gueygel writes

When a track ends, upplay takes up to 9 minutes to play the next track in the playlist. If I double click it, it starts instantly. I’ve also observed that, sometimes, when the next track starts playing, upplay still shows as if the one sounding were the previous one. Then, when it ends, it repeats the same track but this time upplay shows the correct one. I wonder whether these issues are related. I encountered a similar syncing problem using an old version of BubbleUPNP on my phone (a control point app for Android), but eventually it got fixed.

I’m using Ubuntu 15.04 + rygel as media server, upplay installed on the same system, and a pair of DLNA-enabled speakers (KEF X300A Wireless).

Above all, thanks a lot for the good work! I’ve been looking for something like this on linux for a long time, besides the rudimentary gupnp…

DoomHammer writes

Might be related to poorly implemented MediaRenderer. I was having similar issues when testing my Renderer before it got properly implemented ;)

medoc92 writes

Hi,

This is probably some incompatibility between upplay and the KEF renderer. To investigate, it would be useful to enable the libupnp log file and run upplay. You can do it by starting it in the following way on the command line:

 UPPLAY_UPNPLOGFILENAME=/tmp/somefile UPPLAY_UPNPLOGLEVEL=6 upplay

Then you need a way to mark the point in the log file where things stop (then start). Given the delay, you should have enough time to just read the file in an editor and get the current end line number. Then attach the trace, or, if it’s too big we’ll find another way, and tell me the line number where it stopped. Hopefully I’ll be able to see what’s happening.

The strange thing actually is why it starts again at all after all this time !

It would also be useful if you could cut/paste a number of console (upplay output) lines before and after the gap, to complement the lib output.

gueygel writes

Hi, thanks a lot for your prompt reply!

I wasn’t able to do the testing until now. Actually, this time the next track didn’t start at all… Upplay interface was marking the second track in play mode but the slider was stuck at 00:00 for more than 20 minutes. Then I just quit the program.

I attach the upplay console output starting from when the first track ends. The log file is pretty big. The same moment occurs around line 53777.

Hope this helps!

medoc92 writes

Hi,

As far as I can see in the log, the Renderer accepts the "nextAvTransportURI" command for transitionning to the next track, and then just does nothing (does not even signal that it’s stopped).

I’d suggest checking for a firmware upgrade.

Also, it would be interesting to try with another "known good" controller, like Bubble UPNP from Android for example, there is a free version.

Not exactly the same because the report is for a list of flac tracks and says that things worked fine with mp3. Still, the firmware update appears to be the first thing to try.

gueygel writes

OK, although I’m not very optimistic about waiting for firmware upgrades, I’ll try and contact KEF about it.

Now that I remember, they have their own Android control point app, "KEF DMC App". Never used it because Bubble seemed much more polished but I’ll give it a go.

Thanks for your help!

medoc92 writes

Did you actually try with Bubble ? The other guy seemed to have some success at least with mp3 files.

gueygel writes

I did use Bubble for a while and wasn’t having this issue (as I mentioned in the first post), but it failed the other day when I was trying upplay in the same manner. I didn’t check whether the files I used for testing were flac or mp3 though… Will try again.

medoc92 writes

Sorry, I had not noted that Bubble ended up working last time (bad memory + several things at the same time - > should reread from start each time…)

Raspberry PI + Hifi-Berry + alim/box < 100 $ ---- > aux input

:)

gueygel writes

No worries mate ;) Yeah I’ve thought about the PI more than once… I haven’t given up yet because it’s so frustrating to have DLNA-enabled speakers that can’t function wirelessly!

medoc92 writes

Nothing I can do at this point I’m afraid…