Re: 2.6.3x-rty where x > 3 and y >= 1?

[Robin Gareus <robin@gareus.org>]

On 11/19/10 22:18, Niccolò Belli wrote:
> Il 11/11/2010 23:49, Ng Oon-Ee ha scritto:
>>> Glad to hear that, I can't use 2.6.33-rt anymore because of problems
>>> with my intel video card and my band is starting to blame because of xruns.
>>>
>>> Darkbasic
>>
>> In the meantime, the -bfs kernel does improve the situation somewhat.
> 
> I just compiled 2.6.36-bfs, but unfortunately alsa git does not compile
> against it (I need alsa git because of M-Audio Fast Track Ultra support).
> 
Don't worry. It's not worth the hype it generated. I'm just back from
rolling a new kernel for 64studio. I tried a couple of candidates
including -bfs.

2.6.33.7-rt29 is _by far_ the best option for low-latency audio.
Well, actually it's the only option (not counting older rt-linux
releases) which provides reliable <=4 ms latency (16 or 32 audio-frames
per period) without any x-runs!

Vanilla 2.6.36 and -bfs kernel both perform quite well. If >=15 ms
latency is sufficient for your use-case and you can live with some
occasional x-run that is.

I'm looking forward to a new -rt patch.

Cheers!
robin

PS. Note that sound latency here is not calculated (eg. 32*2/48kHz) but
measured with jack_delay from http://www.kokkinizita.net/linuxaudio/
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html