View Full Version : eac3to - audio conversion tool
komisar
5th December 2008, 13:34
Ryu77, try:
md c:\as
copy "%CommonProgramFiles%\ArcSoft\Bin\*.*" c:\as
copy "%CommonProgramFiles%\ArcSoft\MPEG Engine\*.*" c:\as
and add to you PATH environment variable c:\as
Ryu77
5th December 2008, 13:42
Ryu77, try:
md c:\as
copy "%CommonProgramFiles%\ArcSoft\Bin\*.*" c:\as
copy "%CommonProgramFiles%\ArcSoft\MPEG Engine\*.*" c:\as
and add to you PATH environment variable c:\as
Ok, I've added those two lines "%CommonProgramFiles%\ArcSoft\Bin\*.*" & "%CommonProgramFiles%\ArcSoft\MPEG Engine\*.*" without the quotation marks... I am not sure what you mean by adding c:\as?
komisar
5th December 2008, 13:51
Ryu77, open cmd.exe and enter this 3 command. Go to "MyComputer->Properties->Advanced->Environment Variables" and add in "User variables" variable Path with value c:\as. If you already have Path variable, add to end of value ;c:\as
Ryu77
5th December 2008, 14:22
Thank you for your help komisar... But still no luck.
I have added the C:\Program Files\Common Files\ArcSoft\Bin and C:\as to the path in environment variables and I have run those command prompts you suggested but it's still not recognising the DTS decoder.
I'm starting to feel like I was better off with a "trial" version.
komisar
5th December 2008, 14:34
Ryu77, one more explanation. You need all files from ArcSoft\Bin and ArcSoft\MPEG Engine in ONE folder (in my example c:\as). Also you need point Path to this folder (in my example c:\as).
P.S. I also spent a lot of time searching for the decision. And now this work for me...
Ryu77
5th December 2008, 14:57
Ryu77, one more explanation. You need all files from ArcSoft\Bin and ArcSoft\MPEG Engine in ONE folder (in my example c:\as). Also you need point Path to this folder (in my example c:\as).
P.S. I also spent a lot of time searching for the decision. And now this work for me...
I do appreciate your persistence with helping me but nothing seems to be working.
I really can't understand this. I checked my other PC that has the DTS decoder working and I can see that C:\Program Files\Common Files\ArcSoft\Bin has been already added to the environment variables. This must have been done automatically with the older trial install. I guess newer retail versions don't do that anymore. However, one would think manually adding it in as we have done would rectify this... In my case this doesn't seem to work. :-(
asarian
5th December 2008, 15:24
Thank you for your help komisar... But still no luck.
I have added the C:\Program Files\Common Files\ArcSoft\Bin and C:\as to the path in environment variables and I have run those command prompts you suggested but it's still not recognising the DTS decoder.
I'm starting to feel like I was better off with a "trial" version.
TotalMedia Theatre, in the commercial variant of TotalMedia Extreme (which I have purchased too), is at:
C:\Program Files\ArcSoft\TotalMedia Extreme\Digital Theatre
You don't need to add anything to your path; though I'd add eac3to to your path, like: "C:\Program Files\eac3to" (or wherever it resides on your system)
Ryu77
5th December 2008, 15:36
Thank you asarian but I purchased TotalMedia Theatre not the Extreme suite.
I fixed my problem now...
This is what I did...
1) Uninstalled full retail version, then did a restart.
2) Installed trial version, tested eac3to and it worked, then restarted.
3) Tested eac3to again, still good. Ran full retail install again but chose "repair".
