science.
• Commercial telephone bandwidth standards derive from interwar human listening tests — A piece on how the telephone bandwidth we still use today was set by 1920s-30s test subjects, a legacy infrastructure standard encoded in hardware for a century.
• Perceptual auditory models will eventually replace traditional audio metrics — Angle: predicting which traditional measurement regimes are on the verge of perceptual-model replacement — a technology-adoption pattern worth tracing across industries.
• Reverberation time maps directly onto equivalent signal-to-noise ratio — Angle for an essay: reverberation is speech’s ‘acoustic noise’ — a measurable equivalence between room reflections and additive interference that lets architects design for intelligibility.
• Reverberator pre-delay controls perceived source distance independently of loudness ratio — Article angle: the subtle audio tools engineers use to fake depth — pre-delay versus direct-to-reverberant ratio as two independent distance levers.
• Speech intelligibility alone cannot capture loss of speaker identity in codecs — Possible essay angle: the gap between intelligibility and identity in speech codecs, and what acoustics reveals about the limits of optimization metrics.
• Consumer devices are engineered to sound masculine and powerful, not merely quiet. — Pitch for an essay on the gendering of product sounds — how appliances are audio-engineered to perform masculinity, complementing the existing article on immodesty moral panics which also treats consumer/identity signaling.
• Sound design in interfaces is an extension of product sound quality — Sonification of drill angle is a concrete application case worth building into a broader article on acoustic UX and data sonification.
• Music databases need automatic high-level content querying beyond raw audio signals. — The concrete query example could anchor an essay on the gap between signal-level storage and semantic-level search in media libraries.
• MIR metadata is largely assigned by expert listeners, motivating automation as music creation accelerates — Angle: the hidden human labor — armies of expert listeners — behind commercial song-recognition services, and why automation is eroding that labor force.
• PLP coefficients describe speech spectral features compactly and speaker-independently — The trick of bouncing from loudness spectrum back to autocorrelation via inverse FFT before running LP is a genuinely elegant technical maneuver — a piece for audio-tech readers on how perception-aware front-ends work.
• Speech processing can fail independently of general hearing ability — Possible angle: dissociable auditory deficits — many people ‘hear fine’ but can’t parse fast speech — relevant to why captions help, or to audio design for accessibility.
• Post-noise tinnitus and distortion signal auditory overload and hearing-loss risk — Could feed an article on neglected warning signs of hearing damage beyond loudness thresholds. Define
• Noise is defined by annoyance, disturbance, or hearing harm rather than loudness alone — Gives the operative definition of noise as an attention-and-harm-driven concept, not a purely physical one.
• Sound technology work needs acoustics, perception, and signal processing jointly. — Frames the field’s three cornerstones — a core framing device for the whole discipline of communication acoustics.
• Vision dominates audition except where visual events are absent or ambushed — Defines ‘sensory integration’ as fusion of different-modal information into a coherent percept.
• Hearing serves both environmental orientation and communication since early animal evolution — Introduces the two-way categorization of auditory use: orientation in environment vs. communication between subjects.
• Noise is defined as sound that annoys, disturbs, or risks hearing damage — Defines ‘noise’ along a subjective/harm spectrum — annoying, disturbing, or dangerously loud sound.
• Communication acoustics joins basic research and engineering toward shared goals — Establishes the field’s dual research-versus-engineering frame, useful as a definitional anchor for the whole book.
• Very general theories guide thinking but resist direct application to specific cases — Defines theory formation as the third approach, highlighting its generality-applicability tradeoff.
• Computer simulation dramatically expanded modeling’s power over physical models — Defines modeling as simulation of a system, and marks the shift from physical to computational models.
• Physical, biological, and engineered systems share describable general properties — Defines what a systems approach means in this framework — abstraction and modelling based on shared general properties across domains.
• Systems are analyzed through complementary structure and function concepts — Defines core systems-theory vocabulary (structure, function, event, state, control, hierarchy) used across communication acoustics.
• Natural and engineered systems organize as vertical control hierarchies — Defines hierarchy as vertical control flow, with lower units handling detail and higher units strategic control.
• A system’s output depends on its remembered history — Explains system state as memory of prior behavior influencing current output.
• Heterarchies distribute control for versatility while data become information through observation — Defines heterarchy, data, information, and knowledge as a nested ladder.
• Audio and speech technology design should be grounded in perceptual mechanisms — Scopes what ‘perception-based sound technology’ includes: audio, speech, noise techniques, hearing aids, audiology.
• Multidisciplinary problems cannot be mastered within a single knowledge domain — States the core premise justifying why communication acoustics requires spanning acoustics, signal processing, physiology, and engineering.
• Abstracting away details and concretizing them frame scientific research as one cycle — Defines abstraction and concretization as the two poles of a single bidirectional methodology loop.
• Heterarchical, distributed control can outperform strict hierarchy under changing conditions. — Defines heterarchy versus hierarchy, and the data-information-knowledge cascade, with language as knowledge’s transmitter.
• Sound as physical event and sound as auditory perception are distinct — Establishes the core distinction between physical sound events and subjective auditory perceptions, the foundational disambiguation for the whole field.
• Most sound arises from vibrating mechanical objects in a medium — Defines sound physically as wave propagation from vibrating sources, with lightning as the notable exception.
• Compressible air in a volume forms a mechanical spring for resonance — Explains the physical mechanism behind the Helmholtz resonator concept.
• Coupled masses vibrate either along or perpendicular to spring direction — Defines longitudinal versus transversal vibration for coupled spring systems.
• Any vibration decomposes into weighted sums of a system’s modal oscillations — Explains why modes are mathematically fundamental: they form a basis for all possible vibrations.
• Coupled oscillators produce complex resonances shared by instruments, voices, and rooms — Introduces what modal behaviour means for composite systems like strings, air columns, and rooms.
• Opposing traveling waves of the same frequency create standing waves with nodes. — Defines standing waves and nodes, and clarifies the mechanism by which they form.
• Sound pressure is tiny pressure deviation, and human hearing spans a specific Pa range — Defines sound pressure as deviation from static pressure and gives the human hearing range in Pascals.
• Phase alignment determines whether sounds sum by amplitude or by power — Defines coherent vs. incoherent sound sources — the core distinction governing all sound-field superposition.
• The absorption coefficient measures the fraction of incident sound energy absorbed. — Defines the absorption coefficient, a core quantity in room acoustics and acoustic design.
• Relative motion between a sound source and listener shifts perceived pitch — Establishes the mechanism underlying the Doppler effect in acoustics.
• The Helmholtz resonator models confined air as a spring with an air-mass plug — Helmholtz resonator mechanism clearly defined: compressible air volume is the spring, tube air is the mass.
• Coherent sounds interfere by amplitude while incoherent sounds sum by power. — Establishes the coherence/incoherence distinction governing whether sounds add in amplitude or power.
• Sound intensity is a vector describing the directional flow of sound energy. — Defines sound intensity as vectorial flow of energy through a unit area, distinguishing it from scalar pressure.
• Impedance discontinuities split an arriving wave into reflected and transmitted parts — Impedance mismatch as the governing principle of reflection — foundational to understanding horns, ear canals, and architectural acoustics.
• Sound diffraction arises when a wavefront’s edge acts as a secondary source — Clarifies reflection as backscattering and diffraction as scattering from edges and boundaries.
• Early reflections boost perceived loudness while loud late reverberation degrades intelligibility — Clarifies the distinct functional roles of early reflections versus late reverberation in room acoustics — a core concept for acoustic communication design.
• Early reflections boost perceived loudness while loud late reverberation degrades intelligibility — Defines the distinction between early reflections and late reverberation by their perceptual functions and time thresholds.
