Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > General > Audio encoding

Closed Thread
 
Thread Tools Search this Thread Display Modes
Old 23rd August 2008, 08:06   #5881  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
@Madshi
First i did that on Original Blu Ray Disc, i don't think in that last frame would be incomplete. And second time for you, i did that on sample by cutting starting piece.
Secondly that mean, eac3to mux's exact stream to MKV, even if last frame is incomplete. And it only remove last incomplete frame in demuxing.

Hope, i'm getting right. And i never mean to offend you, i just want help you by testing stuff.

Last edited by sehgal.v7; 23rd August 2008 at 08:09.
sehgal.v7 is offline  
Old 23rd August 2008, 09:31   #5882  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by Roscoe62 View Post
I've now uploaded the Mr & Mrs Smith files for you.
Thanks. This problem will be fixed in the next eac3to build.

Quote:
Originally Posted by sehgal.v7 View Post
First i did that on Original Blu Ray Disc, i don't think in that last frame would be incomplete. And second time for you, i did that on sample by cutting starting piece.
Secondly that mean, eac3to mux's exact stream to MKV, even if last frame is incomplete. And it only remove last incomplete frame in demuxing.
The point is that comparing the file size alone does not help. The file can reach bit perfection and still have a different file size (due to frames missing at the end). You need to compare the files with a hex editor or with a good file comparison tool to find out whether you got bit perfection or not.
madshi is offline  
Old 23rd August 2008, 10:45   #5883  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
(A)
1. Extracted 1080p24.vc1 from VC1.ts by Eac3to - (If any frame was missing, it must have been removed)
2. Made M2TS file with Scenarist. Then Extracted 1080p24.vc1 from VC1.m2ts.
------- Compared both files by Hex // both are same.

(B)
3. VC1.m2ts to VC1.mkv with Eac3to.
4. Extracted 1080p24.vc1 from VC1 by MKVExtract.
------- Compared 1080p24.vc1 (A) & 1080p24.vc1 (B) by Hex. Lotsa difference!!

Here's the Report!!
hxxp://rapidshare.com/files/139456657/Report.rar
sehgal.v7 is offline  
Old 23rd August 2008, 10:55   #5884  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by sehgal.v7 View Post
(A)
1. Extracted 1080p24.vc1 from VC1.ts by Eac3to - (If any frame was missing, it must have been removed)
2. Made M2TS file with Scenarist. Then Extracted 1080p24.vc1 from VC1.m2ts.
------- Compared both files by Hex // both are same.

(B)
3. VC1.m2ts to VC1.mkv with Eac3to.
4. Extracted 1080p24.vc1 from VC1 by MKVExtract.
------- Compared 1080p24.vc1 (A) & 1080p24.vc1 (B) by Hex. Lotsa difference!!

Here's the Report!!
hxxp://rapidshare.com/files/139456657/Report.rar
I don't know exactly what you did there, but the "left" file in the comparison looks damaged to me. Look, there's even the text "mkvextract 2.2.0" in the left VC-1 file! That does not make any sense. I think you must have made a mistake somewhere...
madshi is offline  
Old 23rd August 2008, 11:03   #5885  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
@Madshi
1. eac3to VC1.ts -demux
2. eac3to VC1.m2ts -demux
3. eac3to VC1.m2ts VC1.mkv
4. mkvextract tracks VC1.mkv 1:1080p24.vc1 --raw

You can test yourself. http://rapidshare.com/files/139461876/VC1.m2ts by Scenarist.

Last edited by sehgal.v7; 23rd August 2008 at 11:13.
sehgal.v7 is offline  
Old 23rd August 2008, 11:13   #5886  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by sehgal.v7 View Post
4. mkvextract tracks VC1.mkv 1:1080p24.vc1 --raw
That command line is wrong. Just read the last 15 comments in this thread and you'll see why.
madshi is offline  
Old 23rd August 2008, 11:31   #5887  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
Now got it right but still at last http://rapidshare.com/files/139465417/Report_2.rar
mkvextract tracks VC1.mkv --raw 1:1080p24.vc1
sehgal.v7 is offline  
Old 23rd August 2008, 12:16   #5888  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by sehgal.v7 View Post
Now got it right but still at last http://rapidshare.com/files/139465417/Report_2.rar
mkvextract tracks VC1.mkv --raw 1:1080p24.vc1
Yeah, it's exactly as I said:

