GNOME Bugzilla – Bug 742028
Crazy waveforms on each iteration when doing multiple recordings one after another
Last modified: 2015-03-12 04:00:13 UTC
This is a bit hard to describe. The live waveforms seem to go nuts on subsequent recordings (big rhythmic spikes or just 100% plateau), but the recorded sound appears to be fine...
And here's the demonstration: http://jeff.ecchi.ca/public/gnome-sound-recorder-742028.webm
That's really interesting! I had no idea that was happening...I'll see if I can track that down tonight, and if I can't duplicate I'll ask you for some debugging help. Thanks :)
I can't duplicate this. Are there other things you notice that could help me to duplicate: e.g. does it happen more when you are using a particular codec? What type of mic is it happening with? Or anything else you think might be relevant. Thanks!
Also, is it possible for you to post the debug output for the level element (GST_DEBUG=level*:6 ./org.gnome.SoundRecorder) along with a video of the weird behavior? I doubt this is the level's fault but I want to rule that out. Thanks!
You sure about that command? GST_DEBUG=level*:6 ./org.gnome.SoundRecorder bash: ./org.gnome.SoundRecorder: No such file or directory Anyway, I found the trick: the issue does not occur with the Opus codec, but it occurs if you're using Vorbis, FLAC or MP3. I'm using a Logitech USB Desktop Microphone.
(In reply to comment #5) > You sure about that command? > > GST_DEBUG=level*:6 ./org.gnome.SoundRecorder > bash: ./org.gnome.SoundRecorder: No such file or directory Hmm, it works on my machine, but it needs to be run from the correct directory so maybe that's why you get the error :) on Fedora I think the full path should be /usr/share/gnome-sound-recorder/org.gnome.SoundRecorder. If it's not working for you then no worries, though! > > > Anyway, I found the trick: the issue does not occur with the Opus codec, but it > occurs if you're using Vorbis, FLAC or MP3. I'm using a Logitech USB Desktop > Microphone. Yeah, I can't duplicate it using any of the codecs. Next time I'm at the hackerspace I'll try it on other people's machines and see if it happens. That way I can look at the Cairo code as well :)
I wasn't able to duplicate this exact behavior, but I did find some issues during playback. I'll try to spend some time debugging it next weekend, but (without having reviewed my code yet) I think it may be related to the way I'm handling timing. The problem is that there is a difference between the way I handle that for recording vs. playback, so if you are seeing this during recording it migh be something else.
*** This bug has been marked as a duplicate of bug 746065 ***