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

Reply
 
Thread Tools Search this Thread Display Modes
Old 30th January 2009, 12:07   #8081  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
Quote:
Originally Posted by bigdog660 View Post
Hello Madshi! Finally, here are my test results for all three versions.
Could you please keep that Hellboy source, so that you can retest later with v3.06? That would be nice - thanks!
madshi is offline   Reply With Quote
Old 30th January 2009, 13:29   #8082  |  Link
bigdog660
Registered User
 
Join Date: Mar 2004
Posts: 61
Quote:
Originally Posted by madshi View Post
Could you please keep that Hellboy source, so that you can retest later with v3.06? That would be nice - thanks!
No problem... thanks!
bigdog660 is offline   Reply With Quote
Old 30th January 2009, 14:31   #8083  |  Link
tebasuna51
Moderator
 
tebasuna51's Avatar
 
Join Date: Feb 2005
Location: Spain
Posts: 5,275
Quote:
Originally Posted by Blue_MiSfit View Post
The clipping MP2 is uploading. I re-verified that both MPG -> MP2 -> AC3 and MPG -> AC3 both produce clipping, but MPG -> MP2 -> 32float WAV (with Foobar) > AC3 (with eac3to) works fine.
Here is a sample (12 sec) with the problem isolated, is decoded fine with Foobar2000, BeSweet, NicAudio and also ffmpeg, but using eac3to-libav there are a inversion at clipping points than cause audible cliks (see the png included)

I hope than help to fix the problem.
__________________
BeHappy, AviSynth audio transcoder, in Doom9 forums. NicAudio, BassAudio, audio decoders.
tebasuna51 is offline   Reply With Quote
Old 30th January 2009, 14:31   #8084  |  Link
honai
Guest
 
Posts: n/a
WALL-E US vs. WALL-E Germany

I have both the US and German Blu-rays of WALL-E.

Surprisingly, eac3to shows a runtime of 1:38:25 (00095.mpls) for the US disc, but 1:32:43 (00119.mpls) for the German disc.

I have noticed that the US version consists basically of a single large M2TS, but the German version is split up into lots of smaller segments. eac3to also reports many overlaps for the German disc.

Since I don't believe that they cut 6 minutes of footage for the German market, is it possible that eac3to realizes the gaps incorrectly? (The resulting audio files for the German disc are in fact 1:33:35 long.)

Last edited by honai; 30th January 2009 at 14:38.
  Reply With Quote
Old 30th January 2009, 15:09   #8085  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
Quote:
Originally Posted by tebasuna51 View Post
I hope than help to fix the problem.
It will help, thanks!

Quote:
Originally Posted by honai View Post
I have both the US and German Blu-rays of WALL-E.

Surprisingly, eac3to shows a runtime of 1:38:25 (00095.mpls) for the US disc, but 1:32:43 (00119.mpls) for the German disc.
Strange. My "Englisch.flac" from the US disc is 1:38:11 long, so 1:32:43 is definitely too short. Does the eac3to title listing (which you get if you simply do "eac3to blurayDrive:") also list 1:32:43? If so, the overlap removal can not be responsible for the too short runtime, because the overlap removal is only post processing. Are there no other playlists with a bigger runtime on the German disc? Have you tried actually processing the disc? Maybe the runtime display is simply wrong and processing the disc will still work nevertheless?

Maybe you could upload the CLIPINF and PLAYLIST folders? Zipped it should be only a few KBs. Thanks!

P.S: What does playlist "00082.mpls" from the German Blu-Ray say? You can do "eac3to bluRayFolder\PLAYLIST\00082.mpls", if this playlist isn't offered by default.

Last edited by madshi; 30th January 2009 at 15:12.
madshi is offline   Reply With Quote
Old 30th January 2009, 15:22   #8086  |  Link
honai
Guest
 
Posts: n/a
Thanks. Yes, I already checked out that possibility. For 00082.mpls it says 1:32:31 for the German disc. I have seen somewhere else that for this playlist the runtime is supposed to be 1:38:32, that's why I wondered if maybe eac3to is making a mistake.

I have also observed that for 00119.mpls (German) eac3to initially reported 1:32:43 runtime in the tracklisting, when selecting the title for demuxing it shows 1:33:33, but demuxed/processed runtime was 1:33:35, so it seems that those are guesses by eac3to anyway, right? I'm especially wondering about the discrepancy between the initial title listing and the listing when that title is selected.

And yes, 00119.mpls and 00120.mpls are the longest-running titles displayed.

Here's the requested archive:

http://www.sendspace.com/file/ufd3xb

Thanks for looking into this, much appreciated.

EDIT:

After verifying each file from the 00082.mpls I found that one of my M2TS was corrupt. Doh! Will try again, but I guess this fixes it. Sorry for wasting your time.

