Re: Freq to bark domain (Matt Flax )


Subject: Re: Freq to bark domain
From:    Matt Flax  <flatmax(at)IEEE.ORG>
Date:    Fri, 31 May 2002 03:04:57 +1000

On Thu, May 30, 2002 at 10:47:47AM -0400, Vivek Srivastava wrote: > Hello List, > I have this quick question to ask about mapping from frequency to bark = > domain. Suppose I have values of a function(say some transfer function = > of a filter) which are at different frequency samples. I now need to = > find the equivalent values in bark scale. I know the conversion of = > frequency to Bark scale(used the Traunm=FCller equation) but then how do = > I map these values which I have in frequency domain to bark domain? = > Should I just take the average of those values(i.e values which I have = > in the freq domain) that correspond to each interval in the bark domain( = > I have taken intervals of 0.2 bark on the z domain i.e one or more = > frequency domain values will give me a single value for each interval) ? > Any help in this will be appreciated. > Thanks a lot, > With Best Regards, > Vivek Srivastava > ************************************************* > Graduate Research Assistant > Alexandria Research Institute,Virginia Tech > 703-535-3445(Office) > 703-582-9456(Mobile) > vivs(at)vt.edu > ************************************************* You can do something like averaging, this I guess is an issue of interpolation to preserve inferred detail. Perhaps I can suggest a relevant paper : "Bark and ERB bilinear transforms", Smith, J.O., III Abel, J.S. Speech and Audio Processing, IEEE Transactions on,page(s): 697 - 708 Nov. 1999, Volume: 7 Issue: 6 >From the abstract : "A filter design example is included that illustrates improvements due to carrying out the design over a Bark scale." check it out Matt -- http://mffm.darktech.org WSOLA TimeScale Audio Mod : http://mffmtimescale.sourceforge.net/ FFTw C++ : http://mffmfftwrapper.sourceforge.net/ Vector Bass : http://mffmvectorbass.sourceforge.net/ Multimedia Time Code : http://mffmtimecode.sourceforge.net/


This message came from the mail archive
http://www.auditory.org/postings/2002/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University