• Sound from a source falls 6 dB per distance doubling until reverberation takes over — Defines the reverberation distance — the crossover point where direct and reverberant fields are equal, a key room-acoustics concept.
• In steady state, reverberant sound level is uniform throughout a room. — Explains the core prediction of statistical reverberation theory — that reverberant field level does not depend on receiver position.
• Above a critical frequency, room acoustics become statistically tractable as diffuse fields. — Defines critical frequency and diffuse field, and explains why diffuse fields allow statistical rather than exact modelling.
• Modal density scales with spatial dimension, from flat in 1D to squared in 3D. — Explains the concept of modal density and how it scales with the dimensionality of the acoustic system.
• Room reflections can be modelled as sources in mirrored rooms — Explains the image source method concept clearly: reflections modeled as sources in image rooms.
• Ray-tracing and image-source methods fail on low-frequency diffraction at corners — Introduces FEM and explains why it beats geometric methods where diffraction matters.
• Room acoustics can be simulated by statistically tracing sound rays until their energy decays — Explains how ray-tracing simulation works conceptually for a reader new to acoustic modeling.
• Non-linear acoustics governs high-pressure waves like shocks and cavitation — Contrasts linear vs non-linear acoustics domains, naming shock waves and cavitation.
• Above a critical frequency, room acoustics become statistically tractable as diffuse fields. — Explains the critical frequency and the statistical definition of the diffuse field.
• Room reflections can be modelled as sources in mirrored rooms — Explains the image source method: each wall reflection maps to a virtual source behind the wall in a mirrored room.
• Converting continuous sound to digital requires time sampling then numeric quantization — Clean step-by-step definition of the analogue-to-digital signal chain that underpins all acoustic communication processing.
• A sound signal is the registered time-varying vibration of a medium — Defines what a sound signal fundamentally is before any processing or representation happens.
• LTI systems cannot create new frequency components from their input signals — Defines the key distinguishing property of LTI systems in a crisp, load-bearing way.
• Non-linear and time-variant systems generate frequency components absent from the input — Contrast pair defining time-variant vs non-linear, both of which produce new spectral content.
• The Fourier transform trades costly time-domain convolution for simple frequency-domain multiplication — Explains the convolution theorem mechanism that underpins digital signal processing efficiency.
• The Fourier transform measures how a signal matches each sinusoid pair — Gives an intuitive, non-formulaic gloss of the Fourier transform as correlation with sine/cosine basis functions — useful for explaining spectral analysis.
• Any continuous signal decomposes into sinusoidal partials and convolution simplifies in frequency domain — Explains why the Fourier approach is central to all audio processing: decomposition into partials plus the convolution-to-multiplication trick are the load-bearing ideas.
• The auditory system largely ignores signal phase, driving spectral analysis conventions — Explains why spectral analysis conventionally means magnitude-only: a perceptual fact about the ear shapes a whole technical tradition.
• Group delay of modulation is the phase representation most relevant to hearing — Distinguishes group delay from phase delay — both terms get clarified, with the auditory-relevance ranking stated.
• Signal cropping is equivalent to applying a rectangular window — Clarifies that simply cropping a signal segment is itself a windowing operation, with specific spectral consequences.
• Short-time Fourier analysis converts signals into spectrograms via windowed transforms — Defines hop size, short-time Fourier analysis, and spectrogram in one compact passage.
• Undersampled signals alias because high-frequency components mirror below the Nyquist frequency — Defines the Nyquist/sampling theorem and aliasing, including the standard low-pass anti-aliasing filter remedy.
• Autocorrelation reveals a signal’s fundamental periodicity through periodic peaks. — Explains autocorrelation as the tool for finding waveform periodicity, with the peak-multiple structure.
• Each added quantization bit yields 6dB SNR improvement — Quantization noise is defined and quantified against bit depth.
• Autocorrelation reveals a signal’s fundamental periodicity through periodic peaks. — Defines autocorrelation as signal-vs-itself comparison for periodicity detection, with the concrete voiced-speech example.
• Standard audio sampling rates range up from 8kHz speech to 192kHz ultra-high fidelity — Defines PCM coding and maps each application domain to its canonical sampling rate.
• Digital signal processing is inherently band-limited by reconstruction filters — Explains why DSP operates only on band-limited signals via the reconstruction filter.
• Linear prediction models signals as all-pole IIR systems predicting samples from prior ones — Explains the core conceptual move of LP modelling: treating a signal as if produced by an all-pole IIR system.
• Feedback (IIR) filters can become unstable, while feedforward (FIR) filters cannot — Clarifies why IIR filters risk instability through feedback poles while FIR filters are inherently stable.
• LP inverse filtering flattens signals into spectrally whitened residuals — Defines synthesis filter, inverse filter, and the whitening operation in one compact causal sketch.
• Air absorption and room acoustics can be modeled as low-pass and arbitrary filters — Defines what a filter does to a signal before the taxonomy of filter types follows.
• Adaptive filtering changes filter coefficients over time in response to signals. — Clear technical definition of what ‘adaptation’ means in the filtering context.
• Speech recognition picks the model maximizing the probability of observed features — Clarifies why the model is called ‘hidden’: the underlying state sequence is never observed, only the emitted feature vectors are.
• The similarity between artificial and biological neural networks is only superficial. — Defines ANNs and notes the deliberate caveat that their resemblance to brain networks is not strong.
• A good microphone needs flat response from 20 Hz to 20 kHz. — Establishes the formal specification of microphone performance (flat frequency response, low distortion/noise, directional pattern).
• Headphones come in intra-aural, supraaural, and circumaural form factors. — Defines the three headphone construction classes (intra-aural, supraaural, circumaural) as standard technical vocabulary.
• A microphone’s sensing principle determines whether its pickup pattern is omnidirectional, dipolar, or cardioid — Clearly connects physical sensitivity (pressure vs particle velocity) to resulting directional patterns: omnidirectional, dipole, and cardioid.
• Electret microphones trap charge in material so no external voltage is needed — Defines the electret microphone as a self-biased condenser microphone.
• Condenser microphones dominate modern audio because they convert sound to electrical signals efficiently. — Explains the operating principle of condenser microphones: moving electrode changes capacitance and thus voltage.
• Room impulse response measurements cannot be made with true perfect impulses from practical sources. — Explains why impulse response measurement is hard: real impulse sources are imperfect.
• Perfect sound reproduction requires an LTI system with unity transfer function — Establishes the formal definition of perfect sound reproduction as LTI with transfer function equal to 1.
• Sound reinforcement permits only milliseconds of delay while recorded reproduction tolerates seconds — Defines the practical delay tolerance distinction between live sound reinforcement and recorded reproduction — the key numbers (10 ms vs. seconds) are the takeaway.
• Ideal audio reproduction permits gain change and pure delay without degrading quality — Gives the relaxed mathematical definition of ideal sound reproduction: any output that is a scaled, time-shifted copy of the input.
• The technically precise term for frequency response is magnitude response — Clarifies that ‘frequency response’ is a loose common name for the strictly correct ‘magnitude response.’
• Enclosures prevent acoustic short-circuiting that silences low frequencies — Explains ‘short-circuiting’ of pressure difference and the Helmholtz-resonator basis of bass-reflex designs.
• Headphone cushion leakage degrades low-frequency response but spares mid and high frequencies. — Introduces the pressure chamber principle that governs headphone-to-eardrum acoustics.
• Electrostatic transduction, the condenser microphone’s inverse, delivers the highest headphone and speaker quality. — Clarifies the electrostatic principle as the inverse of the condenser microphone mechanism.
• The condenser microphone and its electret variant dominate modern microphone technology — Explains the physical operating principle distinguishing condenser microphones (external voltage charge) from electret microphones (trapped charge).
