Yup, Josep beat me to it. We got great performance with krate between
100-440, and lots of use of the upsamp and interp opcodes. They work
amazingly well. And if you are doing a lot of modulation math ( ie envelopes
modulating lfos and patch settings ) the savings are very significant
between doing the math at krate and then interpolating at the very end on
only the final applied variable, vs doing all the math with higher krate
variables or even arate variables.

Does this make sense? If not I can find an example around here and post it.

Iain


----- Original Message -----
From: "Josep MComajuncosas" 
To: 
Sent: Tuesday, May 25, 2004 7:59 AM
Subject: RE: [CsndTek] A-rate vs K-rate


> Use kr=100 or so, maximum 440, but it's not necessary. Keep as much
> variables as possible working at k-rate: envelopes (sometimes you'll
> need a-rate envelopes), lfo's, GUI-driven controls.... Just at the end
> upsample or interp them to a-rate to get better quality. If necessary
> you can encapsulate sections of code in instruments working at kr=sr.
>
> Josep M Comajuncosas
>
> _______________________________________________
> csoundtekno mailing list
> csoundtekno-N4abDuUB7xo@public.gmane.org
>
> Subscribe, unsubscribe, change mailing list options:
> http://plot.bek.no/mailman/listinfo/csoundtekno

_______________________________________________
csoundtekno mailing list
csoundtekno-N4abDuUB7xo@public.gmane.org

Subscribe, unsubscribe, change mailing list options: