[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: About importance of "phase" in sound recognition



Dear All,

Thanks a lot for all of this great and informative comments, I appreciate them all. like stated in a lot of other comments I had done that randomization of phase of fft coefficients and usually for long signal lengths the distortion was pretty clear, however there was always the stereotype of deafness to short-term phase information as Ken stated.

However based on the comments from Roy and Ken, I wanted to reformulate my question more practically. If the relative phase of activity in different channels doesn't matter, does it mean that if you shift the spike trains in auditory nerve fibers of different CFs, with respect to each other it would not have any effect ? or in other words if you change the order of spikes among different channels in the auditory nerve, it wouldn't have any effect on the perceived input ?
If so, this just seems strange because then it means that apparently there is not collective temporal structure at the auditory nerve level, and perhaps the spike trains in each fiber are at least in a short term sense (as Ken refers to) independent of each other ?
the answer to these questions would be crucial for me as i am trying to design a learning scheme for learning such spike trains and i need to know what is and what is not important.

Also regarding Ken's argument if the processing needs to be short-term is there any, precise definition for this short time period in a biological system ? I mean do you think that in a specific integration window of a typical neuron in the auditory pathway, is adjusted in away that phase (mainly across different afferents ) would not matter in that window ? i.e is it "spike order insensitive" ?!

By the way, I apologize for not providing the reference for the mathematical article that I referred to in the previous email. here is a link to it :

http://www.math.missouri.edu/~pete/pdf/132-painless.pdf

Thanks a lot again, I just can't wait see receive more comments and gain more information from this interesting thread.

Best
Emad


2010/10/8 Laszlo Toth <tothl@xxxxxxxxxxxxxxx>
On Thu, 7 Oct 2010, Kevin Austin wrote:

> When I first learned that phase was not important to hearing, it was in
> a specific context. It was, (for example) with a sawtooth wave, invert
> it. It will sound the same, even though it is 180° out of phase.
> Therefore [?] the phase of the signal can be shifted by 180° (or
> inverted), and the ear is insensitive to this.

Interestingly, a parallel thread is running on the music-dsp mailing list
just about the same topic. You can find the archive here:
http://music.columbia.edu/pipermail/music-dsp/2010-October.txt
In particular, Sampo Syreeni wrote:
"But perhaps the most shocking discovery here is that with sufficiently
large test sets, 180 degree phase differences can be distinguished at a
statistically significant level, monaurally only. That I think was what
lead to the periodicity/rectifier models of pitch perception in the
first place, long before e.g. missing fundamental kind of phenomena were
propertly attributed to the nonlinearity of the cochlear cilia."