Two issues:
First at all, when trying my patch (MEH-SYSTEM) with -nogui option, I was having a "Segment fault"... untill I discovered that the "Robot" FX, from RJDJ was turning DSP on.... and DSP was already on (was turned on with a [loadbang]-[delay 2000]). This crashed Pd. Maybe a bug...?
Attempting to save some CPU (-nogui option), I started using the patch and recorded all the sample banks... then, a lot of noisy clicks started to appear. If I mute all the banks, the clicks are there, at low volume but very audible.
Then, I realized that that noise is there only with -nogui option. Using GUI, it sounds great. Is this a known issue? Because I expected to have the opposite result...
I'm using Pd-Extended 0.43.4, Ubuntu 12.04, Jackd (Jack2?) with 256frames, 2periods, 48000, PulseAudio uninstalled, with a USB soundcard (Encore ENMAB-8CM). It does the same noise with the integrated soundcard.
Here is the WIP (it is 0.7 version... but almost 1.0): http://puredata.hurleur.com/viewtopic.php?pid=37430
If I can give you any other information, just ask for it.
Hi Mario,
I think you might be being bitten by this bug, though I'm not 100% sure.
On 15/08/13 05:22, Mario Mey wrote:
Then, I realized that that noise is there only with -nogui option. Using GUI, it sounds great. Is this a known issue?
Some versions of Pd have a bug where the sample rate isn't set correctly during initialisation before loading patches. This can cause all kinds of strange problems (delay lines having the wrong length, externals crashing...).
A workaround is to load your patches using a loader patch like [loadbang]-[delay 0]-[;pd -open your-real-patch.pd path/to/patch( The attached onload.pd provides a more general approach, but needs at least pd-0.39 (December 2005) to work properly.
As you can see in this test log, Miller's pd 0.41, 0.42, 0.43 have problems - earlier and later versions seem fine. I couldn't test any pd-extended versions (someone with a few versions installed wants to try?) but pd-l2ork git master is buggy too.
$ ./samplerate~-test.sh ~/opt/pd/*/bin/pd Pd version 0.38.4 compiled 08:16:18 Aug 15 2013 SR: 44100 list prepend add ... couldn't create error: inlet: expected '' but got 'pd' ... you might be able to track this down from the Find menu. error: inlet: expected '' but got 'pd' couldn't open MIDI input device 0 couldn't open MIDI output device 0 opened 0 MIDI input device(s) and 0 MIDI output device(s). Pd version 0.39-2 compiled 08:16:43 Aug 15 2013 SR: 44100 SR: 44100 Pd version 0.40-3 compiled 08:17:08 Aug 15 2013 SR: 44100 SR: 44100 Pd version 0.41-4 compiled 08:17:23 Aug 15 2013 SR: 0 SR: 44100 Pd version 0.42-5 compiled 08:17:45 Aug 15 2013 SR: 0 SR: 44100 Pd-0.43.4 ("") compiled 08:18:09 Aug 15 2013 SR: 0 SR: 44100 Pd-0.44.3 ("") compiled 08:20:19 Aug 15 2013 SR: 44100 SR: 44100
Claude
On 15/08/13 05:22, Mario Mey wrote:
I'm using Pd-Extended 0.43.4
Nevermind what I said in my earlier reply, this version doesn't have that particular bug (but 0.41 and 0.42 extended do, and so does pd-l2ork git master as of today):
Pd version 0.41-4extended compiled 17:56:28 Aug 15 2013 SR: 0 SR: 44100
Pd version 0.42-5extended compiled 13:25:06 Aug 15 2013 SR: 0 SR: 44100
Pd-0.43.4 ("extended") compiled 12:23:46 Aug 15 2013 SR: 44100 SR: 44100
Pd-l2ork version 20130724 compiled 18:10:00 Aug 15 2013 SR: 0 SR: 44100
Claude
Thanks for the response, Claude.
I don't understand what "SR" is. Why 0 and 44100?