The only difference between the two files is that one them has one frame missing. One of the files contains 433 frames while the other contains 432 frames. But those 432 frames are 100% bit perfectly identical.
madshi is offline  
Old 23rd August 2008, 12:25   #5889  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
@Madshi
Thanks Madshi, got every single clearance on VC1.
Jus one last ques- Are you planning to add feature to combine AC3 & TrueHD files together? :-)
Me still waiting to burn Blue Planet again ;-)

Last edited by sehgal.v7; 23rd August 2008 at 12:30.
sehgal.v7 is offline  
Old 23rd August 2008, 13:08   #5890  |  Link
Jan Marijniszoon
Registered User
 
Jan Marijniszoon's Avatar
 
Join Date: Mar 2003
Posts: 119
madshi, I now found a workaround (tsMuxer combined with TsRemux) to get TrueHD tracks (with ac3-core) to work on 'homemade' Blu-rays.
I tested it myself and the TrueHD tracks plays perfectly and the core also gets recognized.

So I would like to ask you nicely when you have time for it to check out if it is possible to remove dialog normalisation from these type of tracks. How can it be done? I guess you have to find a way to remove the dialog normalisation from both the headers in the file?
Jan Marijniszoon is offline  
Old 23rd August 2008, 15:52   #5891  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by sehgal.v7 View Post
Jus one last ques- Are you planning to add feature to combine AC3 & TrueHD files together? :-)
I already answered that question.

Quote:
Originally Posted by Jan Marijniszoon View Post
madshi, I now found a workaround (tsMuxer combined with TsRemux) to get TrueHD tracks (with ac3-core) to work on 'homemade' Blu-rays.
I tested it myself and the TrueHD tracks plays perfectly and the core also gets recognized.

So I would like to ask you nicely when you have time for it to check out if it is possible to remove dialog normalisation from these type of tracks. How can it be done? I guess you have to find a way to remove the dialog normalisation from both the headers in the file?
I'll add it to the to do list.
madshi is offline  
Old 23rd August 2008, 16:10   #5892  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
"I may eventually add a feature to create an interweaved TrueHD+AC3 track for Scenarist muxing."
Oki, will be waiting for that.

Secondly, i was jus experimenting with files.. If TrueHD stream contains only 16bit Information, when converting it to WAV, why it outputs 24bit??
Quote:
TrueHD/AC3, 5.1 channels, 48khz, dialnorm: -27dB
Extracting TrueHD stream...
Removing dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
DirectShow reports 5.1 channels, 24 bits, 48khz
Writing WAV...
Creating file "K:\Audio.wav"...
This audio track has a constant bit depth of 16 bits.
Caution: The WAV file is bigger than 2GB.
Some WAV readers might not be able to handle this file correctly.
eac3to processing took 10 minutes, 3 seconds.
Done.
LPCM or RAW wasn't accepted by Scenarist// so had to write WAV file. Is channel configuration for 6CH (L,C,R,LS,RS,lfe)??

-----
Blue Planet contains Dolby TrueHD 5.1 16-bit, how can i force eac3to to output LPCM 16-bit? Eac3to output 24-bit.

Last edited by sehgal.v7; 23rd August 2008 at 17:09.
sehgal.v7 is offline  
Old 23rd August 2008, 16:24   #5893  |  Link
xkodi
Registered User
 
Join Date: Aug 2002
Posts: 221
i also can confirm that VC1 is bit-perfect. so, what we have till now:

- MPEG-2: bit-perfect.
- VC-1: bit-perfect, but usually the last frame is missing (at least on all test i run the last frame is always missing, anyway it's not a problem)
- H.264: AUDs and "filler data" are removed, but as madshi explained they are useless data, all other data are bit-perfect, but always have "the double sequence headers at the beginning of the stream".
xkodi is offline  
Old 23rd August 2008, 16:30   #5894  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
@xKodi
Yeap, Agreed on all points!!
sehgal.v7 is offline  
Old 23rd August 2008, 17:55   #5895  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by sehgal.v7 View Post
Secondly, i was jus experimenting with files.. If TrueHD stream contains only 16bit Information, when converting it to WAV, why it outputs 24bit??
That's a bug in the current version. It will be fixed in the next build.

Quote:
Originally Posted by sehgal.v7 View Post
Is channel configuration for 6CH (L,C,R,LS,RS,lfe)??
WAV channel order is L, R, C, LFE, LS, RS.

