Sub-threshold echoes color sound and can even impart pitch

Evidence

Suppose a sound is repeated with a short delay, of 0.05 to 0.001 s, corresponding to audio frequencies; such repetitions are echoes that are too fast to be heard separately. They do however “color” the sound, even imparting a pitch in some circumstances, as we will discuss further in sections 21.2 and 23.17. This applies to random white noise as well, repeated with a delay as follows: white noise is recorded or generated, and the signal is added to itself with a time delay, say, 5 ms, making a modified file. Clearly, if a large blip exists in the original file, the blip is now also found 5 ms later in the modified file. This gives a sharp peak in the autocorrelation function at 5 ms. — Eric J. Heller, Why You Hear What You Hear- An…, 427

Categorization

• XREF: Explains comb-filtering and the ‘singing’ of delayed white noise — connects to acoustics of small rooms and the ear canal itself.