Predictable sounds carry less information than unpredictable broadband signals

Evidence

How does this relate to an audio signal? Well, consider a sine wave; it sounds pretty boring to listen to because it is very predictable. On the other hand, a piece of music that is jumping around or an instrument whose texture is continuously changing is much more interesting to listen to because it’s more unpredictable or surprising. If we looked at the spectrum of a sine wave, we would find that all the energy is concentrated at one frequency, whereas for the more interesting music signal, it’s spread over lots of frequencies. In fact, the audio signal that carries the most information is either random noise or a single spike that happens at a time you don’t or can’t predict. Interestingly, in both these cases, the spectrum of the signal contains an equal amount of energy at all frequencies. — David M. Howard, Jamie Angus, Acoustics and Psychoacoustics, 2593

Categorization

Category: Causal Relationships

• QUESTION: Worth chasing: the claim that random noise and an unpredictable single spike both have flat spectra yet feel wholly different perceptually — the mechanism linking flat-spectrum to maximal information isn’t fully unpacked.
• XREF: Connects to the reader’s existing notes on keynote sounds vs. signal sounds from The Soundscape — predictability directly parallels how keynotes recede from perception.

Related

• Maximal-information audio signals have flat spectra across all frequencies (similarity 0.91)
• Random noise and unpredictable spikes carry maximum information due to flat spectra. (similarity 0.87)
• Natural sounds are identified by their statistical properties, not by predictable patterns of elements (similarity 0.82)
• Hearing-loss risk dominates noise research because it is predictable from sound properties (similarity 0.80)
• Sound textures are predictable only through their statistical spectrotemporal structure. (similarity 0.80)
• Perceived loudness depends on frequency content and duration, not just sound pressure level (similarity 0.79)
• All information-bearing sound must vary over time (similarity 0.79)
• Auditory textures are stimuli whose only predictability lies in their statistical structure (similarity 0.79)
• Changing a tone’s frequency alone can make listeners perceive motion without source movement (similarity 0.79)
• Broadband noise yields better intensity discrimination than pure tones by engaging multiple frequency regions (similarity 0.78)
• Efficient audio coding discards predictable signal components and encodes only novelty (similarity 0.78)
• Audio symbols with high probability carry less information and need fewer bits (similarity 0.78)
• No sound produces a deterministic internal neural response, even for the same listener (similarity 0.78)
• Real audio signals spend more time at low amplitudes than high amplitudes (similarity 0.78)
• Noise is not unwanted sound but information-poor spectrum useful for site assessment (similarity 0.78)