PDA

View Full Version : should synchronisation with softsqueeze work?



bob findlay (JIC)
2005-01-25, 03:04
I tried to synchronise my squeezebox with my PC running soft squeeze at
the weekend. At best, the PC was lagging by about half a second.

Is this expected? is synchronisation between two squeezeboxes any
better? is there any way I can get softsqueeze to better synchronise?

I was running softsqueeze on the slim server PC. Does that make a
difference?

Cheers,

Bob

rtitmuss
2005-01-25, 14:07
bob findlay (JIC) wrote:

>I tried to synchronise my squeezebox with my PC running soft squeeze at
>the weekend. At best, the PC was lagging by about half a second.
>
>Is this expected? is synchronisation between two squeezeboxes any
>better? is there any way I can get softsqueeze to better synchronise?
>
>I was running softsqueeze on the slim server PC. Does that make a
>difference?
>
>
Bob,

First check out the Softsqueeze FAQ:
http://softsqueeze.sourceforge.net/faq.html#sync

I still have some reports of sync not working correctly, but by your the
description it sounds like you are not using Java 1.5 and the primary
sound driver. When it is working correctly you should not be able to
hear any time difference. And yes, sync between two squeezeboxes is very
good (otherwise I would not have spent many hours getting it to work
correctly with Softsqueeze!).

Regards,
Richard

Joshua Uziel
2005-01-25, 14:40
* Richard Titmuss <richard_titmuss (AT) yahoo (DOT) co.uk> [050125 13:05]:
> I still have some reports of sync not working correctly, but by your
> the description it sounds like you are not using Java 1.5 and the
> primary sound driver. When it is working correctly you should not be
> able to hear any time difference. And yes, sync between two
> squeezeboxes is very good (otherwise I would not have spent many hours
> getting it to work correctly with Softsqueeze!).

I have two Linux machines I've attempted to sync with a squeezebox. On
one, the sync is on towards the beginning of the song and it goes out of
sync as the song goes on. On the other, it stays in sync nicely but is
prone to "ERROR javasound - LineUnavailable" and "ERROR player - Audio
player exception" (log is attached). Even though they're both running
2.6.x kernels with their respective ALSA audio drivers, my guess is that
the differing behavior is due to either the sound chip or driver. This
is with Java 1.5 on both of them.

rtitmuss
2005-01-25, 15:47
Joshua Uziel wrote:

>* Richard Titmuss <richard_titmuss (AT) yahoo (DOT) co.uk> [050125 13:05]:
>
>
>>I still have some reports of sync not working correctly, but by your
>>the description it sounds like you are not using Java 1.5 and the
>>primary sound driver. When it is working correctly you should not be
>>able to hear any time difference. And yes, sync between two
>>squeezeboxes is very good (otherwise I would not have spent many hours
>>getting it to work correctly with Softsqueeze!).
>>
>>
>
>I have two Linux machines I've attempted to sync with a squeezebox. On
>one, the sync is on towards the beginning of the song and it goes out of
>sync as the song goes on. On the other, it stays in sync nicely but is
>prone to "ERROR javasound - LineUnavailable" and "ERROR player - Audio
>player exception" (log is attached). Even though they're both running
>2.6.x kernels with their respective ALSA audio drivers, my guess is that
>the differing behavior is due to either the sound chip or driver. This
>is with Java 1.5 on both of them.
>
>
On the first machine is the track playing at the wrong speed, or do you
get small gaps in the playback that cause it to lose sync? It would also
be worth checking the cpu on this machine during playback, if the cpu is
maxed out this could cause small gaps.

On the second machine could you enable the Java Sound debug and capture
another log. I guess the LineUnavailable, etc. errors occur between
tracks. Does this stop playback?

It will also be helpful to know what the sound cards and ALSA drivers
you are using.