• Perfect sound reproduction tolerates arbitrary signal delay when reproducing recordings. — Defines the practical relaxation of perfect-reproduction: delay is acceptable except in live reinforcement contexts.
• Non-linear and time-variant systems generate frequency components absent from the input — Defines the linear-vs-nonlinear distortion boundary that underlies all audio system characterization.
• Signal processing can compensate for loudspeaker and headphone imperfections. — Defines equalization (EQ) as signal-processing compensation for reproduction-device deviations.
• Total harmonic distortion measures all components a fundamental generates, including non-harmonic ones. — Clarifies THD and subharmonic components, distinguishing harmonic from non-harmonic distortion.
• Equalizers invert measured system impulse responses to flatten magnitude response — Defines parametric equalizer and explains the inverse-filter z-transform approach to flattening an LTI system’s response.
• Voiced speech originates from airflow through the vibrating vocal folds across the glottis — Clarifies larynx anatomy: vocal folds, mucosa, glottis, and their breath vs. speech configuration.
• Vocal tract length differs by sex and its resonances form speech formants. — Defines formants as controllable tract resonances and antiformants as spectral zeros, framing speech as a filter-bank signal-processing problem.
• Phonation types differ by glottal airflow and spectrum shape — The full taxonomy of phonation modes (breathy/modal/pressed plus creaky and whispering) with spectral consequences is clearly and compactly defined here.
• Unvoiced speech arises from turbulence at constrictions or sudden pressure releases — Distinguishes fricatives from plosives by their acoustic generation mechanism — a clear definitional anchor for speech acoustics.
• Speech is a continuous flow that still requires discrete units to be understood — Explains the foundational tension in how speech maps to discrete units, key to understanding acoustic speech coding.
• The International Phonetic Alphabet attempts a universal notation for every speech sound — Defines IPA’s purpose as a universal phonetic transcription system.
• Consonants differ from vowels in intensity, context-dependence, and articulation constriction — Defines consonants by contrast to vowels across three acoustic-articulatory dimensions.
• A full speech communication chain spans brain, vocal organs, channel, and hearing. — The end-to-end source-channel-receiver speech model, including the point that non-linguistic prosody can outweigh words.
• Speech is the original form of linguistic communication, predating writing. — Establishes that human-specific acoustic communication equates to speech, grounding the chapter.
• The International Phonetic Alphabet attempts a universal notation for every speech sound — Defines phonetics and situates IPA as the attempted universal notation.
• Speech segments are classified by abstract classes, concrete sounds, and transition units — Clear technical definitions distinguishing phoneme, phone, allophone, diphone, and triphone units of speech analysis.
• Consonant classes are produced by distinct vocal tract closure configurations — Classifies the six consonant manners of articulation by how the vocal/nasal tract is configured to generate sound.
• Prosody denotes speech features not tied to segmental structure — Introduces prosody/suprasegmental features as a distinct class of speech information.
• Voiced and unvoiced speech differ in whether source is glottal oscillation or frication noise — The source-filter model: the core distinction governing all of speech synthesis and coding.
• The vocal apparatus functions as a pressure source driving a resonant filter system — Establishes the engineering model of speech production: lungs as compressor, vocal tract as filter/encoder.
• Vocal tract acoustics reduce to a cross-sectional area function along one axis — Clarifies the transmission-line framing of the vocal tract central to source-filter speech models.
• Speakers vary pitch to signal emotion, questions, and focus. — Defines intonation as variation of f0 used for emotion, question-marking, focus, and conversational regulation.
• Plosive consonants begin with silence then a burst when the tract opens. — Shows how brief silences function as speech building blocks, relevant to the reader’s silence-and-sound interests.
• Speech production reduces to lungs as compressor driving a glottal oscillator through resonating tracts — The engineering analogy — pressure compressor, mass-spring oscillator, resonator/filter — is the standard mental model for how speech is made.
• The vocal tract can be simulated as a cascade of acoustic transmission line segments — Introduces the Kelly-Lochbaum model, a core computational method in speech synthesis.
• Overtone singing selects harmonics with sharp vocal-tract resonances to create melodies — Defines overtone/throat singing’s acoustic mechanism: fixed vocal-fold pitch plus movable tract resonances picking out harmonics.
• A unified source-filter model describes human and synthetic voice alike — States the chapter’s governing idea: one modeling framework spans biological voice, acoustic sound sources, and electronic/devised speech production.
• Music and speech are utility sounds distinguished by their primary role — This frames what distinguishes music from speech in acoustic-communication terms: aesthetic/emotional evocation versus linguistic content.
• Instrument sounds are endlessly varied yet fall into families sharing generation principles — Establishes the family-of-instruments concept where shared sound-generation principles group countless individual sounds.
• Aerophone design tunes resonance modes so the instrument becomes musically usable — Defines aerophones and how their air-column resonances are shaped by mouthpiece and bell.
• Plucked strings vibrate autonomously after excitation, simplifying their modelling. — Explains why plucked-string synthesis is computationally simple: post-excitation vibration never feeds back into the excitation source.
• Hearing is highly sensitive to temporal changes in sound — Defines the amplitude envelope and spectrogram as the two standard temporal-display methods for sound.
• Musical instrument sounds divide into attack, decay, sustain, and release phases — Defines ADSR-sequence, the standard time-envelope model for synthesizing instrument sounds.
• Bowed string motion is a stick-slip effect producing a sawtooth wave — Describes the stick-slip mechanism clearly enough to serve as a reference definition for bow-string coupling.
• Continuous-excitation instruments like flutes attack slowly while impulse-excited ones do not — Clarifies a perceptual signature distinguishing continuous-excitation from impulsive instruments.
• Computational models can synthesize natural ambient sounds like footsteps and thunder — Defines ‘natural sources’ as a synthesis category distinct from music and speech.
• Hearing is a staged chain from air capture through mechanical conduction to neural processing — Defines the three-part functional division of the peripheral auditory system and the central pathway.
• The human auditory system evolved specifically for spoken-language analysis — Establishes what hearing is for, distinguishing the peripheral capture from central analysis.
• The ear canal acts as a quarter-wavelength resonator amplifying 3-4 kHz by 10 dB — Classic quarter-wavelength resonator example: open tube with higher impedance at the far end, ~22.5 mm tuning 3-4 kHz boosts.
• Cochlear pitch mapping exploits the mass of vibrating liquid varying with route length — Explains the physical mechanism (liquid mass as mass-spring system) behind frequency-to-place coding.
• Hair cells actively amplify cochlear vibration rather than only sensing it — Defines the cochlear amplifier and the concept of an active cochlea.
• Cubic difference tones reveal nonlinear cochlear mechanics even at low sound levels — Defines the cubic difference tone formula fcubic = 2f1-f2 and why its low-excitation generation is anomalous.
• Head shading creates interaural level differences used for sound localization — Establishes the physical basis of interaural time delay and head-shadow level differences, the two classic binaural cues.
• Loud sounds spread cochlear activation broadly and respond faster than soft sounds — Active amplification is the cochlea mechanism that elevates low-level responses and takes time to build up.
• A competing tone at another frequency suppresses a cochlea channel’s response — Two-tone suppression is defined as one tone suppressing activation at another tone’s characteristic frequency.
• The auditory periphery is non-linear and cannot be captured by one transfer function. — Defines firing rate and explains why hearing cannot be modeled as an LTI system.
• Auditory nerve phase locking degrades above 1500 Hz and is weak above 4 kHz — Clearly defines phase locking as neurons syncing firing to a periodic signal’s phase.
• The cochlear nucleus functions like a preamplifier and initial spectral analyzer — Frames auditory brainstem processing in engineering terms, mapping biological stages to signal-processing functions.
