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

Re: Rationale for Critical Bands

Martin, your informal observation of bandwidth "as good as unchanged in this level range" may also be pretty consistent with peripheral models that exhibit a modest bandwidth change over a very large range of levels. Has anyone attempted to test a hypothesis about whether the psychophysical effects are constant, as opposed to somewhat level dependent?

As you and Brian have noted, speech without interference doesn't show much level-dependent effect, but speech in noise does. But neither of these is directly a measure of the critical band; Brian's working hypothesis that the reduction in intelligibility of speech in noise at higher levels is due to wider critical bands sounds plausible. But we'd need a different experiment to test whether the critical bands do actually change; the auditory filter fits for tone in noise detection seem like they do just that. If you reject those results, can you explain why? Or you just focus on different aspects of CB to which those results are not relevant?

I've been reading up on the concept of critical band in IC, which is what you've been talking about. I have this book on IC, which one can read parts of online:
The section "Origins of Spectral Tuning and Resolution" in the chapter by Ehret and Schreiner starts off with something that I think we can all agree on:

"Cochlear filtering represented in the RFs of auditory nerve fibers is the basis for spectral tuning in higher centers of the auditory pathways."

But then they get into an interpretation of "critical bands" that's one I'm not familiar with, but probably is what you are talking about, when they say on p.328 "The origin of neural critical bandwidths with properties of psychophysical critical bands is associated with ICC neural filtering properties" and "The steep slopes of the critical-band borders...". On p.325 they say "Perceptual critical-band properties are absent in auditory nerve fibers (Ehret 1995)"; so I'll guess I'll have to find that and see what they mean. It's in a book that I can't find a copy of: Advances in Hearing Research: Proceedings of the 10th International Symposium on Hearing : Swabian Conference Centre, Irsee, Bavaria 26 June-1 July, 1994, Geoffrey A. Manley, G. M. Klump, C. Koppl, H. Fastl, H. Oeckinghaus (eds.); anybody have a copy?

They also reference these papers for showing that the "critical-band like response" arises first in IC:

Stanley Zerlin
Electrophysiological evidence for the critical band in humans
J. Acoust. Soc. Am. Volume 79, Issue 5, pp. 1612-1616 (May 1986)

Dennis L. Burrows and S. Joseph Barry
Electrophysiological evidence for the critical band in humans: Middle-latency responses
J. Acoust. Soc. Am. Volume 88, Issue 1, pp. 180-184 (July 1990)

So I got these to see what they say. Basically, they're about loudness summation, and finding evoked potentials that show a loudness-correlated CB effect in their response to tone pairs or noise bands. So I think I see now where these guys are coming from -- you can't get loudness summation effects directly in the auditory nerve responses, since you need to aggregrate in some way across frequencies. I can well believe that that is part of what goes on in IC. I still don't see any evidence, however, that the question of whether the CB inferred from such evoked responses is level dependent or not has been addressed by any of these authors. Rather, they appear to assume that it is not level dependent. Am I missing some important piece of the puzzle?

When I said "I don't think there's much actual disagreement" I may have been overlooking something; but it's still not clear to me what it is. I know that Langner and Schreiner and Ehret do great work in IC, but I don't understand from their papers what exactly they have found, or are assuming, about critical bands or their level dependence.

It's interesting that the Schreiner and Langner abstract that you cite starts with "The perception of sound is based on signal processing by a bank of frequency-selective auditory filters, the so-called critical bands." This very direct association of "critical bands" as a set of real discrete "auditory filters" puzzles me. The "critical band" is a psychophysical concept, and the auditory filters usually are, too, though they can have physiological counterparts in some studies. The relationship to real mechanisms is what they're exploring, yet they refer to the auditory filters as if they are specific real things. And this usage seems to be contrary to the loudness summation effects that they seem to emphasize in IC, which is something that follows the filters.

So I guess it all comes down to being careful to say what we mean by critical band or auditory filter when we make claims about it. I'll try to be careful if you will.