[music-dsp] Optimizing Dithering Code

[Chun-Yu]

Hi,

I am trying to optimize a class for dithering and noise-shaping > 16-bit
audio to 16 bits.  Currently, it is quite fast, but I get the feeling that
it could be faster.  I've tried to use SSE code to speed it up, but it
doesn't seem to make much of a difference (probably because I'm not
exactly the greatest assembly programmer).  If anyone would like to take a
look at it and give me ideas on optimizing it, that would be great.  I've
put the code along with a small test program up at
http://php.indiana.edu/~cshei/Dither Code Test.zip .  So far, I've gotten
VC++.NET to process one 32-bit sample -> 16-bit sample in 100 clocks
(without SSE) and 107 (with SSE - my SSE code is actually slower!) on my
866 MHz P-III.  VC++ 6 seems to be a little bit slower than that.

Chun-Yu


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/