Presented to AES Chicago
by Jon Boley
29 January, 2014
Assume we have a periodic signal (e.g., sine wave)
The phase is simply the position
in degrees or radians
...or even Hz×seconds
The instantaneous pressure of 2 or more signals can interact
(contructive & destructive interference)
e.g., to model reflections off surfaces (simple reverb)
(4kHz tone modulated at 500Hz)
Within a critical band, it is easier to detect AM.
But outside, phase doesn't matter.
Zwicker, E., Flottorp, G., & Stevens, S. S. (1957). Critical band width in loudness summation. The Journal of the Acoustical Society of America, 29, 548.
Affected by relative level & frequencies present
As Zwicker & Fastl note, phase changes often manifest as:
>30 degrees offset for a single component in a harmonic complex results in a new auditory stream
Kubovy, Michael, and Ray Jordan. "Tone-segregation by phase: On the phase sensitivity of the single ear." The Journal of the Acoustical Society of America66 (1979): 100.
Play noise in both channels,
but delay a narrow frequency range in one channel
Guess that tune!
This demo only works well with headphones
Heinz et al (2001) built a computer model to predict level discrimination at different frequencies.
It better predicted human performance when across-frequency neural timing was accounted for.
People can easily detect a tone in noise, and typical masking models do not match this ability.
Carney et al (2002) showed that a model that compared the phase across neighboring auditory filters can explain human performance.
Joris et al (2006) showed that some of the internal delay may come from phase differences in the cochlea.
reproduced with permission from Cedolin and Delgutte (2010)
For multiple simultaneous pitches:
Larsen (2008) showed that neural rate is useful for F0>900Hz
but for F0<900Hz, this 'spatio-temporal' derivative is more accurate
Neurons in the cochlear nucleus are sensitive to changes in the phase spectrum
... leading to improved localization?
Wang, G. I., and Delgutte, B. (2012). Sensitivity of Cochlear Nucleus Neurons to Spatio-Temporal Changes in Auditory Nerve Activity, Journal of Neurophysiology, 108, 3172-3195
The claim that the ear is "phase deaf"Laitinen, Mikko-Ville, Sascha Disch, and Ville Pulkki. "Sensitivity of Human Hearing to Changes in Phase Spectrum." Journal of the Audio Engineering Society 61.11 (2013): 860-877.
was already suggested by Ohm ,
and Helmholtz  came to the same conclusion in his tests.
Time-domain behavior... is not a serious factor, as humans are very insensitive to phase shift. Only very large group delays - larger than occur in normal loudspeaker designs - are audible, and even then normal listening room acoustics make the differences even more difficult to hear.Floyd Toole
Lipshitz, Stanley P., Pocock, Mark, and Vanderkooy, John, "On the Audibility of Midrange Phase Distortion in Audio Systems,' J. Audio Eng. Soc., Vol. 30, No, 9, Sept. 1982, pp 580-595.
- Even quite small midrange phase nonlinearities can be audible on suitably chosen signals.
- Audibility is far greater on headphones than on loudspeakers.
- Simple acoustic signals generated anechoically display clear phase audibility on headphones.
- On normal music or speech signals phase distortion appears not to be generally audible, although it was heard with 99% confidence on some recorded vocal material.