[PD] [readanysf~] refuse 96kbps ogg's!?!

Roman Haefeli reduzent at gmail.com
Tue Sep 28 22:41:13 CEST 2010


On Tue, 2010-09-28 at 17:42 +0200, Raphael Raccuia wrote:
> Le mardi 28 septembre 2010 à 17:02 +0200, Roman Haefeli a écrit :
> > Hi
> > 
> > On Tue, 2010-09-28 at 16:49 +0200, Raphael Raccuia wrote:
> > > Seems to be done, as I updated gavl/gmerlin from rdz-ppa (thank's to
> > > you...)
> > > Now all my ogg's pass without any problem, but... Pd crashes on some wav
> > > and mp3 :-(
> > > I don't think I changed anything in my patch, seems to be this update: I
> > > isolated readanysf to try.
> > 
> > gmerlin-avdecoder and gavl are updated on my ppa, but pd-readanysf is
> > not yet. Soon to come. I'll let you know, when I am done.
> Yes, I noticed, no problem, but my version is 0.41 compiled... mp3 and
> wav used to work until today, and dsp/crash also since 0.41...
> Here is terminal output, lasts lines before crash:
> 
> [demuxer] Info: Detected WAV format
> opened /home/raphael/Documents/Studio/Pd/Multi-Player/bank1/46729__jk0666__ocean.wav
> socket receive error: Connection reset by peer (104)
> Erreur de segmentation

I can confirm this problem. All WAV files I tried crash as you describe.
If I interpret things correctly, this is not related to [readanysf~],
but to the libraries libgavl1 and libgmerlin-avdec1. At least the
following applies on my box:

readanysf~ 0.41		new gavl/gmerlin-avdecoder	crash
readanysf~ 0.40		new gavl/gmerlin-avdecoder	crash
readanysf~ 0.41		old gavl/gmerlin-avdecoder	OK
readanysf~ 0.40		old gavl/gmerlin-avdecoder	OK

old: cvs snapshot from 2010-03-09
new: cvs snapshot from 2010-09-27

@August
It could be that at the moment of the snapshot I used, the sources were
in some non-fully-functional state. Do you know if there is a revision
that is supposed to include the fixes that were mentioned in this thread
but nothing more?


> Well, I also updated pd-extended from repo yesterday, but don't think is
> related...

I did not and still the crashes are there.

Roman





More information about the Pd-list mailing list