MKV Dolby Digital - center speaker swapped with surround

[khaindar]

Hello,

first of all thanks a lot for this great software. Everything is just working fine, except one weird thing. When i play mkv's with DD5.1 the voice only comes out of the right surround speaker. With DTS it's working fine. Are there any more sound options i can try to adjust? The only thing i see is the numbers of speakers.

PS3 Media Server 1.4 on Ubuntu 8.10

Thx for the help and Greets

Khaindar

[eetole]

Hi,

I have exactly same problem with Ubuntu 8.04. I have also Vista installed and with that everything works OK. I am using version 1.04 with both OS.

I seems that this is somehow OS related problem. I don't know how it works with OSX.

[Topguy]

I think this is caused by mplayer/mencoder package in ubuntu is too old. Search the forum a little more and you might find a solution.

[az1979]

I have the exact same problem. mplayer version is 1.0rc2, should not be a problem. I transcoded AC3 before, mplayer is capable of passthru, which should do the trick entirely.

Most NAS systems will run on linux, so this is kind of important to work I checked the changelog from rc2 to rc3 (current mplayer). There's nothing interesting in there which could cause that kind of behavior.

I will now try the new mplayer package and report back if this was really the reason. To me, this seems more like a very simple mistake in commandline parsing. But we'll see.

Shagrath: Great work, btw. Me and my friend were planning on writing something similar, because we were very annoyed of twonkymedia, mediatomb and all those problems that come with those. You included lots of features that we didn't even think of yet. Great job, it really seems like you thought this thing through before writing it.

edit: installed rc2 from ppa.launchpad.net - no success... And even if it were mplayers fault, encoding audio with ffmpeg should work, shouldn't it?

[az1979]

Hey everybody,

I was tired of waiting, so I fixed it myself. Just replace the file net/pms/dlna/DLNAMediaAudio.java with this one: http://www.lux-medien.com/DLNAMediaAudio.java and rebuild the jar file, as described in the Howto.

Or just download the rebuilt pms.jar from http://www.lux-medien.com/pms.jar and replace your existing one with this one.

For those interested: The problem was, that PMS rebuilt AC3 Audio and somehow messed up the channels. AC3 audio should just be remuxed without re-encoding, but for that it needs to identify AC3 files correctly. So, this version should work properly once you checked the "remux audio" box.

Shagrath: Identify AC3 files by codecA ac3, a52 AND "liba52" did the trick

Cheers.

Andreas

[krister]

I'm having this issue with movies that are encoded with DTS, any ideas how to fix that?

- Krister

[az1979]

I'm having this issue with movies that are encoded with DTS, any ideas how to fix that?

- Krister

Yes, but just an idea Try setting tsMuxer (Audio High Fidelity) as first application to be used for transcoding audio.

mplayer can also do DTS/AC3 re-encoding, but obviously not the way we want it to. If there's no way to figure this out, I will try and fix the mencoder commandline so it doesn't mix up the channels. I can't really figure out what the problem is, because the mencoder line that is being executed looks pretty okay to me.

I'll look into it. Do you have any DTS sample file you could upload? I will check against it and try to make it work.

Andreas

[az1979]

Tried a few things now.

My receiver is DTS capable, so I can only speak for myself. When I tell PMS to keep DTS inside the stream, everything works fine. The receiver recognizes the DTS stream without any problems.

There was just a lot of static when I switched that feature off. I was too lazy to test it, but here's something for you, that you might wanna try. Again: UNTESTED

In line 267 of TSMuxerVideo.java is an else clause.

replace the three lines from "else" to the next "if" with the following four lines and rebuild.

else {
boolean lSixChannelDTS = false;
lSixChannelDTS = (configuration.getAudioChannelCount()==6 && params.aid.isDTS());
ffmpegLPCMextract = new String [] { configuration.getMencoderPath(), "-ss", "0", fileName, "-quiet", "-quiet", "-really-quiet", "-msglevel", "statusline=-1:mencoder=-1", "-channels", "" + CodecUtil.getAC3ChannelCount(configuration, params.aid), "-ovc", "copy", "-of", "rawaudio", "-mc", "0", "-noskip", "-oac", (params.aid.isAC3() && configuration.isRemuxAC3())?"copy":"lavc", params.aid.isAC3()?"-fafmttag":"-quiet", params.aid.isAC3()?"0x2000":"-quiet", "-lavcopts", "acodec=ac3:abitrate="+CodecUtil.getAC3Bitrate(configuration, params.aid), "-af", "lavcresample=48000",lSixChannelDTS?"channels=6:6:0:4:1:0:2:1:3:2:4:3:5:5":"", "-srate", "48000", singleMediaAudio?"-quiet":"-aid", singleMediaAudio?"-quiet":("" + params.aid.id), "-o", ffAudioPipe[0].getInputPipe() };}

Note that ffmpegLPCMblabla is ONE line, no matter what the forum does to it

After rebuilding, you'll see if DTS will be remuxed accordingly. This one just corrects the channelmapping manually. As I said, for me, remuxing AC3 and passing DTS through are working perfectly.

Just give it a shot.

Andreas

[krister]

woha.. now that looks complex

I'll give it a shot on sunday, i got around the problem with switching to two speaker mode. For some reason this worked for my surround system and it started to output sound the right way. The only problem i got from this was the low-sound-getting-cut-off-to-nothing-problem. That is when the audio in the moive is really low you can hear that the sound is getting turned off, and as soon as the sound starts picking upp it gets turned back on again.

ah well.. i'll first try your rebuild of the jar file.

Krister

[SpEnTBoY]

Ok guys I need some help .I'm experiencing the exact same thing (center swapped with surround speakers) and I've tried some of the suggestions including replacing the .jar file but still with no luck. Here's the details on the file:

(this is from mkvinfo)

Code: Select all

+ EBML head
|+ Doc type: matroska
|+ Doc type version: 1
|+ Doc type read version: 1
+ Segment, size 4695588004
|+ Seek head (subentries will be skipped)
|+ EbmlVoid (size: 4025)
|+ Segment information
| + Timecode scale: 1000000
| + Muxing application: libebml v0.7.7 + libmatroska v0.8.1
| + Writing application: mkvmerge v2.4.1 ('Use Me') built on Dec  5 2008 18:30:05
| + Duration: 6178.635s (01:42:58.635000000)
| + Date: Wed Jan 14 18:17:06 2009 UTC
| + Segment UID: 0xa5 0x1a 0x2b 0xcf 0x0f 0x27 0x88 0x79 0xbf 0x61 0xe7 0xa0 0x70 0x73 0x77 0xdd
|+ Segment tracks
| + A track
|  + Track number: 1
|  + Track UID: 1
|  + Track type: video
|  + Enabled: 1
|  + Default flag: 1
|  + Forced flag: 0
|  + Lacing flag: 0
|  + MinCache: 1
|  + Timecode scale: 1.000000
|  + Max BlockAddition ID: 0
|  + Codec ID: V_MPEG4/ISO/AVC
|  + Codec decode all: 1
|  + CodecPrivate, length 41
|  + Default duration: 41.708ms (23.976 fps for a video track)
|  + Language: eng
|  + Video track
|   + Pixel width: 1280
|   + Pixel height: 536
|   + Interlaced: 0
|   + Display width: 160
|   + Display height: 67
| + A track
|  + Track number: 2
|  + Track UID: 3820497308
|  + Track type: audio
|  + Enabled: 1
|  + Default flag: 1
|  + Forced flag: 0
|  + Lacing flag: 1
|  + MinCache: 0
|  + Timecode scale: 1.000000
|  + Max BlockAddition ID: 0
|  + Codec ID: A_DTS
|  + Codec decode all: 1
|  + Language: und
|  + Audio track
|   + Sampling frequency: 48000.000000
|   + Channels: 6
|+ EbmlVoid (size: 1024)
|+ Cluster


This was me running mpayer from command line with no X just to get some output:

Code: Select all

CPU: AMD Athlon(tm) Processor LE-1640 (Family: 15, Model: 127, Stepping: 2)
CPUflags:  MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 1 SSE2: 1
Compiled with runtime CPU detection.
[mkv] Track ID 1: video (V_MPEG4/ISO/AVC), -vid 0
[mkv] Track ID 2: audio (A_DTS), -aid 0, -alang und
[mkv] Will play video track 1.
Matroska file format detected.
VIDEO:  [avc1]  1280x536  24bpp  23.976 fps    0.0 kbps ( 0.0 kbyte/s)
==========================================================================
Forced audio codec: mad
Opening audio decoder: [libdca] DTS decoding with libdca
Stream with high frequencies VQ coding
AUDIO: 48000 Hz, 2 ch, s16le, 1536.0 kbit/100.00% (ratio: 192000->192000)
Selected audio codec: [dts] afm: libdca (DTS-libdca)
==========================================================================
AO: [pulse] Failed to connect to server: Connection refused
AO: [alsa] 48000Hz 2ch s16le (2 bytes per sample)


Can anyone think of other options I may have to change to get this working? I'm currently transitioning form mediatomb to pms and I love the config etc so far fr pms ... I just can't get it working properly yet lol