eac3to - audio conversion tool

[madshi]

The 3/1.0 channel configuration is not supported by flac encoder, for 4 channels only 2/2.0 is allowed.

You can force flac to encode a wav 3/1.0 using:
--channel-map=none
That's when using the official flac encoder executable file, right? I'm using libFlac. I think libFlac allows me to encode any number of channels between 1..8.

But when you decode the flac, a 2/2.0 channelmask can be send to the player and listen the center channel in left surround speaker.
You seem to be right. The FLAC documentation sais this:

Channel assignment
* 0000-0111 : (number of independent channels)-1. Where defined, the channel order follows SMPTE/ITU-R recommendations. The assignments are as follows:
o 1 channel: mono
o 2 channels: left, right
o 3 channels: left, right, center
o 4 channels: left, right, back left, back right
o 5 channels: left, right, center, back/surround left, back/surround right
o 6 channels: left, right, center, LFE, back/surround left, back/surround right
o 7 channels: not defined
o 8 channels: not defined
* 1000 : left/side stereo: channel 0 is the left channel, channel 1 is the side(difference) channel
* 1001 : right/side stereo: channel 0 is the side(difference) channel, channel 1 is the right channel
* 1010 : mid/side stereo: channel 0 is the mid(average) channel, channel 1 is the side(difference) channel
* 1011-1111 : reserved
So as far as I can see if I encode 4 channels, FLAC will see that as 2/2.0.

However, I've once seen a FLAC file with a Windows WAV style channelmask field in the metadata. Now I don't know whether this is a secretly supported option (having channelmask information in the metadata). Probably not. So I guess FLAC doesn't like 3/1.0.

EDIT: To use flac with 3/1.0 channels maybe we need split the surround channel in:
BL = 0.707 x S
BR = 0.707 x S
and use a 3/2.0 configuration supported by flac.
Yeah, sounds like a good workaround.

[tebasuna51]

However, I've once seen a FLAC file with a Windows WAV style channelmask field in the metadata. Now I don't know whether this is a secretly supported option (having channelmask information in the metadata). Probably not. So I guess FLAC doesn't like 3/1.0.

Using WAVE_FORMAT_EXTENSIBLE header wav's with the ChannelMask set properly I obtain:
C-0.0.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x0004
L-0.0.wav: wrote 116967 bytes, ratio=0,061 (OK mask=0x0001)
1-0.1.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x000C
2-0.0.wav: wrote 276354 bytes, ratio=0,072 (OK mask=0x0003)
2-0.1.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x000B
2-1.0.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x0103
3-0.0.wav: wrote 382988 bytes, ratio=0,066 (OK mask=0x0007)
2-1.1.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x010B
2-2.0.wav: wrote 525781 bytes, ratio=0,068 (OK mask=0x0033)
3-0.1.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x000F
3-1.0.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x0107
2-2.1.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x003B
3-1.1.wav: ERROR: WAVEFORMATEXTENSIBLE chunk with unsupported channel mask=0x010F
3-2.0.wav: wrote 632393 bytes, ratio=0,066 (OK mask=0x0037)
3-2.1.wav: wrote 737200 bytes, ratio=0,064 (OK mask=0x003F)

BTW, when flac decode to wav these files make standard pcm wav file headers for mono and stereo.
If channels are >2 use WAVE_FORMAT_EXTENSIBLE header wav's with the ChannelMask set properly for 3 and 4 channels (always 3/0.0 and 2/2.0), but with 5 and 6 channels I obtain:

mask=0x0037 (FL FR FC BL BR) -> flac -> mask=0x0607 (FL FR FC SL SR)
mask=0x003F (FL FR FC LF BL BR) -> flac -> mask=0x060F (FL FR FC LF SL SR)

Seems flac developpers mistake the:
"back/surround left, back/surround right" (BL, BR)
with the
"SideLeft, SideRight" (SL, SR)

[madshi]

Seems flac developpers mistake the:
"back/surround left, back/surround right" (BL, BR)
with the
"SideLeft, SideRight" (SL, SR)
Yep, seems like that!

[kmitalian]

I shall try all different kinds of WAV formats to see which it accepts, i have tried converting it to LPCM and it wouldnt accept that, so strike that!

[PHD_1976]

What kind of commands do i need to transcode a DolbyTrueHD track into LPCM Multichannel so its compliant for Scenarist to accept that file as a multi-channel PCM so I can make a Blu-Ray disc out of it?

You should type the following:
eac3to inputfile.thd outputfile.wavs
and then run each .wav file through MUI Generator via multichannel LPCM input (available since 4.3 version)

If TrueHD track is taken from Blu-Ray you can import it to Scenarist as TrueHD stream instead of making PCM from it, but not without the help of madshi's great tool.:)

[scarbrtj]

Nero is the default EAC3 decoder for eac3to. Casablanca HD-DVD English track is a 64kbs mono EAC3 track. The Nero decoding via eac3to leaves the resultant AC3 track with "shrieks" and "whistles" and "crackles" but it's rather intermittent. However, invoking "-sonic" makes an AC3 file which is smooth, no noise.

Why would this be?

[kmitalian]

I have tried importing the DolbyTrueHD signal into the MUI Generator but it wont work, its due that the Dolby TrueHD tracks i have aernt coming up as TrueHD, they are coming up as MLP?

the reason why PCM is better is cause it still decodes it into its Full Form and i do have a Sony BDPS1000 i can test it on!!!

[PHD_1976]

I have tried importing the DolbyTrueHD signal into the MUI Generator but it wont work, its due that the Dolby TrueHD tracks i have aernt coming up as TrueHD, they are coming up as MLP?

the reason why PCM is better is cause it still decodes it into its Full Form and i do have a Sony BDPS1000 i can test it on!!!