It's all good now. The trial (or older versions) must do something that the newer full version doesn't. Oh well, it's all working now... Yay. :-D
For anyone that's interested in purchasing ArcSoft TotalMedia Theatre or TotalMedia Extreme, the code that I used to get 40% off is "earlyxmas09". Just type this into the field marked "Please enter your coupon code here:" and you will receive 40% off. :D
Thunderbolt8
5th December 2008, 15:53
got an Arcsoft error when decoding a movie's DTS-HD MA track to flac:
eac3to v2.79
command line: eac3to x:\path 1) 4: x:\test.flac
------------------------------------------------------------------------------
M2TS, 1 video track, 2 audio tracks, 6 subtitle tracks, 3:08:47
1: Chapters, 10 chapters
2: h264/AVC, 1080p24 (16:9)
3: RAW/PCM, Swedish, 1.0 channels, 24 bits, 48khz
4: DTS Master Audio, Swedish, 1.0 channels, 24 bits, 48khz
(core: DTS, 1.0 channels, 24 bits, 768kbps, 48khz)
5: Subtitle (PGS), Dutch
6: Subtitle (PGS), Swedish
7: Subtitle (PGS), Norwegian
8: Subtitle (PGS), Finnish
9: Subtitle (PGS), English
10: Subtitle (PGS), Swedish
[a04] Extracting audio track number 4...
[a04] Decoding with ArcSoft DTS Decoder...
[a04] The ArcSoft DTS Decoder reported an error while decoding.
Aborted at file position 16384.
the movie consists of 2 main .m2ts files, its similar with the 2nd one (position 32768). I've added 2x50mb samples of both here:
http://www.sendspace.com/file/d5yek8
and this happened when I used the -sonic switch for that track instead:
eac3to v2.79
command line: eac3to x:\path 1) 2: G:\test.mkv 3: G:\testPCM.flac 4: G:\testDTSMA.flac -sonic 4: G:\test.dtsma 9: G:\test.sup
------------------------------------------------------------------------------
M2TS, 1 video track, 2 audio tracks, 6 subtitle tracks, 3:08:47
1: Chapters, 10 chapters
2: h264/AVC, 1080p24 (16:9)
3: RAW/PCM, Swedish, 1.0 channels, 24 bits, 48khz
4: DTS Master Audio, Swedish, 1.0 channels, 24 bits, 48khz
(core: DTS, 1.0 channels, 24 bits, 768kbps, 48khz)
5: Subtitle (PGS), Dutch
6: Subtitle (PGS), Swedish
7: Subtitle (PGS), Norwegian
8: Subtitle (PGS), Finnish
9: Subtitle (PGS), English
10: Subtitle (PGS), Swedish
[v02] Extracting video track number 2...
[a03] Extracting audio track number 3...
[a04] Extracting audio track number 4...
[s09] Extracting subtitle track number 9...
[a04] Extracting audio track number 4...
[a03] Reading RAW/PCM...
[v02] Muxing video to Matroska...
[a04] Decoding with DirectShow (Sonic Audio Decoder)...
[a03] Swapping endian...
[a03] Encoding FLAC with libFlac...
[a04] Creating file "G:\test.dtsma"...
[a03] Creating file "G:\testPCM.flac"...
[s09] Creating file "G:\test.sup"...
[a03] The original audio track has a constant bit depth of 24 bits.
[a03] The processed audio track has a constant bit depth of 24 bits.
[a04] The last DTS frame is incomplete and thus gets skipped.
[a04] The FLAC encoder didn't receive the format information.
Aborted at file position 45477599232.
-sonic doesn't give me any output file, nor the "Creating file" message and aborted in the end :S
btw. when I decode the 1.0 RAW/PCM track to .flac I get the "Swapping endian" message, it this intended to appear for 1.0 or 2.0 tracks?
WildTexasChef
6th December 2008, 22:14
Am I mistaken or will eac3to NOT convert DTS audio to AC3?
I tried running eac3to to convert a DTS audio track to AC3 at a bit rate of 448 and it would not convert. It would work fine at a 640k bit rate but not at 448.
Any ideas?
WTC
nautilus7
6th December 2008, 22:16
Post you exact command line and log file.
n0mag!c
6th December 2008, 22:29
You would need to use the PRO version of DIRAC in order to do this.
