I made a comment on the patch tracker concerning this, directly on the jack patch. However, looking at the patch, I cannot see how it would effect pd's audio when running in real-time. All I know is I ran pd-0.40.3-extended from April 15th yesterday, using jack in real-time, for a few hours and the audio was almost flawless. <br>
<br><div class="gmail_quote">On Wed, May 21, 2008 at 3:34 PM, Rich E <<a href="mailto:reakinator@gmail.com">reakinator@gmail.com</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br><br><div class="gmail_quote"><div><div></div><div class="Wj3C7c">On Wed, May 21, 2008 at 3:20 PM, Ken Restivo <<a href="mailto:ken@restivo.org" target="_blank">ken@restivo.org</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div><br>
On Wed, May 21, 2008 at 02:23:20PM -0700, Rich E wrote:<br>
> ><br>
> > IIRC, Ubuntu "RT" kernel is not actually Real Time (i.e. with Ingo<br>
> > patches), just a "desktop RT" which is not very useful for audio and<br>
> > certainly not enough for JACK apps of any heft.<br>
> ><br>
><br>
> Where are you getting this info from? This is not what I have read, but<br>
> maybe I am missing something somewhere. I thought that the Ubuntu kernel<br>
> packagers recently switched to using full realtime preemption... I find this<br>
<br>
</div>My info is from having installed Ubuntu and used it... but that was a few years ago. Glad they're using Ingo's patches now.<br>
<div><br>
<br>
> in the package description of linux-image-2.6.22-14-rt :<br>
> Ingo Molnar's full real time preemption patch (2.6.22.1-rt9)<br>
><br>
> Here is the output of uname -a:<br>
> Linux pal 2.6.22-14-rt #1 SMP PREEMPT RT Tue Feb 12 09:57:10 UTC 2008 i686<br>
> GNU/Linux<br>
><br>
> .. and here are the kernel settings from cat /boot/config-2.6.22-14-rt |<br>
> grep PREEMPT<br>
> # CONFIG_PREEMPT_NONE is not set<br>
> # CONFIG_PREEMPT_VOLUNTARY is not set<br>
> # CONFIG_PREEMPT_DESKTOP is not set<br>
> CONFIG_PREEMPT_RT=y<br>
> CONFIG_PREEMPT=y<br>
> CONFIG_PREEMPT_SOFTIRQS=y<br>
> CONFIG_PREEMPT_HARDIRQS=y<br>
> CONFIG_PREEMPT_BKL=y<br>
> CONFIG_PREEMPT_RCU=y<br>
> # CONFIG_DEBUG_PREEMPT is not set<br>
> # CONFIG_CRITICAL_PREEMPT_TIMING is not set<br>
><br>
> Looks realtime to me.. also the latency is really low (5-10ms<br>
> conservatively), much lower than I have experienced on other OS's, with or<br>
> without realtime. But, if you know something I don't... I have open ears.<br>
<br>
</div>Nope, you got real-time there, for sure. I was wrong then.<br>
<div><br>
> Ubuntu runs so nicely on a laptop though... would hate to switch after<br>
> finding such an easy-to-operate linux distribution because pd and only pd<br>
> doesn't like jack in realtime.<br>
><br>
<br>
</div>Dunno. You're playing Xrun whack-a-mole. Done that; it's not fun.<br>
</blockquote></div></div><div><br>Ken, this made me laugh for quite a while. Thank you.<br> </div><div class="Ih2E3d"><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
Here's what I did to RT-ify my system (on Debian, but Ubuntu is based on it):<br>
<a href="http://www.restivo.org/blog/archives/preparing-a-debian-etch-system-for-audio-use" target="_blank">http://www.restivo.org/blog/archives/preparing-a-debian-etch-system-for-audio-use</a><br>
<br>
Also, I expect you know to use chrt to give the IRQ of your audio interface top priority, higher than JACK and any other process. That's a crucial step. If I omit that, then I get clicks and pops too, which is why I wrote a script to do it automatically. You could also install and configure the "rtirq" package, which works great too.<br>
<font color="#888888"></font></blockquote></div><div><br>I had it on my last setup, but for got to do it this time. Thanks for reminding me :) Also, as my other post mentions, I sometimes got clicks with pd even with the script. I actually had to change the irq priority of pd to get rid of them, which is bad since it changes all the threads pd has instead of only the scheduler. I would get "I/O stuck... closing audio" all the time.<br>
</div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><font color="#888888"><br>
-ken<br>
</font></blockquote></div><br>
</blockquote></div><br>