• The superior colliculus is the only mammalian brain nucleus with a topographical auditory map — Clarifies what the superior colliculus is and does: integrates auditory and visual spatial maps for orienting head and gaze.
• Auditory models link acoustic stimulus features to perceived attributes computationally. — Defines what computational modeling of auditory perception aims to do — map ear canal signals to what subjects report hearing.
• Auditory nerve phase locking degrades above 1500 Hz and is weak above 4 kHz — Defines phase locking as synchronized firing to a stimulus phase, with its frequency ceiling in humans.
• The cochlear nucleus functions like a preamplifier and initial spectral analyzer — Provides the engineering-style functional decomposition of the first auditory-nerve relay stage.
• Computational auditory models serve knowledge, audio-quality measurement, and diagnostics — Clear account of the three distinct application purposes computational auditory models serve, distinguishing them from one another.
• Psychoacoustics developed largely through communication technology challenges — Establishes psychoacoustics as the study of systemic and informational properties of hearing, independent of but correlated with physiology.
• Auditory experimentation resembles measuring an unreliable device against an unpredictable system — Frames the core methodological difficulty of psychoacoustics: a one-line characterization of why auditory experiments need careful controls.
• Psychophysical functions map continuous physical sound scales onto varied perceptual scales. — Defines psychophysical function and distinguishes the psychometric function as the yes/no probability variant; foundational for designing perception experiments.
• Auditory perception must be studied indirectly through measurable external reactions. — Introduces the reaction/description function concept distinguishing internal auditory events from externally observable behavior.
• Perceived loudness depends on frequency content and duration, not just sound pressure level — Clarifies the psychophysical distinction between a sound event’s physical properties and the resulting auditory attributes.
• Perceived-loudness matching often needs separate listening tests to achieve — Distinguishes physical equal-loudness from perceptual equalization — a key methodological distinction.
• Perceptual thresholds blur because listener internal noise varies by personality — Introduces the psychoacoustic threshold as a probabilistic region, not a hard cutoff.
• Trained listeners improve test sensitivity but misrepresent general audience opinion — Distinguishes analytic trained-listener panels from population-representative panels for sound quality research.
• Mean Opinion Score scales standardize audio-quality evaluation on a 1-to-5 range — Defines MOS scale and its use in evaluating audio quality.
• Test method means the logic selecting sound attributes across successive tasks — Precisely distinguishes ‘method’ (selection logic) from the procedural presentation of tasks.
• Constant-stims method maps full psychophysical functions rather than single thresholds — Explains the method of constant stimuli, the gold-standard threshold technique.
• Forced-choice guessing floors push threshold criterion above 50 percent detection — Explains why threshold criteria sit at 71-75 percent rather than 50 percent in forced-choice tasks.
• Psychophysical scales are built by adjusting sounds to match reference ratios — Clarifies magnitude production and ratio production as technique variants for deriving relative and absolute psychophysical scales.
• Tracking audiometry maps hearing thresholds but ignores response bias — Defines the tracking method and its flagship example, Békésy audiometry.
• Introspection alone cannot serve as accepted scientific auditory observation — Defines introspection in the psychoacoustic context and explains why it fails as a method.
• Psychophysical functions map continuous physical sound scales onto varied perceptual scales. — Distinguishes psychophysical functions from psychometric functions, which map continuous scale to a yes/no detection probability.
• Paired comparison against anchors reduces bias in absolute judgments — Introduces the anchor/comparison paradigm and the MUSHRA test name.
• Thresholds and auditory scales link physical sound attributes to perceived auditory events — Distinguishes sound event (physical) from auditory event (perceptual), the field’s central definitional pair.
• Audio codec testing requires methods that reveal unknown perceptual attributes — Introduces descriptive sensory analysis as a method family for mapping unknown perceptual attribute palettes.
• Descriptive sensory analysis maps perceived sound differences onto an N-dimensional attribute space — Clarifies descriptive sensory analysis as projecting sounds into an N-dimensional space of elicited attributes — a methodologically exact framing useful later.
• Verbal elicitation dominates sensory science via consensus or individual vocabularies — Distinguishes consensus-vocabulary from individual-vocabulary descriptor methods — the two branches of verbal elicitation.
• Assessors can use their own vocabulary instead of agreeing on descriptors — Clarifies what free-choice profiling is and why it saves training effort.
• Multidimensional scaling maps perceived sound dissimilarities without labeling the perceptual dimensions — Explains indirect elicitation and multidimensional scaling as methods that derive perceptual attributes without direct subject labeling.
• Perceptual structure analysis indirectly recovers individual feature-usage in auditory perception. — Defines perceptual structure analysis as separating feature identification from feature labelling.
• Careful listening tests validly reveal how acoustic attributes become auditory attributes. — Defines descriptive sensory analysis and the psychophysical transfer function concept central to communication acoustics.
• Analysis of variance tests whether any independent variable affects a dependent variable. — Defines ANOVA’s core question and the role of posthoc tests in psychoacoustic experiment analysis.
• Subjective listening tests map manipulated acoustic variables to measured responses. — Clarifies how independent/dependent variable terminology maps onto stimulus manipulation and subject response in listening tests.
• Sound attributes are independent variables manipulated to elicit dependent responses. — Defines how acoustic attributes map onto the independent/dependent variable vocabulary of experimental design.
• Sound audibility depends on both its level and its frequency — Establishes the two physical axes delimiting the existence of an auditory event: SPL and frequency, with the pain threshold as upper limit.
• The phon scale defines loudness relative to equal-loudness matches at 1 kHz — Definition of the phon unit anchors loudness measurement.
• Hearing’s lower limit is the minimum audible field; upper is pain — Defines MAF and threshold of pain as the boundaries of auditory working range.
• One sound can mask another even when their spectra do not overlap. — Clarifies that masking does not require spectral overlap — a non-obvious mechanism underpinning codec design.
• A-weighted decibels are a rough practical approximation of perceived loudness — Defines the A-weighted sound level curve family and its practical rationale.
• The hearing threshold is measured from sounds reaching the subject’s binaural position. — Defines the hearing threshold and minimum audible field (MAF) calibration method.
• Complex real-world maskers are less effective than simple energetic ones predict — Introduces the energetic-vs-informational masking distinction, a core concept.
• Hearing groups frequency bands by shared envelope modulation, reducing masking — Defines co-modulation masking release as a named phenomenon before explaining its mechanism.
• Tone-on-tone masking can be detected through beating and combination tones alone — Explains what beats are in context.
• Coherent amplitude modulation across frequency bands may drive auditory source grouping. — Introduces the modulation filter bank concept — a proposed mechanism for how the ear groups signal components into sources.
• Sounds mask both preceding and following sounds but forward masking dominates. — Defines temporal — backward vs forward — masking, extending the classical simultaneous-masking picture.
• Each inner hair cell responds best to one frequency plus a widening critical band — Defines the critical band / auditory filter concept and traces it to cochlear hair-cell tuning.
• Hearing segregates sounds by frequency unless they fall within a critical band — Defines frequency resolution — the ability of hearing to segregate sounds separated in frequency.
• Loudness stays constant below critical bandwidth then grows beyond it — This operationally defines the critical band: the bandwidth below which loudness is constant regardless of spectral spread.
• Loudness stays constant below critical bandwidth then grows beyond it — Concretely defines the critical-band concept operationally as a 160 Hz knee at 1 kHz, which is the load-bearing empirical anchor for the whole critical-band framework.
• Wider bandwidth increases perceived loudness at equal sound pressure — Defines critical bandwidth operationally: the bandwidth point above which perceived loudness begins to increase.
• Auditory bands from the same source are summed to compute total loudness — Explains the mechanism by which broadband signals seem louder: same-source auditory bands are integrated and summed.