EDIT:

Could you please add a sanity check if one of the M2TS files referenced in the playlist has a size of 0 bytes or is missing altogether from HDD? I have noticed that in cases like above eac3to will silently ignore the missing/corrupt file. Thanks!

Last edited by honai; 30th January 2009 at 16:13.
  Reply With Quote
Old 31st January 2009, 20:00   #8087  |  Link
Snowknight26
Registered User
 
Join Date: Aug 2007
Posts: 1,381
eac3to doesn't seem to encode 32-bit FLACs, but works fine if I remove -down32, resulting in a 24-bit FLAC:

Quote:
eac3to v3.05
command line: eac3to.exe "Z:\Encoding Tools\temp\audio\2fast2furious.eac3" "Z:\Encoding Tools\temp\audio\2fast2furious.flac" -down32
------------------------------------------------------------------------------
E-AC3, 5.1 channels, 1:47:36, 1536kbps, 48khz
The Nero decoder doesn't seem to work, will use libav instead.
Decoding with libav/ffmpeg...
Remapping channels...
Reducing depth from 64 to 32 bits...
The FLAC encoder received a non-supported data format.
Aborted at file position 262144.
Also, any reason why creating a silent AC3 (E-AC3?) frame would fail?
Quote:
eac3to v3.05
command line: eac3to.exe "G:\hotfuzz" 1) 4: ..\temp\audio\hotfuzz.eac3
------------------------------------------------------------------------------
EVO, 1 video track, 6 audio tracks, 7 subtitle tracks, 2:00:52
4: E-AC3 EX, English, 5.1 channels, 1536kbps, 48khz, dialnorm: -27dB, 133ms
[a04] Extracting audio track number 4...
[a04] Removing E-AC3 dialog normalization...
[a04] Applying (E-)AC3 delay...
[a04] Creating silent AC3 frame failed. <WARNING>
[a04] Creating file "..\temp\audio\hotfuzz.eac3"...
Video track 3 contains 173880 frames.
eac3to processing took 3 minutes, 8 seconds.
Done.
Snowknight26 is offline   Reply With Quote
Old 31st January 2009, 20:15   #8088  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
Quote:
Originally Posted by honai View Post
Could you please add a sanity check if one of the M2TS files referenced in the playlist has a size of 0 bytes or is missing altogether from HDD? I have noticed that in cases like above eac3to will silently ignore the missing/corrupt file. Thanks!
Will add a warning...

Quote:
Originally Posted by Snowknight26 View Post
eac3to doesn't seem to encode 32-bit FLACs, but works fine if I remove -down32, resulting in a 24-bit FLAC
FLAC does not support 32bit, highest bitdepth supported is 24bit. BTW, some experts believe that even 16bit is good enough, others disagree. But all experts agree that 24bit is more than enough. So there's no reason to go 32bit, anyway, unless you want to do further processing on the file...

Quote:
Originally Posted by Snowknight26 View Post
Also, any reason why creating a silent AC3 (E-AC3?) frame would fail?
Silent frame creation only works for AC3 because Aften doesn't support E-AC3 encoding yet. It's a bug that eac3to even tries to encode a silent E-AC3 frame, but it's only a cosmetical problem.
madshi is offline   Reply With Quote
Old 31st January 2009, 20:38   #8089  |  Link
Snowknight26
Registered User
 
Join Date: Aug 2007
Posts: 1,381
Quote:
Originally Posted by madshi View Post
FLAC does not support 32bit, highest bitdepth supported is 24bit.
http://flac.sourceforge.net/faq.html#general__samples
Quote:
FLAC supports linear PCM samples with a resolution between 4 and 32 bits per sample.
:\

Quote:
Originally Posted by madshi View Post
So there's no reason to go 32bit, anyway, unless you want to do further processing on the file...
...or for Fun. I was just trying to see how big a 64-bit floating point FLAC file (from an EAC3 file) would be, but since it doesn't support floating point precision, I wen't for 32, then 24bit when that didn't work.

Last edited by Snowknight26; 31st January 2009 at 21:19.
Snowknight26 is offline   Reply With Quote
Old 31st January 2009, 23:14   #8090  |  Link
Inspector.Gadget
Registered User
 
Join Date: May 2008
Posts: 1,618
Splitting audio tracks by chapters

Hey Madshi, quick question:

I'm using eac3to through the HD-DVD/Blu-ray Streams Extractor to pull some audio off a concert Blu-ray. Currently, I'm muxing the audio (extracted as AC3, FLAC, and DTS) to MKV with MKVMerge, splitting the resulting .MKA file by chapter, and then demuxing the audio from each resulting .MKA segment. I'm doing this in order to split the audio tracks by song. Is there a faster way to do this, perhaps during extraction or transcoding? Thanks.

