Re: [fluid-dev] OOM killer synths and ALSA overruns

[Josh Green]

On Tue, 2007-01-02 at 18:31 +0100, Frank Barknecht wrote:
> Hallo,
> Josh Green hat gesagt: // Josh Green wrote:
> 
> > As for good low-latency Linux response, not sure what the best
> > options are these days, since I haven't played with it much with
> > newer kernels.  I have heard though that unpatched the 2.6.x kernels
> > can be either very bad or moderately OK as far as latency, depending
> > on which particular version, etc. 
> 
> You really lag a bit behind here. ;) 
> 

Yep ;)  Haven't played much with low latency these days.  I'm usually
programming, which doesn't require very low latency (and no, I haven't
yet gotten around to tuning my computer keyboard IRQ, he he).  When I do
get around to playing music, its usually just improvising with a MIDI
keyboard and I don't really care if a few dropouts occur.  Yeah sure, at
some point I'll get around to actually recording/composing something
(thereby requiring lower latency), but right now I'm happy with it.

> To quickly get you almost up to date I would recommend Lee Revell's
> paper from the LAC 2006. Generally a stock 2.6 kernel now provides
> reasonable latency for many day-to-day audio work scenarios. I run
> stock kernels since about 2.6.16 (also because I became lazy and
> didn't want to follow the sometimes unstable RT-kernels) and it's
> more than enough for me.
> 

Just to nit pick on my statement, I said 2.6.x varies..  Older 2.6 was
crap.  And a few of the recent 2.6.x kernels were also rather bad.  Nice
to hear that they are paying more attention to it now though.  Thanks
for keeping us up to date on it and replying to my perfectly useless
statement.

> Ciao

Cheers.
        Josh