I dare to disagree with this statement. If you figure this out further this would mean that each time when we do stretching we'll get different result, not the same. Our fear and ignorance are warmed up by smart marketing stuff. But this is easy to check up! We must process stereo file and two mono channels independently and just compare results.
If you have the situation that the relative phase between channels matters, it is imperative to use the multi-channel
processing mode of DIRAC STUDIO and PRO (all channels are being processed at the same time).
I don't really think that this algorithm PURPOSELY produces different result each time it runs.
vucloutr
7th December 2008, 12:25
Hi, I got a little question regarding AAC encoding.
eac3to can encode AAC in quality ranging from (-quality=) 0.01 to 0.99
but the Nero AAC encoder which eac3to uses can do a float from 0..1.
So how is the eac3to range mapped to the Nero AAC encoder range?
Does eac3to 0.99 mean Nero 1.00 or does eac3to 0.99 mean Nero 0.99 (which is what I assume) ?
If it is the latter and only a minor change to make:
Could you be so kind to allow also -quality=0.00 and 1.00, so that one can use the full range the Neor AAC encoder offers ?
madshi
7th December 2008, 13:00
got an Arcsoft error when decoding a movie's DTS-HD MA track to flac
Interesting sample, thanks.
It seems that neither ArcSoft nor Sonic can decode this one. They don't seem to like mono DTS tracks. You'll have to use either Nero or libav to decode it. Of course both only decode the core... :(
If you figure this out further this would mean that each time when we do stretching we'll get different result, not the same.
No, that's not true.
Our fear and ignorance are warmed up by smart marketing stuff.
I talked to the developer of the algorithm, not to the marketing department... ;)
I don't really think that this algorithm PURPOSELY produces different result each time it runs.
It produces the same result when running on the same data. But the part you didn't understand is that the algorithm adopts itself to the waveform of the audio track. So if you split a stereo file into 2 separate mono files and process each separately from each other, the DIRAC algorithm may do different things to each channel, depending on the waveform. And this can distort stereo phase information. Same thing with surround tracks.
If you use phase locked processing (only available in the non-free DIRAC versions), the DIRAC algorithm makes sure that it does exactly the same thing to all channels. This makes sure that stereo and surround phase information is not distorted.
Could you be so kind to allow also -quality=0.00 and 1.00, so that one can use the full range the Neor AAC encoder offers ?
Will be added in the next build.
Thunderbolt8
7th December 2008, 13:28
Interesting sample, thanks.
It seems that neither ArcSoft nor Sonic can decode this one. They don't seem to like mono DTS tracks. You'll have to use either Nero or libav to decode it. Of course both only decode the core... :(does any of these 2 still gives out a slightly better quality or are both equally bad here?
madshi
7th December 2008, 13:47
does any of these 2 still gives out a slightly better quality or are both equally bad here?
I don't know. Nero has a reference decoder, so it might be better. Or not...
xkodi
7th December 2008, 14:47
@madshi
actually, Thunderbolt8's DTS-HD MA mono samples work with Arcsoft in GraphEdit.
i set the "ArcSoft Audio Decoder HD" to decode to "2 channel stereo" and the result is file with empty Right channel (all zeroes) and all the information is in the Left channel. so, it is save to assume that the decode is bit-perfect.
however, if i use eac3to to force Arcsoft to use 2 channels:
C:\eac3to279>eac3to.exe c:\FAAsplit\FAAsplit1.m2ts 3:c:\1.wav -2
i get error:
[a03] Decoding with ArcSoft DTS Decoder...
[a03] The ArcSoft DTS Decoder reported an error while decoding.
Aborted at file position 16384.
so, actually eac3to is wrong - somehow it should decode with ArcSoft to 2 channels like GraphEdit do and then remove the empty Right channel.
please, add to your TODO list - fix for decoding mono tracks with ArcSoft.
[edit] Sonic also decode the mono samples in GraphEdit.
[edit2] Sonic and Arcsoft decodes in GraphEdit are bit identical, but duration is 28-29sec as the core. so, Sonic and Arcsoft decode the DTS core not the DTS-HD MA in GraphEdit with these mono samples.