Last edited by Inspector.Gadget; 31st January 2009 at 23:17.
Inspector.Gadget is offline   Reply With Quote
Old 1st February 2009, 00:21   #8091  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
Quote:
Originally Posted by Snowknight26 View Post
Maybe the spec supports it, but I think the official encoder does not. See here:

http://www.hydrogenaudio.org/forums/...hp/t23613.html

It's possible that this post is outdated, though, not sure...

Quote:
Originally Posted by Inspector.Gadget View Post
Is there a faster way to do this, perhaps during extraction or transcoding?
Right now it's not possible with eac3to, if that's what you're asking. Don't know what you can or can not do with other tools.
madshi is offline   Reply With Quote
Old 1st February 2009, 00:34   #8092  |  Link
Snowknight26
Registered User
 
Join Date: Aug 2007
Posts: 1,381
Quote:
Originally Posted by madshi View Post
Maybe the spec supports it, but I think the official encoder does not. See here:

http://www.hydrogenaudio.org/forums/...hp/t23613.html

It's possible that this post is outdated, though, not sure...
Oh well, not like it was an issue to begin with.
Snowknight26 is offline   Reply With Quote
Old 1st February 2009, 05:19   #8093  |  Link
Blue_MiSfit
Derek Prestegard IRL
 
Blue_MiSfit's Avatar
 
Join Date: Nov 2003
Location: Los Angeles
Posts: 5,307
Hi Madshi,

Did you get a chance to peek at my problematic MP2? It's the one with clipping etc..

Thanks,
~MiSfit
Blue_MiSfit is offline   Reply With Quote
Old 1st February 2009, 08:39   #8094  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
Quote:
Originally Posted by Blue_MiSfit View Post
Did you get a chance to peek at my problematic MP2? It's the one with clipping etc..
Yes. It decodes without clipping when using "-nero" or "-sonic", so the issue is related to libav decoding. I'm checked and libav returns the data to me with clipping in the data and I don't really see how I can detect and remove that. I think I'll have to update to the latest libav SVN version and maybe do some patches to get floating point from libav instead of integer samples. Then I might get it solved. However, this is not likely to happen this week. Maybe next. For now you can use Nero or Sonic for decoding...
madshi is offline   Reply With Quote
Old 1st February 2009, 08:57   #8095  |  Link
Snowknight26
Registered User
 
Join Date: Aug 2007
Posts: 1,381
Madshi, if you are thinking about doing that (the patch for floating point support), you might wish to submit your changes to the SVN as well, for everyone's benefit.
Snowknight26 is offline   Reply With Quote
Old 1st February 2009, 18:49   #8096  |  Link
GZZ
Registered User
 
Join Date: Jan 2002
Posts: 433
I think I found a small cosmetic error when parsing a playlist file:

Result:

Quote:
d:\eac3to>eac3to "H:\BDMV\PLAYLIST\00000.mpls"
1) 00000.mpls, 0:50:10
[4+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+
5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+
5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5+5].m2ts
- h264/AVC, 1080p24 /1.001 (16:9)
- RAW/PCM, English, multi-channel, 48khz
Not sure it should say 5+5+5+5+5+5...... it should just be [4+5].m2ts

I have attached the playlist file here, for you to test it.

http://hbar.dk/files/00000.rar
GZZ is offline   Reply With Quote
Old 1st February 2009, 20:26   #8097  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
Quote:
Originally Posted by GZZ View Post
I think I found a small cosmetic error when parsing a playlist file:

Not sure it should say 5+5+5+5+5+5...... it should just be [4+5].m2ts
Some playlists are really that stupid. I don't think it's a bug.
madshi is offline   Reply With Quote
Old 1st February 2009, 20:39   #8098  |  Link
madshi
Registered Developer
 
Join Date: Sep 2006
Posts: 8,985
eac3to v3.06 released

http://madshi.net/eac3to.zip

Code:
* added MKV reading/parsing support
* added demux support for MKV (E-)AC3, DTS(-HD), AAC, MPx, FLAC and WAV tracks
* added demux support for MKV "modern style" MPEG2, VC-1 and h264/AVC tracks
* reading from (HD) DVD and Blu-Ray drives uses different reading APIs now
* empty tracks in TS/m2ts container are not listed, anymore
* for 24.000 fps video tracks a little warning is displayed now
* when demuxing subtitle files, the number of captions is added to the filename
* timestamp derived FPS is used for gap checking instead of video bitstream FPS
* fixed: 44.1khz AC3 encoding was still broken
* fixed: zero byte stripping pass was done for true 24bit TrueHD tracks
* fixed: downconverting WAV files with 0x3f channel mask didn't work
* fixed: log output "remaining delay [...]" was sometimes wrong for AC3 tracks
* fixed: silent frame creation was tried for E-AC3 although it can't work
Please note that MKV read support is not complete yet. E.g. subtitles and chapters can't be demuxed yet. Also some MKV files muxed with old versions of mkvtoolnix, gdsmux or other muxing tools may not work correctly. If you run into any trouble with audio/video tracks, just let me know.