Following Maelstorm suggestion, I'm using 48000. He says the soundcard "clocks better when you set the sample rate to something that easily divides by milliseconds, and it seems to be true. I get less xruns when using 48000 than 44100"
El 15/08/13 14:11, Claude Heiland-Allen escribió:
On 15/08/13 05:22, Mario Mey wrote:
I'm using Pd-Extended 0.43.4
Nevermind what I said in my earlier reply, this version doesn't have that particular bug (but 0.41 and 0.42 extended do, and so does pd-l2ork git master as of today):
Pd version 0.41-4extended compiled 17:56:28 Aug 15 2013 SR: 0 SR: 44100
Pd version 0.42-5extended compiled 13:25:06 Aug 15 2013 SR: 0 SR: 44100
Pd-0.43.4 ("extended") compiled 12:23:46 Aug 15 2013 SR: 44100 SR: 44100
Pd-l2ork version 20130724 compiled 18:10:00 Aug 15 2013 SR: 0 SR: 44100
Claude
Hi,
On 15/08/13 20:35, Mario Mey wrote:
I don't understand what "SR" is. Why 0 and 44100?
Ah sorry for being unclear - the first one is the output of [samplerate~] when a patch is loaded from the command line, the second is the outut when the patch is loaded after a [delay]. See the attachments in my other reply:
http://lists.puredata.info/pipermail/pd-list/2013-08/103730.html
If it's ever 0, that might cause problems.
Anyway, this bug is fixed in the latest pd and pd-extended (which you are using), sorry for the noise.
El 15/08/13 14:11, Claude Heiland-Allen escribió:
On 15/08/13 05:22, Mario Mey wrote:
I'm using Pd-Extended 0.43.4
Nevermind what I said in my earlier reply, this version doesn't have that particular bug (but 0.41 and 0.42 extended do, and so does pd-l2ork git master as of today):
Pd version 0.41-4extended compiled 17:56:28 Aug 15 2013 SR: 0 SR: 44100
Pd version 0.42-5extended compiled 13:25:06 Aug 15 2013 SR: 0 SR: 44100
Pd-0.43.4 ("extended") compiled 12:23:46 Aug 15 2013 SR: 44100 SR: 44100
Pd-l2ork version 20130724 compiled 18:10:00 Aug 15 2013 SR: 0 SR: 44100
Claude
Sorry for re-floating this thread... but only Claude answer it and he was talking about previous version of Pd, so, it doesn't apply in my Pd-Extende version. I paste again the mail:
-------- Mensaje original -------- Asunto: Worse performance with -nogui. Fecha: Thu, 15 Aug 2013 01:22:36 -0300 De: Mario Mey mariomey@gmail.com Para: pd-list pd-list@iem.at
Two issues:
First at all, when trying my patch (MEH-SYSTEM) with -nogui option, I was having a "Segment fault"... untill I discovered that the "Robot" FX, from RJDJ was turning DSP on.... and DSP was already on (was turned on with a [loadbang]-[delay 2000]). This crashed Pd. Maybe a bug...?
Attempting to save some CPU (-nogui option), I started using the patch and recorded all the sample banks... then, a lot of noisy clicks started to appear. If I mute all the banks, the clicks are there, at low volume but very audible.
Then, I realized that that noise is there only with -nogui option. Using GUI, it sounds great. Is this a known issue? Because I expected to have the opposite result...
I'm using Pd-Extended 0.43.4, Ubuntu 12.04, Jackd (Jack2?) with 256frames, 2periods, 48000, PulseAudio uninstalled, with a USB soundcard (Encore ENMAB-8CM). It does the same noise with the integrated soundcard.
Here is the WIP (last version: 0.91): http://puredata.hurleur.com/viewtopic.php?pid=37430
If I can give you any other information, just ask for it.
EDIT: I'm trying to get no noise, because I'm trying to use 128 frames in Jack, instead of 256, to get less latency.
-
Claude Heiland-Allen -
Mario Mey