Subject: Re: [linux-audio-dev] matrix routing
From: Steve Harris (S.W.Harris_AT_ecs.soton.ac.uk)
Date: Sun May 20 2001 - 14:26:43 EEST
On Fri, May 18, 2001 at 01:18:04PM -0400, Paul Davis wrote:
> well, its not as bad as I indicated in another message. i realized (i
> think!) that if we require a plugin to either do in-place processing,
Sounds good.
> you know what else is good here? all the plugins can use the "run()"
> (assignment) approach when processing data in their output buffer, but
> the engine (not the plugin) can optimize moving data from the input
> sources into the output buffers. for example:
Thats definatly good.
> in the above case, i would just run the most-recently created of plugin{A,B}
> first, which is as optimal as the other way around.
Yes, I think that is fine, as long as the logic is known and guaranteed,
there are times when the order the engine executes in can be important.
You don't want to load your network into another server (assuming there is
more than one) and find it sounds different.
> this way, the loop you wrote gets a partial execution when just before
> we execute each plugin, affecting only the ins/outs that involve that
> plugin. sound OK?
Sounds OK, it remains to be seen wether its too CPU intensive or not, I
guess it depends on how many non-zero gains there are. This would be a
really good place to apply SSE instructions, but I don't think many
versions of gcc support them yet.
- Steve
This archive was generated by hypermail 2b28 : Sun May 20 2001 - 13:50:17 EEST