Talk:Bark scale

meaning of critical band rate
The meaning of this term is not explained. I can guess at it, but ...! yoyo 08:02, 17 March 2006 (UTC)

critical band edges
Saying that the critical band edges fall at 0, 100, 200, 300, etc Hz seems to imply that the frequency spectrum is partitioned precisely by those frequencies into non-overlapping regions. But the explanation at critical band implies rather that at any frequency, there are sounds at both higher and lower pitch that the initial pitch will mask. If that is so, what I think this article ought to be saying is that the critical bandwidth increases (non-linearly), and tabulating the bandwidth for a sequence of increasing initial pitches. Unless the basilar membrane has a number of sharp discontinuities in its response curve, I don't see how it could possibly partition the spectrum in the manner indicated. yoyo 08:02, 17 March 2006 (UTC)

--> Hey, there is definitely a difference between the Bark Scale and the basilar membrane. That's an obvious and silly statement, because of course they're different--the Bark Scale has to do with measurement and is discontinuous (as all analysis is inherently), and the basilar membrane (organ of Corti) is the human organ, and IS continuous. I'm not familiar with the history of the Bark Scale, but I would presume that the "edges" of the bandwidth were determined after determining the centre frequencies--which this article doesn't mention, but https://ccrma.stanford.edu/~jos/bbt/Bark_Frequency_Scale.html mentions. And, as that page mentions, and as I was trying to clarify above, we have to remember not to confuse the measurement (Bark) with the human organ... the human organ doesn't sense these numbers, we APPLY these numbers (not completely arbitrarily, but we're still applying them) and build knowledge out from their application. So Bark, for example, probably started with 1000 Hz as a centre frequency and conducted the rest of the research centring on that--it looks as though, for whatever reason I'm not sure about, the ratio between centres is 1.17... i guess that's what they determined to be the just-noticeable-difference of frequency 24.84.231.7 (talk) 05:21, 23 November 2011 (UTC)p-man

--> Update: I have now updated the article to include the centre frequencies 24.84.231.7 (talk) 20:13, 23 November 2011 (UTC) p-man

--> Update: I have added a quote by Zwicker that hopefully clarifies the issue of critical band width versus edge/center values. What I think would be more useful than a list of the Bark Scale edges and centers would be a an equation and/or graph that shows how the width of the critical band increases with frequency. I'll examine more of the research and see if I can find this information. CarlGrundstrom (talk) 18:07, 20 February 2013 (UTC)

needs more detail
In my textbook, Keith Johnson's Acoustic and Auditory Phonetics, the explanation of the Bark scale is much more detailed, describing the motivation for using the bark scale, relating it to the cochlea, and with tables showing its effects. In fact, I don't think it even mentions the equation. Believe it or not, 13*arctan is hard for most people to visualize. This article needs figures and more explanation, and the equation should be deemphasized. —Preceding unsigned comment added by Depaderico (talk • contribs) 17:29, 8 October 2007 (UTC)

Still needs more detail. --Phil Wolff (talk) 02:56, 13 September 2012 (UTC)

question
The first equation on this page matches the band edges, but the other equations don't match those band edges at all. Why is that? — Preceding unsigned comment added by 68.34.19.73 (talk) 19:05, 13 November 2011 (UTC)

First Bark scale change
The first Bark scale (centre) has been changed recently from 50 to 60Hz. The original Zwicker paper was definitely clear that this was 50. I can't see any justification to change it to 60. Obviously it can be contested that this isn't the right number, but I think that if we accept the derivation of the Bark scale was defined by Zwicker then the number should be 50. Yes?