>From: fons adriaensen <[EMAIL PROTECTED]> > >On Thu, May 19, 2005 at 08:57:24PM +0300, Juhana Sadeharju wrote: > >> Yes. The priority should depend on the buffer size. >> Longer buffers would run with non-soft-RT schdulers. > >That problem is really the same as what you get with partitioned >convolution.
That was the problem which did lead me to develop the system a few year ago. Any hardware can be used in the system for lowest latencies. (hw -- RT Linux or soft-RT -- user scheduling) Juhana -- http://music.columbia.edu/mailman/listinfo/linux-graphics-dev for developers of open source graphics software