Regards,
Richard

Joshua Uziel
2005-01-25, 17:28
* Richard Titmuss <richard_titmuss (AT) yahoo (DOT) co.uk> [050125 14:44]:
> On the first machine is the track playing at the wrong speed, or do you
> get small gaps in the playback that cause it to lose sync? It would also
> be worth checking the cpu on this machine during playback, if the cpu is
> maxed out this could cause small gaps.

I've not noticed any gaps... one difference I failed to mention is that
this system is wired while system two was tested wireless. You can
throw the cpu speed out the window since both systems are suitably
powered... the first one has an Athlon 2200+ and system two has a
Pentium M 1.7GHz. My guess is that the track is playing at the wrong
speed by a pinch... towards the end of a ~4 minute song, I'd say it gets
out of sync by about a second.

> On the second machine could you enable the Java Sound debug and capture
> another log. I guess the LineUnavailable, etc. errors occur between
> tracks. Does this stop playback?

Sure, I'll try it again when I get home tonight... it does happen
between tracks, but this was with quick testing so I only noticed it
when I was skipping from track to track rather than it moving on from
one track to the next by itself.

> It will also be helpful to know what the sound cards and ALSA drivers
> you are using.

System one identifies it's audio chip as a

"Multimedia audio controller: VIA Technologies, Inc. VT8233/A/8235/8237
AC97 Audio Controller"

and uses the "snd_via82xx" ALSA audio driver.

System two has a

"Multimedia audio controller: Intel Corp. 82801DB/DBL/DBM
(ICH4/ICH4-L/ICH4-M) AC'97 Audio Controller"

and uses the "snd_intel8x0" ALSA audio driver. Let me know if you want
more detailed info like PCI ids, etc.

rtitmuss
2005-01-31, 13:48
On 26 Jan 2005, at 00:28, Joshua Uziel wrote:

> * Richard Titmuss <richard_titmuss (AT) yahoo (DOT) co.uk> [050125 14:44]:
>> On the first machine is the track playing at the wrong speed, or do
>> you
>> get small gaps in the playback that cause it to lose sync? It would
>> also
>> be worth checking the cpu on this machine during playback, if the cpu
>> is
>> maxed out this could cause small gaps.
>
> I've not noticed any gaps... one difference I failed to mention is that
> this system is wired while system two was tested wireless. You can
> throw the cpu speed out the window since both systems are suitably
> powered... the first one has an Athlon 2200+ and system two has a
> Pentium M 1.7GHz. My guess is that the track is playing at the wrong
> speed by a pinch... towards the end of a ~4 minute song, I'd say it
> gets
> out of sync by about a second.

Not sure I can fix this one. Do other music applications play at the
correct speed on this machine?

>
>> On the second machine could you enable the Java Sound debug and
>> capture
>> another log. I guess the LineUnavailable, etc. errors occur between
>> tracks. Does this stop playback?
>
> Sure, I'll try it again when I get home tonight... it does happen
> between tracks, but this was with quick testing so I only noticed it
> when I was skipping from track to track rather than it moving on from
> one track to the next by itself.

I think I understand what's happening here, it should only be a problem
when skipping between tracks and not when allowing it to play normally.
I will be making some more improvements to the audio layer in
Softsqueeze soon, and will fix this then.

Regards,
Richard

Joshua Uziel
2005-01-31, 14:43
* Richard Titmuss <richard_titmuss (AT) yahoo (DOT) co.uk> [050131 12:49]:
> Not sure I can fix this one. Do other music applications play at the
> correct speed on this machine?

As far as I can tell, it's playing at the correct speed with
SoftSqueeze. If it isn't, then it's so close that I probably won't
notice it.

> I think I understand what's happening here, it should only be a
> problem when skipping between tracks and not when allowing it to play
> normally. I will be making some more improvements to the audio layer
> in Softsqueeze soon, and will fix this then.

Yeah, I never got you that log dump, which I can do if you still want
it. Either way, I'll be testing with both machines when you throw
another release over the wall.