Re: real-time threaded IO with low latency (audio)

Daniel Kobras (daniel.kobras@student.uni-tuebingen.de)
Sat, 24 Jul 1999 15:43:39 +0200 (CEST)


On Sat, 24 Jul 1999, Gerhard Mack wrote:

> On Fri, 23 Jul 1999, Ingo Molnar wrote:
>
> > the whole 'why sound skips' problem i think lacks one major necessery
> > component: information. Nobody really ever analyzed what the problem
> > really is/was. I do have a awe64 soundcard (which i almost never use), if
> > you (or anyone else) could point to a method to make sound-playback
> > produce bad quality, i'd be happy to analyze the problem. Is running
> > mpg123 enough to see the problem? What should i do to make sound skip?

> Mpg123 will skip as the disk usage gets heavy.
>
> Somewhere between 2.2.8 and 2.3.4 something drastically reduced the sound
> dropouts on my system. I used to be able to make the sound skip even with
> "mpg123 -b 1024" but that is no longer the case.

In my experience setting the buffer size to a higher value won't do you a
blind good as the problem is not input bound but rather on the output
side. Heavy disk i/o locks out mpg123's buffer process, therefore
audio_play_samples() doesn't get run often enough. So to me it looks like
mingos additional schedule()s are certainly in the right ballpark. For a
simple test case on sound skippishness, disable DMA on an IDE disk and run
a find on it while playing sound. [Or be the proud owner of any IDE
chipset that wasn't supported in DMA mode prior to unified ide. :-/]

Daniel.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/