[gmpi] Re: NAMM follow-up, some major decisions to make

Paul Davis <paul@xxxxxxxxxxxxxxxxxxxxx> writes:

>>My guess is tho', this would require support in the OS for very 
>>efficient, deterministic, low-latency context switching and fast IPC- 
>>not sure how well current desktop OS match up to this :-/
>
> s/current/a particular current/   :)
>
> FWIW, using the 2.6.11 linux kernel with the new "Realtime Preemption"
> patch, I now know of at least a couple of people successfully running
> their linux systems as real time fx processors with a dropout-free
> interrupt interval of 16 frames at 48kHz (0.3msec per buffer = 0.6msec
> roundtrip latency). Processor load is a bit high (around 25%), but it
> seems perfectly viable given the right hardware.

Mac OS X also handles low-latency scheduling reliably.  

Recent Linux tests ran quite well, even without the low-latency
patches Paul mentions.  Those fixes are gradually being integrated
into the mainstream kernel.  Now, even the vanilla Linux 2.6.10 kernel
works stably at relatively low latencies.

My tests ran 20 JACK clients using POSIX SCHED_FIFO with a 64-frame
cycle at 44,100Hz (1.45 msec per cycle, about 3msec end-to-end
latency).  The CPU is a mere 1.5GHz Athlon 1800+ XP.  A second set
were done with an ardour compile running continuously in the
background.

Those results do require tuning the hardware and system configuration
to eliminate other sources of systemic PC latency problems, of course.
-- 
  joq

----------------------------------------------------------------------
Generalized Music Plugin Interface (GMPI) public discussion list
Participation in this list is contingent upon your abiding by the
following rules:  Please stay on topic.  You are responsible for your own
words.  Please respect your fellow subscribers.  Please do not
redistribute anyone else's words without their permission.

Archive: http://www.freelists.org/archives/gmpi
Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe

Other related posts: