Subject: [linux-audio-user] Re: [Jackit-devel] [Fwd: Re: Linux audio latency]
From: Lee Revell (rlrevell_AT_joe-job.com)
Date: Sun Oct 03 2004 - 01:26:10 EEST
On Sat, 2004-10-02 at 18:03, John Hedditch wrote:
> > On Sat, 2004-10-02 at 17:31, Florian Schmidt wrote:
> > > Also maybe the alsa drivers themselfes might be problematic/buggy. If
> > > the above minimal alsa app still produces xruns, then either alsa lib or
> > > the drivers themselfes are to blame [probably].
> > >
> > > Of course such a test is not a proof but might give important clues.
>
> How about the following?
> arecord -t raw -r96000 -fU24_LE -c2 --period-size=64 --buffer-size=128 fish
>
> This work just fine (no xruns).
>
> If we tried to drop down to 32 frames per period it (unsurprisingly)
> breaks, however, and we get
> the following:
> arecord -t raw -r96000 -fU24_LE -c2 --period-size=32 --buffer-size=64 fish
> Recording raw data 'fish' : Unsigned 24 bit Little Endian, Rate 96000 Hz, Stereo
> overrun!!! (at least 0.087 ms long)
> overrun!!! (at least 0.053 ms long)
> overrun!!! (at least 0.152 ms long)
> overrun!!! (at least 0.055 ms long)
> overrun!!! (at least 0.055 ms long)
> overrun!!! (at least 0.054 ms long)
> overrun!!! (at least 0.064 ms long)
> overrun!!! (at least 0.070 ms long)
> overrun!!! (at least 0.166 ms long)
> overrun!!! (at least 0.175 ms long)
> overrun!!! (at least 0.165 ms long)
> overrun!!! (at least 0.175 ms long)
>
> Anyway, this tends to support the jack client / jackd hypothesis, yes?
>
Right, this is exactly the expected behavior with the VP kernel. 64
frames at 96kHz (or 32 at 48kHz) is about the lower limit of PC
hardware, I would expect 32 @ 96 to break.
Try to isolate the jack clients that don't work...
Lee
This archive was generated by hypermail 2b28 : Sun Oct 03 2004 - 01:28:42 EEST