• Notched-masker tests prevent off-frequency listening to measure filter width — Defines ERB bands and off-frequency listening, and the notched-noise method used to measure auditory filter bandwidths.
• The Bark scale is linear below 500 Hz and logarithmic above it. — Defines the Bark scale’s specific linear-to-logarithmic behavior, which is core acoustic perception vocabulary.
• Pitch is not uniquely tied to a single physical frequency — Gives the ANSI standard definition of pitch and clarifies its loose relation to physical repetition frequency.
• Pitch is the perceptual property ordering sounds from low to high — ANSI’s formal definition of pitch, with its caveat about ambiguous-pitch sounds.
• Summing a signal with its delayed copy creates perceived pitch. — Defines repetition pitch and connects it to the comb-filter effect — a clear conceptual framework.
• Amplitude modulation of noise creates pitch with a flat spectrum — Introduces SAM noise and the striking fact that pitch can arise from temporal (not spectral) structure.
• Humans perceive a fundamental frequency even when it is absent from the signal — Defines virtual pitch and missing fundamental perception.
• Place and timing theories fail to explain some complex pitch perception phenomena — Introduces the place theory of pitch perception and demonstrates its limits.
• Place and timing theories fail to explain some complex pitch perception phenomena — Distinguishes the two classical theories of pitch perception and their current combined status.
• Perceived loudness can be measured by iterative ratio-matching across subjects — Explains the experimental methodology (iterative doubling/halving with averaging) behind the loudness psychophysical function.
• The phon defines loudness level equal to dB SPL at 1 kHz — Defines the phon and how equal loudness curves anchor it at 1 kHz.
• The mel scale was derived by asking listeners to double perceived pitch — Defines the mel scale, the perceptual pitch scale underlying mel-frequency cepstral coefficients used widely in speech recognition.
• Listeners perceive the fundamental frequency even when its lowest partials are absent — Defines the missing-fundamental phenomenon and introduces synonyms virtual/residue pitch.
• Loudness is perceptual while intensity is physical and must be distinguished — Establishes the ANSI operational definition of loudness as a perceptual scale, distinct from physical SPL.
• Frequency spread in the cochlea is asymmetric with steeper low-frequency slopes — Quantifies the auditory spreading function: steep low-frequency slope (20 dB/ERB) versus gentler high-frequency slope (8 dB/ERB), anchored to basilar membrane motion.
• Perceived loudness of concurrent tones hinges on frequency separation via three distinct mechanisms — Explains how beating-driven amplitude summing, power summing, and critical-band separation each dominate loudness perception in a different frequency regime.
• The basilar membrane’s fixed-spread frequency resolution shrinks relative bandwidth as frequency rises — Defines the spreading function and its physiological correlate on the basilar membrane.
• Hearing forms loudness by integrating sound energy over about 100-200 ms — Names and explains the temporal integration mechanism of loudness perception.
• Time-varying sound loudness is set by integrated loudness peaks, not moment-to-moment level — Clarifies how ‘loudness peak’ and short-time integration jointly account for percepts of dynamically changing sounds.
• Timbre distinguishes sounds sharing pitch, loudness, and duration. — Classic definition of timbre in terms of what remains after pitch, loudness, and duration are matched.
• Perceived sharpness is stable across subjects and rises with spectral centre of gravity — Defines the psychoacoustic quantity ‘sharpness’ as a measurable, subject-independent perceptual dimension.
• Hearing repackages one-dimensional sound signals into multidimensional time-frequency objects — Introduces the conceptual shift from signals to auditory streams and object perception that frames the whole chapter.
• Perceived sharpness is stable across subjects and rises with spectral centre of gravity — Defines the psychoacoustic quantity sharpness and grounds it in a repeatable, subject-independent measurement procedure.
• Fluctuation strength is standardized against a 4 Hz amplitude-modulated reference tone — Defines the vacil, the measurement unit for fluctuation strength.
• Slow sound-level changes are heard as fluctuation and fast ones as roughness — Defines fluctuation strength, roughness, and impulsiveness as distinct psychoacoustic quantities tied to modulation rate.
• Arbitrary-sound roughness is estimable by summing per-critical-band contributions — The summing-over-critical-bands approach is a practical modeling method for engineering sound evaluation.
• Fast amplitude modulations within a critical band produce roughness perception — Defines roughness as a psychoacoustic quantity caused by 15-300 Hz modulations inside a critical band.
• Auditory perception runs parallel bottom-up and top-down processes — Two-stream model of auditory perception: automatic bottom-up ordering plus selective top-down attention.
• The brain organizes simultaneous sound events into separate auditory streams via auditory scene analysis. — Introduces auditory streaming and auditory scene analysis, core concepts for understanding how listeners parse complex acoustic scenes.
• Psychoacoustic tonality measures tone-character, distinct from musical tonality terms — Gives the crucial disambiguation between psychoacoustic ‘tonality’ (quantity of tone-likeness) and the unrelated musical terms tonal/atonal/microtonal — a persistent terminological confusion worth nailing down.
• Auditory cortex specifically attenuates responses to frequently repeated sounds — Defines stimulus-specific adaptation: decreased neural response to a frequent standard stimulus at the auditory cortex.
• Hearing perceives synchronized neural firing across critical bands as buzzy timbre — Explains the mechanism by which phase coherence across critical bands produces a ‘buzzy’ perceptual character.
• Synchronized phase harmonics produce a ‘buzzy’ sharpness perceptible without room reflections — Introduces the technical term ‘buzzyness’ for the percept of phase-aligned harmonic partials.
• Group delay affects perceived quality only when it varies with frequency — Defines group delay and the important distinction between constant delay (harmless one-way) and frequency-dependent delay (audible degradation).
• Interval dissonance increases with position higher in the harmonic series — Explains the mechanism by which harmonic-partial interference, not simply the fundamental ratio, produces perceived dissonance — the perceptual basis for why 3:2 is consonant but 5:4 less so.
• Peak dissonance between two tones occurs at a quarter critical band separation — Defines the quantitative relationship between critical band spacing and perceived consonance/dissonance, a core perceptual-acoustics anchor.
• Two close sine tones produce amplitude beats at their frequency difference — Defines ‘beating’ as the amplitude fluctuation arising from two nearby-frequency tones.
• Auditory perception runs parallel bottom-up and top-down processes — Defines the two parallel perceptual mechanisms (bottom-up primitive ordering vs. top-down selective attention) that structure how listeners parse sound.
• Perception is a brain interpretation that need not use all sensory detail — Clarifies the sensation-versus-perception distinction, a foundational definition for acoustic communication.
• Auditory cortex adaptation decreases response to frequent sounds, shaping perception — Defines stimulus-specific adaptation (SSA), the cortical decrease in response to standard repeated stimuli - a key concept distinguishing peripheral from central auditory adaptation.
• Western major scales consist of five whole tones and two semitones in fixed arrangement — Explains the structural intervals and construction of the diatonic scale.
• The outer ear’s resonant cavities enable elevation perception where other cues fail — Explains the mechanism by which pinna anatomy encodes sound-source elevation, a core definitional concept in spatial hearing.
• Head-related transfer functions are measured with in-ear microphones in anechoic chambers — Explains the standard measurement procedure for HRTFs, the data underlying binaural and spatial audio.
• HRTF definition normalizes ear responses by the head’s absence — Mathematical definition of the transfer function, essential for anyone building spatial audio systems.
• A dedicated cone-of-confusion coordinate system describes hard-to-discriminate sound directions. — Explains the cone of confusion and the coordinate system built around it.
• Spatial hearing is described in three distinct coordinate systems with different strengths. — Clearly explains the three coordinate systems (Cartesian, spherical, cone-of-confusion) used to describe spatial hearing.