[edit3] Nero doesn't work at all here with these mono samples. libav is OK.
madshi
7th December 2008, 15:26
@xkodi, which exact graph are you using? And are you using the m2ts file or the demuxed DTS file? Are you sure that the splitter you used didn't output the PCM file instead of the DTS file? On a quick check I couldn't get the DTS track to work in GraphEdit with ArcSoft or Sonic, but of course I might have missed something...
yonta
7th December 2008, 15:51
evo audio delay detection has a problem with v2.79
v2.78
D:\samples>eac3to.exe "The.Interpreter.evo"
EVO, 1 video track, 5 audio tracks, 5 subtitle tracks, 0:01:02
1: h264/AVC, 1080p24 /1.001 (16:9) with pulldown flags
2: E-AC3, 5.1 channels, 1536kbps, 48khz, dialnorm: -27dB, -125ms
3: DTS, 5.1 channels, 24 bits, 1509kbps, 48khz, dialnorm: -4dB, -125ms
4: E-AC3, 5.1 channels, 768kbps, 48khz, dialnorm: -27dB, -125ms
5: E-AC3, 5.1 channels, 768kbps, 48khz, dialnorm: -27dB, -125ms
6: E-AC3 Surround, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB, -29ms
v2.79
D:\samples>eac3to "The.Interpreter.evo"
EVO, 1 video track, 5 audio tracks, 5 subtitle tracks, 0:01:02
1: h264/AVC, 1080p24 /1.001 (16:9) with pulldown flags
2: E-AC3, 5.1 channels, 1536kbps, 48khz, dialnorm: -27dB, -543ms
3: DTS, 5.1 channels, 24 bits, 1509kbps, 48khz, dialnorm: -4dB, -543ms
4: E-AC3, 5.1 channels, 768kbps, 48khz, dialnorm: -27dB, -543ms
5: E-AC3, 5.1 channels, 768kbps, 48khz, dialnorm: -27dB, -543ms
6: E-AC3 Surround, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB, -447ms
madshi
7th December 2008, 16:10
evo audio delay detection has a problem with v2.79
Have you tried which delay is more correct? I've changed to a different VOB/EVO delay calculation method in v2.79, but I found myself that it doesn't work very well, so I've already changed in back for the next build.
xkodi
7th December 2008, 16:14
@madshi
i missed the fact that there is also LPCM track and yes you are right i actually decoded the LPCM track instead of the DTS-HD MA track. so, nothing you can do, Arcsoft and Sonic simply can't handle such tracks in GraphEdit too.
yonta
7th December 2008, 16:15
Have you tried which delay is more correct? I've changed to a different VOB/EVO delay calculation method in v2.79, but I found myself that it doesn't work very well, so I've already changed in back for the next build.
v2.78 is way more correct.
I would say, in this specific case, v2.78 is correct while v2.79 is not.
Tried some more files.
Some are correct while others are about 300~400ms off.
laserfan
7th December 2008, 16:46
Not a bug, but maybe this would bother someone else:
eac3to v2.79
command line: eac3to coredaudio.1.dts audio.ac3 -448
------------------------------------------------------------------------------
DTS, 5.1 channels, 2:29:59, 24 bits, 1509kbps, 48khz
The ArcSoft decoder doesn't seem to work, will use Sonic instead.
Decoding with DirectShow (Sonic Audio Decoder)...
DirectShow reports 5.1 channels, 24 bits, 48khz
Encoding AC3 <448kbps> with libAften...
Creating file "audio.ac3"...
eac3to processing took 28 minutes, 11 seconds.
Done.
I don't have Arcsoft on my PC so this took me aback when I saw it. Or maybe I should ask too: Is Arcsoft preferred over Sonic for DTS decoding? I was thinking of getting their TMT anyway...
poisondeathray
7th December 2008, 16:49
I don't have Arcsoft on my PC so this took me aback when I saw it. Or maybe I should ask too: Is Arcsoft preferred over Sonic for DTS decoding? I was thinking of getting their TMT anyway...
