Re: Low Latency vs. Real Time Kernel

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On Tue, Feb 14, 2012 at 12:21 PM, Brendan Jones <brendan.jones.it@xxxxxxxxx> wrote:
On 02/14/2012 08:31 PM, Brian Monroe wrote:
I've been spending a lot of time on the #opensourcemusicians channel
talking to Ubuntu Studio users about their kernel and latency times
they're getting. Seems like most of them are using g a stock kernel with
the preemptive option enabled and they are getting great latency results
(2ms)while utilizing the @audio group on their user. I ended up
compiling my own low latency kernel and I haven't had any issues with it
yet. If this is what we are missing for the spin I'd be happy to
maintain packaging for the kernel.  I know ccrma has been behind a few
kernel releases.

Latency times are really relative things based on many factors. Have you been getting better latency times than ta

yeah.  

Currently 3.2.2-1.rt10.1 is in CCRMA testing (which is only one behind upstream which I believe Fernando has already started working on)

The patches are taken from here: http://www.kernel.org/pub/linux/kernel/projects/rt/



I saw the instructions for adding the real time patch for a tick less
kernel and from what I can tell it wouldn't be hard to get that rolling
as well.

I'm not entirely sure what ccrma does differently with their kernels
compared to other Linux users, and I'm still a bit of a noob so I could
be off base with this, but I would reason that we should be able to just
utilize the same settings to archive similar performance enhancements.

That question might be better directed towards CCRMA.

I haven't got that far yet but I would imagine that any kernel shipped with a Fedora endorsed spin would have to be generated by the kernel team (ie, from the Fedora build system). Having said that, latency times achieved using the stock kernels w/threadirqs and our standard jack limits.conf are more than sufficient for most users.

Is there a descent chance they would build a kernel for us? At least the preemptive option if not the rt patch that Fernando is applying? The latency times I've been seeing for stock kernels haven't been great. They're fine for mixdowns or mastering but trying to record live or using it for performances I just don't see myself being able to use it without some serious work. 

But maybe we're not trying to reach that user group? 

I just think it'd be neat to give someone all the tools they need for audio creation, with little to no configuration (like a custom kernel)

On a side note, I think Macs have almost all of the market share for this. I also think that linux in general has been moving towards a more user friendly experience, and for audio I would imagine having a spin that isn't going to require that you install a new repository to get a new kernel, a lot of musicians are nerdy, but they're not that nerdy. 

Now I'm just thinking out loud with this, and I don't know what the fedora culture is like but even if we could get it into standard repos would be a big deal. Even if that means we don't ship with the kernel installed, call it an "unstable-testing" kernel or something, that would be a big win for making things more practical for people using fedora for audio. I think people would try using it if there was ease of access. 
 



I thought I read that ccrma uses a unique scheduler, but if we could get
a 2ms latency time without it, the point may be moot.


Brendan

_______________________________________________
music mailing list
music@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/music

_______________________________________________
music mailing list
music@xxxxxxxxxxxxxxxxxxxxxxx
https://admin.fedoraproject.org/mailman/listinfo/music

[Index of Archives]     [Linux Audio Users]     [ALSA Users]     [Fedora Development]     [Fedora Desktop]     [Fedora Users]     [Gimp]     [Yosemite News]

  Powered by Linux