• The ear sits 8 cm off-center, so source-ear distance changes with azimuth — Explains why HRIR arrival time shifts with source direction: geometric offset of the ear from the head’s rotational center.
• Shoulder reflections create a distinguishable 1-ms delay in head-related impulse responses — Explains how shoulder reflection produces a measurable, systematic delay in HRIRs — a concrete mechanism connecting anatomy to acoustic signal structure.
• Auditory spatial perception emerges from weighted combination of localization cues — Defines ‘localization cues’ as stimulus properties relevant to spatial perception whose weighted combination forms spatial attributes of auditory events.
• Interaural level differences and time shifts trade off to control lateralization — Explains how ITD relates to perceived lateral position across regimes, including breakdown.
• The head shadows and reflects sound to create level differences between the ears. — Defines ILD as a scattering-based binaural localization cue.
• Cone of confusion makes time and level cues ambiguous for elevation — Explains the cone-of-confusion problem motivating spectral and dynamic localization cues.
• Headphone audio internalizes because static binaural cues signal an inside-the-head source — Explains the mechanism of headphone sound externalization failure via the brain’s inference rule that constant-under-head-motion means inside the head.
• Separate sensory modalities fuse into one coherent perception unless cues conflict — Names and defines multimodal perception and the fusion of independent sensory partial percepts.
• The precedence effect suppresses spatial identity of reflections, not their audibility — Defines the precedence effect as the mechanism suppressing early reflections in direction perception.
• Speech elements near masked threshold carry intelligibility only under binaural listening. — Defines BILD (binaural intelligibility level difference) as the measurement metric for the binaural advantage.
• The ear achieves spatial hearing through signal analysis rather than directional receptors — Distinguishes the eye’s directional-receptor architecture from the ear’s single-cochlea analysis architecture for spatial separation.
• Sound localization fuses independent sensory evidence into one coherent percept — Defines multimodal perception and the fusion of partial percepts — a framework concept for how spatial perception actually works.
• Perceived auditory distance has an acoustic horizon under 10 meters — Introduces the useful term ‘acoustic horizon’ for the finite perceived-distance ceiling in free field.
• The auditory system cancels room coloration to recover source signals — Defines the term ‘binaural decolouration’ — the mechanism by which listeners ignore room spectral contributions in ear canal signals.
• Listeners perceptually subtract room coloration from source signals, called binaural decolouration. — Clear technical definition of binaural decolouration plus its hypothesized mechanism (central binaural spectrum summation masking room-response dips).
• Exaggerated interaural level differences signal that a sound source is near — Clarifies how ITD and ILD diverge as proximity cues — the head shadow produces distance-dependent ILD for nearby sources.
• Hearing acts as a spectral analyser extracting quantifiable perceptual metrics from sound — Sets out how the peripheral hearing system is modelled as a spectral analyser, a foundational concept for understanding what auditory models actually compute.
• DFT-based psychoacoustic models derive an auditory excitation spectrum on the ERB scale — Introduces the ‘auditory spectrum’ — cochlear excitation level dB vs. frequency on the ERB scale — a core operational definition for this chapter’s modelling approach.
• Psychoacoustic models explain experimental results without addressing physiology — Distinguishes auditory model from psychoacoustic and perceptual models — useful terminology for reading later chapters on applications.
• Frequency masking is modelled as convolution with a spreading function — Clarifies that excitation-pattern spreading due to masking is computationally implemented as a convolution with a spreading function approximating the masking curve.
• Filter bank models outperform DFT in mimicking auditory temporal resolution — Clarifies why filter bank models are preferred over DFT-based approaches for auditory modeling.
• Gammatone filter banks model human auditory filtering using measured nerve responses — Explains what a gammatone filter is and why its shape results from measured auditory nerve impulse responses.
• Pitch perception can be modeled by autocorrelation across auditory filter bands — Defines the correlogram and summary autocorrelation, concrete terms in pitch modeling.
• Hair-cell adaptation to sound arises from neurotransmitter pool depletion and recycling dynamics — Defines the Meddis hair-cell model: transmitter pool, cleft, reprocessing store, and factory, with loss-rate l providing adaptation.
• Pitch perception is encoded by the timing of neural firing in the cochlea — Introduces the cochleogram, correlogram, and summary autocorrelation function as the processing stages of the auditory model.
• Auditory nerve adaptation operates like automatic gain control. — Clarifies that neural adaptation models share an AGC functional basis.
• Loudness models need 100-200 ms time constants to capture temporal integration and post-masking. — Specifies the concrete time constants (100–200 ms) used to emulate temporal integration and post-masking in auditory models.
• Neurotransmitter loss in the synaptic cleft produces auditory adaptation. — Explainable mechanism: adaptation to sustained sound arises from transmitter depletion rather than neural fatigue.
• Jeffress coincidence detection explains how neurons encode interaural time differences. — Explains the Jeffress place-code model — how delay lines make each neuron respond maximally to one specific ITD.
• Interaural time differences are computed by cross-correlating the two ear signals — Explains how IACC computation emulates the Jeffress coincidence-detection model to extract ITD cues.
• The IACC maximum value estimates interaural coherence between the two ear signals — Defines interaural coherence as derived from the normalized IACC maximum.
• Jeffress coincidence detection explains how neurons encode interaural time differences. — Clear mechanistic explanation of the Jeffress coincidence-detection model underlying most binaural algorithms.
• High-frequency auditory analysis preserves only the temporal envelope of signals — Explains what IACC (interaural cross-correlation) functions represent and how they degrade at high frequencies.
• Equalization-cancellation subtracts masker signals to reveal hidden targets — Explains the equalization-cancellation model (Durlach 1963) for detecting signals in masking noise via subtraction after equalizing ITD/ILD.
• Two brain models differ on whether localization uses narrow or wide receptive fields — Distills the core theoretical contrast between Jeffress-type coincidence detection and count-comparison models of binaural localization.
• Binaural unmasking can be modelled as equalizing then subtracting two ear signals — Clear lay-accessible account of what binaural unmasking is and the EC model’s mechanism.
• Count-comparison and coincidence-detection models imply different brain receptive-field architectures for direction coding. — Distinguishes count-comparison (two broad hemifield receptive fields, left/right coordinate output) from Jeffress coincidence-detection (narrowly tuned directional neurons) model architectures.
• Sound reproduction’s unifying trait is its endpoint in the human listener — Defines sound reproduction and marks the human-listener endpoint as the field’s distinctive property.
• Authentic sound reproduction is judged by perceptual identity, not signal fidelity — Defines authenticity in reproduction as perceptual equivalence, not waveform fidelity.
• Audio engineering is the production of meaningful sound signals across media. — Defines the content/engineering split used throughout the book.
• The Blumlein XY pair captures reverberance fully but causes phase artefacts from side sound. — Defines the Blumlein pair XY microphone arrangement and its side-lobe phase reversal quirk.
• The ORTF technique blends coincident and spaced microphone characteristics — A clear technical definition of the ORTF microphone technique and how spacing/angle trade off time vs. amplitude differences.
• Spherical harmonics act as basis functions for building any microphone pattern — Explains B-format spherical harmonics and the additive property of directional patterns as basis functions.
• First-order Ambisonics reproduction is restricted to three loudspeakers — The 2N + 1 loudspeaker limit for a given microphone order N is a crisp quantitative rule worth remembering.
• DirAC divides captured sound into diffuse and non-diffuse streams — Explains the diffuseness parameter (0-1) and the diffuse/non-diffuse stream split that implements the DirAC assumption.
• Amplitude panning creates auditory objects with no real physical source — Defines virtual source positioning as perceived-direction control via monophonic reproduction.