That is answered on page 1 under "evaluation of available decoders"
Cheers
asarian
7th December 2008, 17:31
eac3to v2.79
command line: eac3to coredaudio.1.dts audio.ac3 -448
------------------------------------------------------------------------------
DTS, 5.1 channels, 2:29:59, 24 bits, 1509kbps, 48khz
The ArcSoft decoder doesn't seem to work, will use Sonic instead.
Decoding with DirectShow (Sonic Audio Decoder)...
DirectShow reports 5.1 channels, 24 bits, 48khz
Encoding AC3 <448kbps> with libAften...
Creating file "audio.ac3"...
eac3to processing took 28 minutes, 11 seconds.
Done.
Just out of curiosity, why would you downgrade the AC3 stream to 448kbps (from 1536kbps DTS), instead of just using the default 640kbps?
Thunderbolt8
7th December 2008, 17:47
Have you tried which delay is more correct? I've changed to a different VOB/EVO delay calculation method in v2.79, but I found myself that it doesn't work very well, so I've already changed in back for the next build.
is this problem only restricted to delay or does it affect gaps/overlapping as well?
n0mag!c
7th December 2008, 19:54
It produces the same result when running on the same data. But the part you didn't understand is that the algorithm adopts itself to the waveform of the audio track. So if you split a stereo file into 2 separate mono files and process each separately from each other, the DIRAC algorithm may do different things to each channel, depending on the waveform. And this can distort stereo phase information. Same thing with surround tracks.
Aha, I've figured that out lately, DIRAC algorithm (due to its purpose) can shift phase in different degree depending on the input data.
If you use phase locked processing (only available in the non-free DIRAC versions), the DIRAC algorithm makes sure that it does exactly the same thing to all channels. This makes sure that stereo and surround phase information is not distorted.
Phase locked processing isn't available in free stereo version? Damn. At least we could feed it with pairs (C+R, C+L, C+SL, C+SR, C+LFE) to get phase synced channels but lose in speed.
madshi
7th December 2008, 20:32
is this problem only restricted to delay or does it affect gaps/overlapping as well?
Only delay. So no need to worry...
Phase locked processing isn't available in free stereo version? Damn. At least we could feed it with pairs (C+R, C+L, C+SL, C+SR, C+LFE) to get phase synced channels but lose in speed.
Yeah, unfortunately the free version can't even do stereo phase locked processing. Only mono. That's why I think it's useless for our purposes. The only situation where the free DIRAC version can be used is for producers/studios/musicians who record and edit one track after the other in mono.
madshi
7th December 2008, 22:45
eac3to v2.80 released
http://madshi.net/eac3to.zip
* fixed: FLAC files with missing runtime information were not accepted
* gone back to old VOB/EVO auto delay calculation method, more reliable for me
* improved TS broadcast audio delay detection
* added support for constant bitrate AAC encoding
* added support for AAC encoding 0.00 and 1.00 quality
nautilus7
7th December 2008, 22:53
Thanks man. I really love to know that i can expect a new version every Sunday night.
Thunderbolt8
8th December 2008, 01:10
Thanks man. I really love to know that i can expect a new version every Sunday night.
+1 :thanks:
can you please comment on the improved .ts audio delay correction, in far was it unprecise before? only a few ms? or XXms? do I need to check again now those I've already remuxed?
vucloutr
8th December 2008, 01:30
eac3to v2.80 released
http://madshi.net/eac3to.zip
* ...
* added support for AAC encoding 0.00 and 1.00 qualityThank you ! :)
PS: just the description isn't updated yet ;)
madshi
8th December 2008, 08:26
Thanks man. I really love to know that i can expect a new version every Sunday night.
At one time in the past I realized that I was spending too much time on a non-profit software like eac3to. So now I'm limiting myself to only work on it on Sundays... ;) Of course there's no guarantee that I'll have time every Sunday. But so far there are usually at least a few bugfixes every week...
can you please comment on the improved .ts audio delay correction, in far was it unprecise before? only a few ms? or XXms? do I need to check again now those I've already remuxed?
It could be quite incorrect (up to several hundreds of ms) for such movies where there were video frames before the first sequence header in the source file. That is not the case for any HD DVD or Blu-Ray movies I know. Also cleanly cut broadcast movies shouldn't have this, either. But randomly cut broadcasts could show this problem. You would have noticed that, though, cause the delay was usually way off in such cases. So if you did a quick audio sync check on every movie and you found it to be ok, there's nothing to worry about...
PS: just the description isn't updated yet ;)
Done.
FredThompson
8th December 2008, 16:48
available Intro and Trial (Try and Buy) versions of Arcsoft TotalMediaTheatre (TMT) and TotalMediaExtreme (TME) and which of them work with eac3to:
1. TMT Intro version 2.1.6.120 - "TotalMediaTheatre_Retail_INTRO_E.exe" (md5 77ea5ecdc52af9f20d646cc352389cc2):
http://base.arcsoft.com/downloads/intro/TotalMediaTheatre_Retail_INTRO_E.exeThe link is broken for me. That might be redirection blocking, I don't know. Your referenced post is a few months old so I'm querying to see if there is updated sourcing information. Goal is 7.1 with as little overhead as possible, of course.
jimz06
8th December 2008, 17:44
A search on the Forums reveals that at least several people on this thread have managed to remux the Babel HD DVD so I was hoping that one of those fortunate people could share some wisdom.
I've lost track of the various ways and tools I've used produce a file. The last attempt was to use eac3to to extract the AC3 track, EVOdemux to extract the H264 file, and then mkvmerge to put them together. The end result always seems to be the same, a break up of the picture every 10 seconds or so and loss of audio sync. When I play the H264 stream alone it plays fine without the breakups and the framerate counter in MPC says it is playing at 24 fps. I can listen to the AC3 stream separately also. Am I missing some critical setting on any of these tools that is causing me fits? All of the Blu Rays I've remuxed have been a dream compared to this HD DVD.
Beastie Boy
8th December 2008, 17:59
@jimz06
Use eac3to to produce a mkv file containing the video (just use .mkv as the output extension). Then use mkvmerge to join your video and audio.
Cheers, Beastie.
hubblec4
8th December 2008, 17:59
eac3to v2.80 released
http://madshi.net/eac3to.zip
* added support for constant bitrate AAC encoding
quality
how i can realize that?
hubble
nautilus7
8th December 2008, 18:09
I suppose like you specify the bitrate for other format encoding. Try it.
Thunderbolt8
8th December 2008, 18:22
madshi, regarding that DTS-HD MA 1.0 movie I sent you 2 samples from, there is one gap/overlap detected for the DTS-HD MA track between the sync point of both main m2ts files. but when I process the LPCM -> FLAC track from it, I dont get this message. is this ok ? (I know that LPCM frames are rather short and the delay difference would be indiscernible, but could this be a general bug, or did eac3to only decide that the correction of this minor delay wouldnt be worth it and therefore did not put out a gap/overlap message?)
nurbs
8th December 2008, 18:47
how i can realize that?
Haven't tried yet, but I guess its the -xxx switch, same as ac3.
BTW thanks madshi
Jeff Flowerday
8th December 2008, 19:41
A search on the Forums reveals that at least several people on this thread have managed to remux the Babel HD DVD so I was hoping that one of those fortunate people could share some wisdom.
I've lost track of the various ways and tools I've used produce a file. The last attempt was to use eac3to to extract the AC3 track, EVOdemux to extract the H264 file, and then mkvmerge to put them together. The end result always seems to be the same, a break up of the picture every 10 seconds or so and loss of audio sync. When I play the H264 stream alone it plays fine without the breakups and the framerate counter in MPC says it is playing at 24 fps. I can listen to the AC3 stream separately also. Am I missing some critical setting on any of these tools that is causing me fits? All of the Blu Rays I've remuxed have been a dream compared to this HD DVD.
Sounds more like a play back problem, try reinstalling haali media splitter.
Devrethman
8th December 2008, 22:49
Okay, stupid n00b question here: I just bought Band of Brothers on Blu-ray, it's the first disc I've seen that has DTS "Master Audio" which I have gathered to be the same thing as DTS-HD. When I run eac3to to check what the track order is, it says that the audio conversion is not supported. However, the first post also says that the ArcSoft decoder (which I have) works well for DTS-HD, and when I start decoding it to FLAC, it decodes with the arcsoft decoder just fine. is it actually decoding the HD track or just the DTS core? If the latter, is it possible to get it to decode the whole track or is that not supported yet?
Thunderbolt8
8th December 2008, 23:29
got this message with the german dances with wolves BD:
eac3to v2.80
command line: X:\eac3to\eac3to X:\movie 1) 2: X:\dances.mkv 4: X:\dances.flac 4: X:\dances.dtsma 11: X:\dances.sup
------------------------------------------------------------------------------
M2TS, 1 video track, 4 audio tracks, 5 subtitle tracks, 3:56:38
1: Chapters, 31 chapters
2: h264/AVC, 1080p24 /1.001 (16:9)
3: DTS Master Audio, German, 7.1 channels, 16 bits, 48khz
(core: DTS, 5.1 channels, 16 bits, 1509kbps, 48khz)
4: DTS Master Audio, English, 7.1 channels, 16 bits, 48khz
(core: DTS, 5.1 channels, 16 bits, 1509kbps, 48khz)
5: AC3, English, 2.0 channels, 192kbps, 48khz
6: AC3, English, 2.0 channels, 192kbps, 48khz
7: Subtitle (PGS), German
8: Subtitle (PGS), German
9: Subtitle (PGS), German
10: Subtitle (PGS), German
11: Subtitle (PGS), English
[v02] Extracting video track number 2...
[a04] Extracting audio track number 4...
[s11] Extracting subtitle track number 11...
[a04] Extracting audio track number 4...
[v02] Muxing video to Matroska...
[a04] Decoding with ArcSoft DTS Decoder...
[a04] Encoding FLAC with libFlac...
[a04] Creating file "G:\dances.dtsma"...
[a04] Creating file "G:\dances.flac"...
[s11] Creating file "G:\dances.sup"...
[a04] The last DTS frame is incomplete and thus gets skipped.
[a04] Original audio track, L+R+C+BL+BR+SL+SR: constant bit depth of 16 bits.
[a04] Processed audio track, LFE: bitdepth analyzation failed.
[a04] Processed audio track, L+R+C+BL+BR+SL+SR: constant bit depth of 16 bits.
[a04] Processed audio track, LFE: bitdepth analyzation failed.
[a04] The last DTS frame is incomplete and thus gets skipped.
Added fps value to MKV header.
Video track 2 contains 340422 frames.
eac3to processing took 1 hour, 2 minutes.
Done.
50mb sample:
http://www.sendspace.com/file/xkpcl3
madshi
8th December 2008, 23:33
The last attempt was to use eac3to to extract the AC3 track, EVOdemux to extract the H264 file, and then mkvmerge to put them together. The end result always seems to be the same, a break up of the picture every 10 seconds or so and loss of audio sync.
As Beastie Boy suggested, try "eac3to HdDvdSourceFolder movie.mkv". Then play the MKV file just as it is (without the audio). Are there still break ups? If so, probably your playback system is borked (as suggested by Jeff Flowerday). If there are no breaks up when playing the video only MKV, mux the audio track to the video MKV by using mkvtoolnix. Does the final result play fine? If not, again probably something is wrong with your DirectShow filter setup. Try different video/audio decoders and maybe also different video/audio renderers...
how i can realize that?
nautilus7 and nurbs have guessed right. It's the same way you can specify bitrates for AC3 and DTS. E.g. "eac3to source.whatever dest.aac -192".
