Bugzilla – Bug 4391
Softsqueeze cannot play certain FLAC files
Last modified: 2008-12-18 11:12:53 UTC
This is with SlimServer v6.5.1 - 10386 - Windows XP - EN - cp1252; Perl v5.8.8 MSWin32-x86-multi-thread; MySQL v5.0.22-community-nt. I have some FLAC files that won't play using Softsqueeze v3.0.1. After pressing play, the Softsqueeze display show the track info and elapsed time, but no audio appears and the elapsed time stays at 00:00. The Softsqueeze GUI tends to become unresponsive. This doesn't affect all FLAC, only some. I've tried JLayer/MP3 Plugin, and that doesn't make any difference. They files play okay in Foobar2000 v0.9.4.1. Strangely, forcing SlimServer to transcode to WAV or MP3 doesn't make any different. The FLACs still don't play. I've tried a few decoding/encoding steps to try and find the cause: Original FLAC - Does not play Original FLAC > Remove comments with FB2K - Does not play Original FLAC > Convert to WAV with FB2K > Convert to FLAC with FLAC.EXE - Plays OK Original FLAC > Convert to WAV with FB2K > Convert to FLAC with FB2K - Plays OK Original FLAC > Convert to WAV with FB2K > Convert to FLAC with FB2K > Add comments with FB2K - Does not play Original FLAC > Convert to WAV with FB2K > Convert to FLAC with FB2K > Add different comments with FB2K - Does not play Original FLAC > Convert to WAV with FB2K > Convert to FLAC with FLAC.EXE > Add comments with FB2K - Does not play Original FLAC > Convert to WAV with FB2K > Convert to FLAC with FLAC.EXE & Add comments with METAFLAC - Does not play It just looks like any FLAC file made from the problem one won't play. Thanks.
Created attachment 1651 [details] Debug output from Softsqueeze v3.0.1 when configured to use Java MP3 Plugin
Created attachment 1652 [details] Debug output from Softsqueeze v3.0.1 when configured to use JLayer
Created attachment 1653 [details] Original problem FLAC (26.4 MB)
Do your Squeezeboxes play this FLAC okay, Nigel?
Yes, the SB2s I have play it fine.
Richard please let me know if there's anything additional I can or should do to help with this one.
Looks like it works if we upgrade to jflac 1.2
so, are we calling this fixed at change 15211 then?
With change 14762 actually.
Woops, wrong bug. No I want to wait until I have an "official" SoftSqueeze build to test it with. That is likely to take a while, what with other priorities. Anyone prepared to hack up their own SoftSqueeze from change 15211 is welcome to test it too.
Fixed with SQ 3.6
That's great. I'll test it once a Win32 build of v3.6 is available (on SourceForge?).
I've re-opened this bug, but feel free to tell me to create a new one. When playing the comment 3 FLAC with no volume adjustment, the file plays fine with Softsqueeze v3.6. I do get a "0 AudioDecoder-0] WARN javasound.verbose - audioDecoder did not write (0) the same number of bytes as read (16384)" error - does that matter? However, if I set the volume adjustment for the Softsqueeze player to "Smart gain" (my normal setting), the file doesn't play and a console error is displayed: 8 [AudioMixer-0] WARN javasound - audio mixer exception java.lang.IllegalArgumentException: Requested value 17.360008 exceeds allowable maximum value 6.0206. at javax.sound.sampled.FloatControl.setValue(Unknown Source) at com.sun.media.sound.DirectAudioDevice$DirectDL$Gain.setValue(Unknown Source) at org.titmuss.softsqueeze.audio.AudioMixer.setVolume(AudioMixer.java:198) at org.titmuss.softsqueeze.audio.AudioMixer.bufferEvent(AudioMixer.java:639) at org.titmuss.softsqueeze.audio.AudioBuffer.read(AudioBuffer.java:445) at org.titmuss.softsqueeze.audio.AudioMixer.playSamples(AudioMixer.java:488) at org.titmuss.softsqueeze.audio.AudioMixer.run(AudioMixer.java:421) at java.lang.Thread.run(Unknown Source) The track does have large positive RG values (+20.32/+0.105469 for track/peak and + 17.36/+0.202728 for album/peak). Is the +6.0206 dB gain a sanity limit to avoid clipping? If so, should Softsqueeze just use the +6.0206 dB maximum gain when the gain is larger than this? I wonder if this is at all related to bug 5119. Bug 5119 talks about tracks with positive gain causing clipping, and the need to take the peak again to to account. Would doing that help this problem? Thanks.
Yes, please file this as a separate bug, as this is quite a different issue.
I've opened bug 6850 for comment 13. Thanks for fixing the orginal problem.
This bug is being closed since it was resolved for a version which is now released! Please download the new version of SqueezeCenter (formerly SlimServer) at http://www.slimdevices.com/su_downloads.html If you are still seeing this bug, please re-open it and we will consider it for a future release.