• Three-dimensional VBAP extends pair panning to position sources on loudspeaker triangles — Defines 3D VBAP as a triangular loudspeaker-vector-base extension of two-dimensional pairwise panning.
• Audio production divides into recording, mixing, mastering, and live sound stages — Defines the production chain of recording, mixing, mastering, live sound, studio, and audio format in one coherent framework.
• Object-based audio decouples soundtracks from specific loudspeaker layouts — Defines object-based audio: sound sent as objects with rendering metadata rather than fixed channels.
• Mono reproduction exaggerates recording-room coloration that stereo cancels — Explains the mechanism of binaural decolouration and why it fails to compensate the recording room in mono reproduction — a clear technical definition of the phenomenon.
• Directional coincident microphones suppress room reverberation, killing sense of space — Explains mechanically why coherent/coincident techniques sacrifice spatial ambience.
• Instrument directionality shapes the captured spectrum depending on mic placement — Introduces the concept of ambience microphones that capture room response, complementing close mics.
• Close-miked mono reproduction stays clear because the room barely affects the signal — Defines when room acoustics matter: only when direct sound level is comparable to or below reverberant field level.
• Spot miking captures each instrument with its own near microphone for independent channels — Defines the spot/close-miking technique and its goal of per-source channel independence.
• Inverted-polarity (antiphasic) stereo signals cancel lows and smear image — Defines the engineer’s term ‘phasy’ and gives the underlying mechanism of phase cancellation in antiphasic reproduction.
• Binaural audio reproduces ear canal signals from real or virtual scenarios. — Gives a working definition of binaural technique centered on reproducing exact ear-canal signals.
• Headphone binaural reproduction fails without dynamic head-movement cues — Clarifies why headphone playback yields inside-the-head localization: dynamic cues absent.
• Head-tracked HRTF filtering restores realistic external sound localization — Defines head tracking and its sampling rate as a mechanism for virtual audio realism.
• Compression, limiting, and expansion attenuate or boost audio differently by amplitude. — Clear three-way distinction between compressor, limiter, and expander behavior.
• Microphone directivity shapes how reverberant a convolved signal sounds — Clarifies why impulse-response-based convolution captures a specific mic/source pair rather than a neutral room.
• FFT block convolution trades speed for double-block latency in reverbs — Explains the tradeoff behind FFT-based convolution reverberation, a core DSP technique.
• Nonlinear spatial analysis distributes measured impulse responses across loudspeaker arrangements — Defines nonlinear impulse-response reproduction: instantaneous spatial analysis spreads the measured response to each channel.
• Shoulder reflections imprint identifiable arcs on measured head-related impulse responses. — Explains how to read HRIR arc structure as causal traces of shoulder reflections.
• Auditory spatial perception emerges from weighted combination of localization cues — Defines the concept of localization cues and how they combine into spatial perception.
• Sound source location is encoded in waveform distortions between the ears — Defines binaural vs monaural localization cues and introduces ITD/ILD as the two primary ones — a foundational vocabulary passage for the reader’s applications focus.
• Stereo signals can be processed for headphones to recreate natural binaural sound fields — Defines HRTFs as free-field transfer functions from a source to both ears, and introduces BRIRs as the binaural counterpart.
• Frequency-domain processing enables efficient audio coding for monophonic and multichannel signals — Establishes why time-frequency transforms matter and the aliasing constraint that shapes transform choice.
• Time-frequency modulation and amplitude modulation are distinct operations — Draws a sharp line between two meanings of ‘modulation’ - a recurring terminological trap in audio DSP.
• Non-redundant time-frequency transforms introduce aliasing artifacts when signals are modified in the frequency domain. — Defines non-redundant vs redundant transforms and links redundancy to aliasing avoidance.
• MDCT excels for coding but lacks robustness to time-frequency domain edits — Key distinction between a coding transform (MDCT, in MP3/AAC) and a processing transform (STFT).
• Each STFT sample represents a circularly continuous tone — Explains the circular/tone interpretation central to understanding STFT bin meaning.
• MDCT exploits inter-frame aliasing cancellation to achieve critical sampling — Explains MDCT’s core mechanism: half-overlap frames whose odd/even symmetry generates aliasing that the next frame cancels.
• Decorrelation enables creating independent channels while preserving perceptual quality. — Defines decorrelation and ties it to the concrete upmixing/surround use case.
• Spatial audio is synthesized by controlling band energies and inter-channel interdependencies. — Frames the core parameters (energy, cross-term) underpinning parametric spatial-audio coding.
• Decorrelators degrade audio quality on transient-rich signals like applause. — Defines what decorrelators do (alter time/phase structure per frequency band) and precisely why they fail on applause-like content.
• Head and pinna scattering imprint direction-dependent spectral notches onto the ear signal. — Defines spectral cues as the direction-dependent magnitude spectrum alteration caused by head and pinna scattering.
• Bandwidth extension and error concealment are speech-enhancement techniques for missing or narrow data. — Precise definitions of two applied speech-enhancement subfields that a practitioner would need to distinguish.
• Speech coding is fundamentally a quality-versus-bitrate optimization tradeoff. — Defines the core framing of speech coding as bitrate-quality optimization, useful as a conceptual anchor for the whole chapter.
• The 1930s vocoder pioneered reducing speech to slowly-varying parameters to cut bandwidth — Explains how the vocoder works mechanically and why it saved bandwidth, defining the term historically.
• Wideband speech coding spans 50 Hz to 7000 Hz for brighter fuller quality — Clarifies the technical definition of wideband versus superwideband and fullband in speech coding.
• Modern synthetic voices learn from single-speaker speech databases rather than hand-crafted phonetic rules. — Defines data-based synthesis as distinct from knowledge-based synthesis, establishing the core terminological split of the chapter.
• Data-based speech synthesis splits into sample-reassembly and parameter-model approaches. — Clarifies the two distinct technical approaches that both fall under data-based synthesis.
• Speech synthesis selects units by minimizing target and concatenation costs. — Defines the two cost metrics (target and concatenation cost) that drive unit-selection synthesis.
• Language knowledge beyond acoustics is required for automatic speech recognition. — Explains why grammar and vocabulary knowledge must be layered atop HMM phone models.
• Synthesis and recognition demand language knowledge databases that coding does not. — Cleanly draws the conceptual line between coding (no language knowledge) versus synthesis/recognition (language knowledge required).
• Head movements resolve front-back ambiguity by creating direction-specific binaural cue changes — Defines the dynamic-cue mechanism for resolving the cone of confusion, including the key implicit static-source assumption.
• Sound quality means judged preferability, not category membership — Clarifies the two senses of ‘quality’ and which one the book uses.
• Speech intelligibility measures the proportion of correctly identified transmitted messages — Defines subjective and objective speech intelligibility measurement in clear, concrete terms.
• Squaring the pressure signal makes noise and reverberation measurable on a common scale — Distinguishes intensity (net energy flow) from pressure-squared energy in STI literature.
• Sound quality cannot be assessed except through psychophysical experiments with human subjects — Establishes sound quality as fundamentally subjective and defines how it is studied.
• Comparative MOS ratings allow a test sound to score above the reference. — Introduces the comparative MOS (CMOS) scale from -3 to 3 for ranking against a reference.
• Quality and impairment are measured with distinct MOS variants. — Distinguishes MOS (direct quality) from DMOS (degradation between reference and test).
• PAQM estimates audible quality degradation by comparing specific loudness across critical bands. — Explains what PAQM is and how it derives from critical-band specific loudness.
• Auditory-model quality metrics fail on acoustically captured signals — Clarifies what PEAQ can and cannot do, demarcating its limited area of applicability.
• Binaural intelligibility level difference measures speech unmasking benefit — Defines the BILD metric used to quantify how much two-ear listening improves speech intelligibility in noise.
• Unlike the eye, the ear has no primary directional sensitivity — Clarifies the structural basis of why spatial hearing must be computed from canal signals rather than directly sensed.
• Speech intelligibility estimates compress 98 MTF values into one index — Documents how the full MTF matrix gets reduced to the single-number STI scвход value that practitioners actually use.
• MTF measurement uses amplitude-modulated noise bands across 0.63 to 12.5 Hz — Spells out the concrete operational definition of MTF measurement as an emissions-and-probe procedure.
• STI reliability is judged against its 0.15 category step size — Specifies the quality-category granularity used to judge measurement repeatability.
• Sound perception is shaped by the listener’s expectations and context of use — Distinguishes product sound quality (context-dependent interpretation) from raw auditory attributes
• All sound quality judgments derive from properties of the human auditory system. — Defines ‘product sound quality’ as the unifying thread across all sound quality research and grounds it in auditory-system physiology.
• PESQ works as a signal-difference device rather than a true auditory model — Clarifies the key distinction between objective signal-difference measurement and genuine perceptual/neural modeling — the engineering pragmatism behind PESQ.
• Speech quality can be predicted algorithmically using signals that mimic human auditory processing. — Explains how 3QUEST predicts subjective speech quality using a pipeline mimicking human time-frequency resolution plus a trained pattern-recognition model.
• EQUEST injects temporal masking knowledge into echo annoyance estimation — Introduces EQUEST as a specific instrumental method for echo quality evaluation.
• Concert hall quality is codified into standardized subjective attribute vocabularies — Introduces Beranek’s 18-attribute palette and Lokki’s sensory analysis as the vocabularies for describing hall acoustics.
• Annoyance and disturbance refer to distinct degrees of noise harm — Distinguishes the closely related technical terms annoyance and disturbance in sound-quality research.
• Any sound can function as noise depending on the listener — Defines noise as a subjective listener-dependent category, not a physical property of sound.
• Every sounding product is continuously evaluated for sound quality — A useful working definition of product sound quality as goal-relative adequacy.
• Excitation spread across frequency bands is modelled as convolution with a level-dependent spreading function. — Explains what ‘spreading function’ means computationally — convolution emulating how excitation leaks across frequency bands.
• CASA aims to extract all perceivable information from ear canal signals. — Defines CASA and positions music information retrieval as a subtopic of it.
• Interactive rendering differs from reproduction because user actions alter the sound. — Distinguishes virtual-reality audio rendering from conventional sound reproduction by the presence of user interaction.
• Interactive game audio requires dynamic rendering systems beyond traditional linear production. — Defines dynamic/adaptive audio vs. linear audio as the core distinction of game and VR sound production.
• Virtual audio engines render sounds based on avatar position, not just a fixed mix. — Defines how audio is situated in virtual worlds — audio follows an avatar’s location and orientation the way visuals follow the field of vision.
• Spatially broad sound sources should alter their perceived extent with listener distance — Defines the design requirement of ‘spatial extent’ of a virtual source — a concept central to believable audio VR.
• Sound design in interfaces is an extension of product sound quality — Defines sonic interaction design as a subfield of communication acoustics.
• Computational auditory scene analysis targets human-level performance without imitating the ear — Defines CASA precisely and clarifies the crucial point that it targets human-level performance, not imitation of the human auditory system.
• Music information retrieval serves both casual listeners and music professionals needing structured database queries — Defines music information retrieval (MIR) and who its stakeholders are.
• Blind source separation becomes very difficult when sources exceed microphones — Defines blind source separation and ICA and their key constraint condition.
• Auditory displays deliver computer information to humans through sonification and earcons — Defines auditory displays, earcons, and auditory icons as a coherent terminology set.
• Speech intelligibility is the standard criterion for hearing function — Defines the field’s primary functional metric: speech intelligibility as the criterion for hearing functionality, with music nuance as a secondary casualty of mild loss.
• The 0 dB HL audiometric zero varies across frequency by design — Clarifies that 0 dB HL is frequency-dependent by construction, a subtle and often-misunderstood point about audiometric reference levels.
• Sensation level expresses tone level relative to one individual’s own threshold — Distinguishes SL from HL/SPL reference frames, a foundational distinction for any applied audiometric work.
• Outer hair cell damage compresses dynamic range, causing recruitment and distorted loudness perception — Defines recruitment as the phenomenon where quiet sounds are attenuated but loud sounds are normal due to outer hair cell damage.
• Hearing impairment degrades sensitivity, discrimination, and can cause hyperacusis or tinnitus — Defines the taxonomy of hearing impairment symptoms: threshold shift, discrimination loss, distortion, hyperacusis, and tinnitus.
• Recruitment makes loudness grow faster in impaired ears than normal ears. — Briefly clarifies that ‘recruitment’ is the term for abnormal loudness growth from outer hair cell damage.
• Hearing loss varies independently across different frequency ranges — Explains that sensorineural threshold shift is frequency-dependent, with different mechanisms (functional, linear attenuation, recruitment) in different bands.
• Adaptation, temporal integration, and post-masking can be simulated in a single auditory element. — Explains the architectural components (envelope detection, dual low-pass filters, logarithmic feedback) of a compact auditory model element.
• Cochlear hydrodynamics can be simulated by analog electrical equivalent circuits — Explains the electrical-equivalent-circuit mapping: membrane mass as inductor, damping as resistor, spring storage as capacitor.
• Outer hair cells provide active frequency-selective amplification in the cochlea — Explains the active vs passive amplification mechanism of outer hair cells in the basilar membrane.
• Annoyance and disturbance refer to distinct degrees of noise harm — Formal definition distinguishing objective harmfulness from subjective disturbance/annoyance, with measurement methods for each.
• Tinnitus can arise from unstable outer hair cell oscillation in the cochlea — Clarifies the physiological mechanisms — neural malfunction vs. cochlear oscillation — behind a phantom sound.
• Subtractive microphone designs produce cardioid and dipole patterns via port spacing and delay — Explains the mechanism behind first-order directional microphones and how cardioid patterns are physically produced.
• Beamforming combines multiple microphone signals into directive beams — Defines beamforming, the core technique behind modern smart speakers and hearing aid arrays.
• Cochlear implants bypass the conductive auditory path to stimulate nerve fibers directly — Clear definition of what a cochlear implant mechanically does and when it is indicated.
• Recovery from temporary hearing shift takes roughly double the exposure time — Gives the practical recovery rule and marks the TTS-to-permanent shift boundary.
• Comparing bone- and air-conduction thresholds distinguishes conductive from sensorineural hearing loss — Explains how bone-conduction audiometry works and how its diagnostic logic discriminates conductive from sensorineural impairment.
• Speech audiometry tests intelligibility by varying signal level and speech pace — Defines how speech audiometry is calibrated and administered as a diagnostic method.
• Otoacoustic emissions objectively reveal auditory system functioning and defect location — Explains what EOAEs indicate and limits of their diagnostic specificity.
• Eardrum and middle ear status can be measured objectively via acoustic impedance — Defines tympanometry as an impedance-based diagnostic technique.
• Hearing aids are miniature signal-processing sound systems that must secure adequate signal-to-noise ratio — Frames a hearing aid as a miniature signal-processing reproduction system, with SNR as the central design constraint.
• A healthy cochlea emits new frequencies when stimulated by two tones — Defines what distortion-product otoacoustic emissions are and how they are measured.
• Automatic gain control compresses sound to fit impaired hearing’s reduced dynamic range. — Defines AGC and recruitment-driven compression as responses to level-dependent hearing loss.
• Combining compression with a limiter reduces distortion better than a limiter alone. — Clarifies the distinct roles of noise gate, compression, and limiter in a hearing-aid gain curve.