Hi csounders,

I was wondering if there was a more clever way to automate the creation of mutiple samples (or more specifically, soundfonts) than manually scripting a bunch of .csd files, each having their own wav file output name and single score event that would represent a sample.

My idea is to avoid processor overhead for certain realtime physical modelling orchestras which are complex (e.g. an orchestra rich in instances of the 'mode' opcode) by creating a non-realtime multi-sample set or soundfont and using that in live situations.

It woukd be nice to automate the creation of such multi-samples, but short of creating, or scripting the batch creation of a .csd for each sample, I'm not sure there are other more direct ways within csound to do this....are there?

Best,
AKJ

Hi Aaron,

I'm not sure if this would save all that much time, but you could
create a single CSD that uses macros, then call csound with
--omacro:xxx=yyy and --smacro:xxx=yyy to replace values.  If you wrote
the sample just to outs, you could then automate the CSD calling and
replace the orc and score macros and the -o outputName_xxx.wav.

steven

On Wed, Jul 24, 2013 at 10:12 AM, Aaron Krister Johnson
 wrote:
> Hi csounders,
>
> I was wondering if there was a more clever way to automate the creation of
> mutiple samples (or more specifically, soundfonts) than manually scripting a
> bunch of .csd files, each having their own wav file output name and single
> score event that would represent a sample.
>
> My idea is to avoid processor overhead for certain realtime physical
> modelling orchestras which are complex (e.g. an orchestra rich in instances
> of the 'mode' opcode) by creating a non-realtime multi-sample set or
> soundfont and using that in live situations.
>
> It woukd be nice to automate the creation of such multi-samples, but short
> of creating, or scripting the batch creation of a .csd for each sample, I'm
> not sure there are other more direct ways within csound to do this....are
> there?
>
> Best,
> AKJ