If you don't specify anything for AAC encoding, eac3to will use quality 0.5. If you specify a different quality level, eac3to will use that instead. And if you specify a bitrate, eac3to will use constant bitrate encoding instead of quality based encoding.
madshi, regarding that DTS-HD MA 1.0 movie I sent you 2 samples from, there is one gap/overlap detected for the DTS-HD MA track between the sync point of both main m2ts files. but when I process the LPCM -> FLAC track from it, I dont get this message. is this ok ?
Yes. LPCM gaps usually only get big enough to be reported if there are a whole lot of m2ts parts. The threshould for reporting a gap is 7ms, IIRC.
When I run eac3to to check what the track order is, it says that the audio conversion is not supported.
You most probably have used an incorrect command line then.
However, the first post also says that the ArcSoft decoder (which I have) works well for DTS-HD, and when I start decoding it to FLAC, it decodes with the arcsoft decoder just fine. is it actually decoding the HD track or just the DTS core?
That is explained in the first post of this thread. The full DTS-HD track is decoded bit perfectly, when the ArcSoft decoder is used by eac3to.
madshi
9th December 2008, 00:14
got this message with the german dances with wolves BD:
4: DTS Master Audio, English, 7.1 channels, 16 bits, 48khz
[a04] Original audio track, L+R+C+BL+BR+SL+SR: constant bit depth of 16 bits.
[a04] Processed audio track, LFE: bitdepth analyzation failed.
That is *VERY* bad. It means that the English audio track on the Dances with Wolves Blu-Ray is broken. The LFE is channel is empty. That's why eac3to can't find out which bitdepth it has.
Now it's not the end of the world. We could extract the LFE channel of the German audio track and use that for the English audio track. But for all intends and purposes the Blu-Ray should be considered broken and a repressing should be requested from the studio.
Thunderbolt8
9th December 2008, 02:26
hm how can I do that to have it in flac lossless as well? extract the german dts-hd track to .wavs and then just convert the LFE channel to flac? how to combine it then with the other 4 channels from the english track all together to one single flac track again?
madshi
9th December 2008, 09:35
hm how can I do that to have it in flac lossless as well? extract the german dts-hd track to .wavs and then just convert the LFE channel to flac? how to combine it then with the other 4 channels from the english track all together to one single flac track again?
Decode both the English and German audio tracks to "WAVs". Then swap the LFE WAV channel WAV files of both tracks. Next combine the separate WAV files for the English track to one big WAV file. Unfortunately eac3to can't do that, but I think wavewizard can. Finally do "eac3to combined.wav english.flac" to encode the final FLAC file.
Thunderbolt8
9th December 2008, 18:19
madshi (or someone else), im not familiar with wavewizard, but it looks to me like this programm cant do it. I can change channel mapping with it or convert wav files to another format, but I cant find any option where I can join multiple files to a big single one.
is there perhaps another program who can do this?
b66pak
9th December 2008, 19:17
try using m$ wavavimux (google it - its free) to mux 2, 6 or 8 wav mono channels in one giant .avi which can be used as an input be many audio tools
_
tebasuna51
9th December 2008, 20:20
madshi (or someone else), im not familiar with wavewizard, but it looks to me like this programm cant do it. I can change channel mapping with it or convert wav files to another format, but I cant find any option where I can join multiple files to a big single one.
is there perhaps another program who can do this?
WaveWizard instructions:
- Add monowavs in standard order: FL,FR,FC,LF,SL,SR
- Edit -> Preferences:
All options unchecked but:
Stream manipulation -> Merge files
Output format -> Wave PCM
(or WaveFormatEx but the ChannelMask value is always 0)
- Convert
You can use also the sox command line:
sox -M FL.wav FR.wav FC.wav LF.wav BL.wav BR.wav multichannel.wav
vBulletin® v3.8.11, Copyright ©2000-2025, vBulletin Solutions Inc.