Good news is that MKV video muxing and demuxing is now a transparent operation, if you use eac3to both for muxing + demuxing. In other words: If you mux a video track with eac3to and then demux it again, you'll get the exact same data back. No changes, nothing missing... *)

-------

*) exceptions:

(1) of course eac3to does some video bitstream manipulations like cropping 1088 to 1080, removing padding bytes etc, but these changes are done intentionally and are not caused by muxing/demuxing
(2) VC-1 muxing loses the very last video frame, don't know why
madshi is offline   Reply With Quote
Old 1st February 2009, 20:46   #8099  |  Link
d1g1ta7
Registered User
 
Join Date: Apr 2006
Posts: 19
On the NBC HDTV muxes used to distribute programming to network affiliates in the US, NBC sends the surround audio via 3 stereo tracks - the first contains front left and front right audio, the second contains the center and lfe, the third contains rear left and rear right. Would it be possible to combine these tracks in eac3to, outputting as 5.1?

Last edited by d1g1ta7; 1st February 2009 at 20:49.
d1g1ta7 is offline   Reply With Quote
Old 1st February 2009, 20:47   #8100  |  Link
Snowknight26
Registered User
 
Join Date: Aug 2007
Posts: 1,381
Quote:
C:\unzipped\eac3to>eac3to.exe "Z:\Movies\1408\1408.mkv"
MKV, 1 video track, 1 audio track, 16 subtitle tracks, 1:52:26, 24p /1.001
1: V_MPEG4/ISO/AVC, English, 1920x800p (12:5)
2: DTS, English, 5.1 channels, 16 bits, 1509kbps, 48khz
3: S_TEXT/ASS, Finnish, "Finnish"
4: S_TEXT/ASS, Finnish, "Finnish"
5: S_TEXT/ASS, Finnish, "Finnish"
6: S_TEXT/ASS, Finnish, "Finnish"
7: S_TEXT/ASS, Finnish, "Finnish"
8: S_TEXT/ASS, Finnish, "Finnish"
9: S_TEXT/ASS, Finnish, "Finnish"
10: S_TEXT/ASS, Finnish, "Finnish"
11: S_TEXT/ASS, Finnish, "Finnish"
12: S_TEXT/ASS, Finnish, "Finnish"
13: S_TEXT/ASS, Finnish, "Finnish"
14: S_TEXT/ASS, Finnish, "Finnish"
15: S_TEXT/ASS, Finnish, "Finnish"
16: S_TEXT/ASS, Finnish, "Finnish"
17: S_TEXT/ASS, Finnish, "Finnish"
18: S_TEXT/ASS, Finnish, "Finnish"
Only the last track is Finnish.

And then funny enough, the next mkv I had muxed says
Quote:
C:\unzipped\eac3to>eac3to.exe "Z:\Movies\11-14\11-14.mkv"
MKV, 1 video track, 1 audio track, 1:25:43, 23.975p
1: V_MPEG4/ISO/AVC, English, 1080p (16:9)
2: DTS, English, 5.1 channels, 24 bits, 1509kbps, 48khz
And finally..
Quote:
C:\unzipped\eac3to>eac3to.exe "Z:\Movies\The Third Man\the third man.mkv"
MKV, 1 video track, 4 audio tracks, 1:45:13, 24p /1.001
1: V_MPEG4/ISO/AVC, English, 1436x1080p (359:270)
2: FLAC, English, 1.0 channels, 1:45:13, 16 bits, 48khz
"FLAC 1.0 @ 202kbps"
3: A_VORBIS, English, 1.0 channels, 48khz
"Abridged recording of Graham Greene∆s treatment, read by actor Richard Clarke"
4: A_VORBIS, English, 1.0 channels, 48khz
"Abridged recording of Graham Greene∆s treatment, read by actor Richard Clarke"
5: A_VORBIS, English, 1.0 channels, 48khz
"Abridged recording of Graham Greene∆s treatment, read by actor Richard Clarke"
Only the last track has that description. And that ∆ is really a ī. Guess eac3to doesn't support unicode?

Samples can be provided if needed.

Last edited by Snowknight26; 1st February 2009 at 21:52.
Snowknight26 is offline   Reply With Quote
Reply

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 11:54.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2017, vBulletin Solutions Inc.