Sound Similarity

[Sarah Thompson]

I'd suspect that an approach based on resampling to an equivalent bit
depth/sample rate first, followed by overlapping FFTs with an appropriate
window function, then comparing the results of the FFTs on something like a
'sum of the squares of the differences' basis would be a good starting
point. You could use cluster analysis then to figure out a 'family tree' of
sounds. This wouldn't deal with situations where sounds were offset in time
however - you'd need to figure out how to synchronise everything.

Sarah
-----Original Message-----
From: Brian Whitman <bwhitman at crudites.org>
To: music-dsp at shoko.calarts.edu <music-dsp at shoko.calarts.edu>
Date: 21 September 1999 23:24
Subject: Sound Similarity


>
>Does anyone have any pointers on sound-similarity algorithms? For example,
>comparing two sound files independent of format or bitdepth/frequency
>range and 'score them' based on how 'similar' they were?
>
>
>Brian Whitman
>bwhitman at crudites.org
>http://www.netspace.org/users/bwhitman
>
>
>
>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/