[music-dsp] Whats wrong with my FFT?

Christopher Penrose penrose at sfc.keio.ac.jp
Fri Apr 20 02:41:38 EDT 2001


I hadn't looked at Julius's resample code for several years or heard its 
results since, and my past observations were based on version 1.2 of his 
resample software.  The filter code appears to use a windowed FIR 
filter, so my recollection of IIR artifacts were mistaken.  My apologies.

Christopher
penrose at sfc.keio.ac.jp


On Friday, April 20, 2001, at 02:29  PM, Bob Cain wrote:

>
>
> John Stewart wrote:
>>
>> On Thu, 19 Apr 2001, Bob Cain wrote:
>>> Could you elaborate a bit on this?  I haven't encountered the method.
>>> Given an arbitrary factor, for example a stimulus 2,397,112 samples 
>>> long
>>> and a response 2,394,513 samples long how would you proceed to make 
>>> the
>>> stimulus length equal the response length?  These are real numbers 
>>> for a
>>> sweep based stimulus/response pair that I obtained using a CD player 
>>> and
>>> a DAT.
>>
>>         You might find the following link useful:
>> http://www-ccrma.stanford.edu/~jos/resample/resample.html
>
> Very nice piece of work.  The use of a fixed size table with linear
> interpolation is an obvious way to approximate the sinc function
> coeficients at aribitrary points without evaluating a bunch of trig
> functions but his very detailed and comprehensive error analyis leading
> to necessasary and sufficient conditions for the error to be less than
> quantization error is hardly obvious and an important result.  I haven't
> followed that analysis through because it is quite detailed but I
> suppose I can presume it has been checked by others and found valid.  In
> the end the algorithm is of the order of complexity of a filtering
> operation done by convolution where the filter coeficients vary for each
> output sample and are calculated by a table lookup and a set of linear
> interpolations.
>
> Since the sample rate factor can change continuously this algorithm
> could be a basis for a time/pitch morphing transformation such as is
> used in the Pitch N' Time plugin for Pro Tools, something I've been
> looking for since they don't want to do one for Dirext X.
>
>
> Many Thanks,
>
> Bob
> --
>
> "Things should be described as simply as possible, but no simpler."
>
>                                              A. Einstein
>
>
> ////////////////////////////////////////\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\
> \\\\\\
>
>  To contribute your unused processor cycles to the fight against cancer:
>
>      http://www.intel.com/cure
>
>  To feed a hungry kid today at the cost of a click follow this link:
>
>      http://www.thehungersite.com/cgi-bin/WebObjects/HungerSite
>
> \\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\/////////////////////////////////
> //////
>
> dupswapdrop -- the music-dsp mailing list and website: subscription 
> info,
> FAQ, source code archive, list archive, book reviews, dsp links
> http://shoko.calarts.edu/musicdsp/
>
>

dupswapdrop -- the music-dsp mailing list and website: subscription info,
FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/




More information about the music-dsp mailing list