Quote:
Originally Posted by xkodi View Post
- MPEG-2: bit-perfect.
- VC-1: bit-perfect, but usually the last frame is missing (at least on all test i run the last frame is always missing, anyway it's not a problem)
Correct. You need to use "--raw" for both MPEG2 and VC-1, though, to achieve bit perfectness.

Quote:
Originally Posted by xkodi View Post
- H.264: AUDs and "filler data" are removed, but as madshi explained they are useless data, all other data are bit-perfect, but always have "the double sequence headers at the beginning of the stream".
Correct. In this case you must NOT use "--raw", btw. And I want to add that I don't consider the AUDs to be useless for a continuous stream of bytes. I just consider AUDs to be useless inside of an MKV or MP4 container. IMO an MKV/MP4 demuxer which targets to create a raw h264 file should readd AUDs. Sadly current demuxers don't do that. Maybe some day I'll have to write my own MKV demuxer. But it's not a big deal since AUDs are optional, anyway, and decoders don't need them in any case.
madshi is offline  
Old 23rd August 2008, 18:10   #5896  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
@Madshi
Thanx for fixing bug in next release.
Secondly, Scenarist is accepting L,C,R,LS,RS,lfe config for while.
Screenshot http://rapidshare.com/files/139543028/Untitled.jpg
sehgal.v7 is offline  
Old 23rd August 2008, 18:24   #5897  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 9,140
Quote:
Originally Posted by sehgal.v7 View Post
Secondly, Scenarist is accepting L,C,R,LS,RS,lfe config for while.
I'm not sure but I guess that's the channel order which Scenarist will write into the final m2ts file and *not* the channel order it expects in the wav file. The wav file channel order is standardized and eac3to uses the channel order which everybody uses in wav files. Scenarist expecting a non-standard channel order in the wav file would be quite strange.
madshi is offline  
Old 24th August 2008, 00:30   #5898  |  Link
deacon crusher
Registered User
 
Join Date: Feb 2008
Posts: 14
So, I'm trying to convert CJ7 (the newish Steven Chow movie) to MKV and running into quite a bit of difficulty.

It's a multi-part (But I've dealt with that before by muxing together with tsmuxer.) However, there's some problem with the streams, when running tsmuxer or eac3to I get many alerts of the type

The program channel mapping changes in the middle of the stream.

So, I'm demuxing each seperate m2ts file using eac3to and I'm going to try to then combing together the h264 files into 1 mkv and the ac3 files into 1 ac3. However, I'd love any tips regarding correct command line options for this phase and any gotchas to look out for to make sure I'm getting rid of audio overlap and will have a final product where audio and video are in sync.

EDIT
I guess specifically my question is in madshi's quote he says audio overlap can then be fixed, how??
END EDIT

Thanks

Quote:
Originally Posted by madshi View Post
No. The problem with m2ts multi part movies is that audio often overlaps a little. In order to fix that problem completely I would probably have to write a full featured m2ts muxer. That is not planned. With HD DVD movies it's easier because the audio doesn't overlap. So I just need to join the files and maybe correct a few timestamps. Easy. Not so with Blu-Ray.

You have 2 options:

(1) Either you can let tsMuxeR or TsRemux do the m2ts multipart joining. Neither of them will correct the audio overlaps, though. So the end result may play bumpy on the join points or eventually audio sync will gradually be lost (depends on the media player / splitter).

(2) Or you can let eac3to demux all video and audio tracks. Audio overlaps can then be fixed. Finally you can recreate a new m2ts file by using tsMuxeR or Scenarist.

Last edited by deacon crusher; 24th August 2008 at 01:46. Reason: adding detail to question.
deacon crusher is offline  
Old 24th August 2008, 03:37   #5899  |  Link
sehgal.v7
Registered User
 
Join Date: Jul 2008
Posts: 93
@Madshi
Oki, i'll test and post results back!!
sehgal.v7 is offline  
Old 24th August 2008, 03:54   #5900  |  Link
scarbrtj
Registered User
 
Join Date: Jul 2007
Posts: 48
Madshi:

Re: "The Bank Job" AVC/7.1 DTSHD

The AVC video track duration is 1:54:34.075

-core eac3to extraction yields a 1:52:01.749 DTS file

If I instead convert to AC3 with eac3to, we get a 1:54:34.075 file... and muxing this to the video gives perfect sync.

Obviously, a mux of the -core DTS and AVC video gives a very big video/audio sync problem.

But, I'd like to get the -core DTS and AVC together... thoughts?
scarbrtj is offline  
Closed Thread

Tags
eac3to

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 07:59.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.