I really don't recall where this pd came from, but command line says:
pedro@arkana:~$ pdextended -version
Pd version 0.42-5extended-20100530
compiled 09:02:05 May 30 2010

And yes, I'm on karmic (9.10).

>On monday, I will try to run my patch with the 0.42.5 rc4 on an Hardy... we will see if it works or >not...
In my mind it doesn't "compile" that the issue would be much Ubuntu-version-dependent, but if you say you could easily run 4 instances of this exact patch in Hardy and cannot no longer in Lucid.. something is going on (as Americans say: something "fishy").

Good luck with that :)
Pedro

On Sun, Jul 11, 2010 at 1:39 AM, Olivier Baudu <lamouraupeuple@gmail.com> wrote:
Thank you Pedro for all your tests and posts...

You said you are on an Ubuntu 9.10.
Do you use the pd-extended 0.42.5 nightly auto-builded or did you build a 0.41.4 for 9.10 ?

On monday, I will try to run my patch with the 0.42.5 rc4 on an Hardy... we will see if it works or not...

My problem is that I can't install Ubuntu 8.04 on my new laptop because it's too young for all his hardware to be recognized.
And I confirm that everything is working fine with an Hardy, Pd-ext 0.41.4 and jack from the official repository.

01ivier.




2010/7/7 Pedro Lopes <pedro.lopes@ist.utl.pt>

I was on the train so I had to kill some time :)

I tried running jack with buffer size 4k and it worked a lot better, using H4 usb audio. Just one instance dropped jack percentage from ~15%to 1/2 %.

It stopped giving A/D/A sync errors in pd console too. But when I fire the second instance... the CPU just bursts to load the patch and once it run with a lot of buffer glitches and another time it didn't even load and jack got so slow that was zombified.

Weird thing is a lot of times, the second instance dies, and jack survives... thus could imply that pd is actually crashing for some odd reason. I should try this in my other Linux too, to see if I can get any meaningful information - so far all I did was blab, sorry.

best of luck with this,
Pedro

p.s: I never got this behaviour with any other pd file, should be interesting to see what others on this list report using your patch. I usually never fire 4 pd instances, but sometimes I use two, on windows I had a stupid habit of using zillions (like 5 or more) instances of pd because I double clicked a file and he always opened a new instance for that =P But never had any trouble.



On Wed, Jul 7, 2010 at 1:33 PM, Pedro Lopes <pedro.lopes@ist.utl.pt> wrote:
Hi Olivier,

I dont have much time on my hands now, but I tested your patch in my Ubuntu 9.10 using jack. results:

1) I lauched the first one, it took quite a while and CPU bursted as soon as the patch opened (meaning.. heavy patch :D)
1.1) I draw and played a sequence, audio is okay.
2) I launched the second. Major CPU burst, that pdextended went almost to 90% cpu all the time
2.1) The audio began to glitch. (but no xruns on jack)
3) I launched a pdextended just to play test audio sinewave, it launched fairly quick (of course the CPU was busy with other instances..so not as quick as usual) and played without glitch.

I need to test more with more time on my hands, but it seems that its more related to how heavy the script is rather than the jack setup. But i really need to test it further, and change jack latency to see if it affects that much... jack never went past 10% os usage, which is normal for that many instances (sometimes i use pd and renoise, and it goes much higher).

My jack setup was:
- rt (I have rt kernel)
- buffer at 512 and 3 period
- using the onboard sound card which is sh*t
- my laptop is also sh*t for nowadays standards (AMD 1.9 Turion X2)

Best regards,
Pedro



On Wed, Jul 7, 2010 at 1:14 AM, Olivier Baudu <lamouraupeuple@gmail.com> wrote:
Hi.

I'm going to try to expose the problems I've met during the last two month dealing with the title of my post.
Sorry if it's confused.

This winter, I used to work on an Ubuntu Hardy, Extended 0.41.4 and with the jackd and qjackctl from the official repository (by the way, my sound card is a firewire Presonus Firepod).
I had no problem to use 4 different instances of Pd at the same time using jackd as sound server. (and I can precise I didn't need RT-kernel to do what I wanted).

I recently change for Ubuntu Lucid.
So I also change for Pd 0.42.5 (rc4) and the new official jackd and qjackctl.

If I use the kernel 2.6.32-23, I can play my patch on one instance, it's stable but the sound crunch a little bit sometimes.
I also can execute my patch on 4 instances, but the sound is awful and 3 minutes later, Jack quit.

So I've try the 2.6.31-11-rt.
In Pd-ext, If I change directly from ALSA to Jack, pd quit. Every time I have to change to OSS before to choice Jack and I can't save the jack configuration.
On one instance the sound is perfect and stable.
It's possible to launch 4 instances without patch and to hear simultaneously the 4 tests tones on my sound card using jack.
But it's impossible to launch even 2 instances with my patch.
Both quit at the same time when I try to pass from OSS to Jack on the second.
Console message just tell me that GUI of Pd have quited.

My patch is the one you can find on this page : http://yamatierea.org/papatchs/squareroom/OSR.htm
Direct link for the patch : http://yamatierea.org/papatchs/squareroom/OSR.zip
But be careful if you look inside... it's still a little bit fuzzy...

Do you think there is a way to solve my problem ?
--> Using my patch, simultaneously on 4 instances of Pd-extended using Jack on Ubuntu 10.04?

Thank you.

01ivier

_______________________________________________
Pd-list@iem.at mailing list
UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list




--
Pedro Lopes
contacto: jazz@radiozero.pt
website: http://web.ist.utl.pt/Pedro.Lopes



--
Pedro Lopes
contacto: jazz@radiozero.pt
website: http://web.ist.utl.pt/Pedro.Lopes




--
Pedro Lopes (ongoing MSc)
contact: pedro.lopes@ist.utl.pt
website: http://web.ist.utl.pt/Pedro.Lopes