quodlibet stops playing...?

[Rene Maurer]

quodlibet (V3) stops playing in my environment (very rarely but it
happens).

It happens when I play the queue or a playlist and it happens only at
the begin off a song (after "song changed").

I have no chance to reproduce it as it happens really rarely. Is this a
known issue (or am I the only with this behavior)?

Beside I have seen glitches which may go to the same direction:

1.
- Play a song in a playlist
- Use the scroll bar assigned to the seek button and go to the end of
the song (i.e. go to 0:00 if "Display remaining time" is selected")
- Don't release cursor within the scroll bar, directly move the cursor
outside the scroll bar and release.

=> The seek button doesn't show playing / remaining time any more.

2.
- Play a song in a playlist
- Use the scroll bar assigned to the seek button and go to the end of
the song (i.e. go to 0:00 if "Display remaining time" is selected")
- Release cursor within the scroll bar so that the next song is played

=> Repeat this until the next song *isn't* played or until the next
song starts playing with a *delay* of some seconds (or even minutes).


I am using ARCH Linux (other distros, other behavior?)

René

[Christoph Reiter]

1. is probably unrelated
for 2. see http://code.google.com/p/quodlibet/issues/detail?id=1177

[Rene Maurer]

Christoph Reiter <reiter.c...@gmail.com> wrote:

Thanks for answering Christoph.

> 1. is probably unrelated
I will do bug report.

> for 2. see http://code.google.com/p/quodlibet/issues/detail?id=1177

Uff, this makes quodlibet not usable at the moment for semi-professional
usage IMHO. What do you recommend?

I have tried to switch to the xine back end, but there are some quirks:
a) Silence (5 seconds) is not played, it is skipped instead. b) Going
back in the progress bar often starts playing the next song. c) Silence
at the and of song seams to be sipped as well.

Is there an example how the xinebe can/should be configured?

[Christoph Reiter]

2013/10/17 Rene Maurer <rene...@gmail.com>:

> Christoph Reiter <reiter.c...@gmail.com> wrote:
>
> Thanks for answering Christoph.
>
>> 1. is probably unrelated
> I will do bug report.

Thanks

>> for 2. see http://code.google.com/p/quodlibet/issues/detail?id=1177
> Uff, this makes quodlibet not usable at the moment for semi-professional
> usage IMHO. What do you recommend?
>
> I have tried to switch to the xine back end, but there are some quirks:
> a) Silence (5 seconds) is not played, it is skipped instead. b) Going
> back in the progress bar often starts playing the next song. c) Silence
> at the and of song seams to be sipped as well.

I could add a config option to disable gapless playback in the
playback prefs, which should fix it. Would that be OK? I still can't
reproduce the problem here...

> Is there an example how the xinebe can/should be configured?

No idea, I just make sure once in a while that the basics still work.

[Rene Maurer]

Christoph Reiter <reiter.c...@gmail.com> wrote:

> I could add a config option to disable gapless playback in the
> playback prefs, which should fix it. Would that be OK?

I think this is OK. But only if it makes sense to for the project to
have this configuration option anyway!

> I still can't reproduce the problem here...

Do you mean the "gstreamer bug"?

On my machine gstreamer reports like described in comment #6
for http://code.google.com/p/quodlibet/issues/detail?id=1177

... when track stops (and maybe restarts) playing I see:

0:10:43.318451554 922 0x7f9f14039e30 WARN audiobasesink
gstaudiobasesink.c:1580:gst_audio_base_sink_get_alignment:<alsasink1>
Unexpected discontinuity in audio timestamps of +0:08:31.512562358,
resyncing

[Christoph Reiter]

2013/10/17 Rene Maurer <rene...@gmail.com>:

> Christoph Reiter <reiter.c...@gmail.com> wrote:
>
>> I could add a config option to disable gapless playback in the
>> playback prefs, which should fix it. Would that be OK?
>
> I think this is OK. But only if it makes sense to for the project to
> have this configuration option anyway!

There shouldn't be such an option, but Gstreamer had problems with
gapless transitions on and off with various versions, so this might
"solve" it once and for all.

>> I still can't reproduce the problem here...
>
> Do you mean the "gstreamer bug"?

Yes.

> On my machine gstreamer reports like described in comment #6
> for http://code.google.com/p/quodlibet/issues/detail?id=1177
>
> ... when track stops (and maybe restarts) playing I see:
>
> 0:10:43.318451554 922 0x7f9f14039e30 WARN audiobasesink
> gstaudiobasesink.c:1580:gst_audio_base_sink_get_alignment:<alsasink1>
> Unexpected discontinuity in audio timestamps of +0:08:31.512562358,
> resyncing

Which GStreamer version btw? (see the QL about dialog for the version)

[Nicola Larosa]

Christoph Reiter wrote:
> Which GStreamer version btw? (see the QL about dialog for the
> version)

Same problem here with GStreamer v.1.0.6.0.

--
Nicola Larosa - http://www.tekNico.net/

[Anarchy] assumes that the burden of proof for anyone in a position
of power and authority lies on them. Their authority is not self-
justifying. They have to give a reason for it, a justification. And
if they can't justify that authority and power and control, which
is the usual case, then the authority ought to be dismantled and
replaced by something more free and just. - Noam Chomsky, May 2013

[Rene Maurer]

Christoph Reiter <reiter.c...@gmail.com> wrote:

> Which GStreamer version btw? (see the QL about dialog for the version)

Audio device: GStreamer (alsasink)
Mutagen: 1.22
GTK+: 3.10.2
PyGObject: 3.10.1
GStreamer: 1.2.0.0

[Christoph Reiter]

2013/10/18 Christoph Reiter <reiter.c...@gmail.com>

>
> 2013/10/17 Rene Maurer <rene...@gmail.com>:
> > Christoph Reiter <reiter.c...@gmail.com> wrote:
> >
> >> I could add a config option to disable gapless playback in the
> >> playback prefs, which should fix it. Would that be OK?
> >
> > I think this is OK. But only if it makes sense to for the project to
> > have this configuration option anyway!
>
> There shouldn't be such an option, but Gstreamer had problems with
> gapless transitions on and off with various versions, so this might
> "solve" it once and for all.

Added on trunk in revision fd7d38ea51e4 [0]

[0] http://code.google.com/p/quodlibet/source/detail?r=fd7d38ea51e42d24020cf63e759b6002da16c943

[Rene Maurer]

Thanks Christoph. It works AFAICS.

I have played with it yesterday and it worked all the time (using a
Lindeman USB-DDC). Whenever I have enabled gapless transition
stop/start playing on song changes was there within minutes.

I will try the the "disabled gapless transition" with a VLink USB/SPDIF
converter this weekend.

René