In order to import TrueHD stream from BD to Scenarist HDMV you have to do the following:
- run TrueHD track via eac3to twice:
1. eac3to track.thd audio1.ac3
2. eac3to track.thd audio1.thd
Then rename audio1.thd to .mlp and feed MUI Generator with audio1.ac3 specifying it as Dolby Lossless.
In next window be sure, that the path to your audio1.mlp is correct. Scenarist should be happy with that.

PCM needs lot of extra space, so sometimes movie doesn't fit BD-25 because of it.

[madshi]

Nero is the default EAC3 decoder for eac3to. Casablanca HD-DVD English track is a 64kbs mono EAC3 track. The Nero decoding via eac3to leaves the resultant AC3 track with "shrieks" and "whistles" and "crackles" but it's rather intermittent. However, invoking "-sonic" makes an AC3 file which is smooth, no noise.

Why would this be?
Can I have a sample, please? You could also try "-libav" to cross check whether it's a problem with the Nero decoder or whether the Sonic decoder does some magic which both libav and Nero don't do.

[earl]

i will demux 2 .evo files from hdd1 to hdd2
eac3to d:\part1.evo+d:\part2.evo -demux e:\
does not work, eac3to put the demuxed files in d:\ and not to e:\
.. why ? .. what make i wrong ?

[madshi]

i will demux 2 .evo files from hdd1 to hdd2
eac3to d:\part1.evo+d:\part2.evo -demux e:\
does not work, eac3to put the demuxed files in d:\ and not to e:\
.. why ? .. what make i wrong ?
Change the current directory to "e:\" (by executing "e:\" and then "cd \" in the command prompt), then run again. When using the "-demux" option, eac3to always demuxes to the current directory.

[Chumbo]

Just wanted to report another "unclean" track but the source this time is an HD DVD. I did demux the track to a .ddp file using EVOdemux, but eac3to 2.36 still wasn't able to convert it.E-AC3, 5.1 channels, 2:06:38, 1536kbit/s, 48khz, dialnorm: -27dB
Removing dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
Disabling DRC for Nero (E-)AC3 decoding...
This track is not clean. Processing aborted.
Please clean the track with delaycut and then retry eac3to.
DirectShow reports 5.1 channels, 24 bits, 48khz
Writing WAVs...
I tried a few older versions too, but they didn't work either. Just fyi. Thanks.

[madshi]

Just wanted to report another "unclean" track but the source this time is an HD DVD. I did demux the track to a .ddp file using EVOdemux, but eac3to 2.36 still wasn't able to convert it.E-AC3, 5.1 channels, 2:06:38, 1536kbit/s, 48khz, dialnorm: -27dB
Removing dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
Disabling DRC for Nero (E-)AC3 decoding...
This track is not clean. Processing aborted.
Please clean the track with delaycut and then retry eac3to.
DirectShow reports 5.1 channels, 24 bits, 48khz
Writing WAVs...
I tried a few older versions too, but they didn't work either. Just fyi. Thanks.
The "unclean" detection for E-AC3 tracks is extremely reliable cause E-AC3 is protected by a CRC. So unless both eac3to and EvoDemux demux the track incorrectly, it is really damaged. Try running it through delaycut 1.3.

[moshmothma]

Madshi, I see that you have replaced the haali source and splitter with your own. Any chance we could see that in a directshow version? If not any chance you could release the code for someone else to port? Thanks as always for your incredible software!!!

[earl]

Change the current directory to "e:\" (by executing "e:\" and then "cd \" in the command prompt), then run again. When using the "-demux" option, eac3to always demuxes to the current directory.

thanks, it works so

i use eac3to from gentoo linux with wine
it is a litle bit tricky, but it works .. i run with wine the cmd.exe, and from there i use eac3to.exe - it works
the directly way with wine and eac3to.exe work also
but with the directly way i have ever the demuxed files in the same directory how the .m2ts/.evo file(s) are
is there no other way to specify the directory for the demuxed files ?

..will a linux version from eac3to come ?


:edit
gkrellm shows me...
wine + eac3to.exe -> 70..90mb/s read (eac3to 1.evo+2.evo) (20% cpu load)
wine + eac3to.exe -> 25..35mb/s read and write (eac3to 1.evo+2.evo -demux) (40% cpu load)
wine + cmd.exe + eac3to.exe -> 10..15mb/s (eac3to 1.evo+2.evo -demux) (75% cpu load)
(c2d 1,86ghz 3x seagate .11 series)


:edit2

c:\programme\eac3to\eac3to.exe z:\movie.m2ts -demux -ignoreDiscon
Hooking GetModuleFileNameA failed (Erfolg).
M2TS, 1 video track, 5 audio tracks
1: h264/AVC, 1080p24 /1.001 (16:9)
2: DTS Master Audio, 6.1 channels, 24 bits, 48khz
3: AC3, 5.1 channels, 448kbit/s, 48khz, dialnorm: -27dB
4: AC3, 5.1 channels, 448kbit/s, 48khz, dialnorm: -27dB
5: AC3, 2.0 channels, 224kbit/s, 48khz, dialnorm: -27dB
6: AC3, 2.0 channels, 224kbit/s, 48khz, dialnorm: -27dB
Hooking WriteConsoleA failed (Erfolg).
Extracting primary video track...
Extracting audio track number 2...
Extracting audio track number 3...
Extracting audio track number 4...
Extracting audio track number 5...
Extracting audio track number 6...
Removing dialog normalization...
Removing dialog normalization...
Removing dialog normalization...
Removing dialog normalization...
Creating file "movie - 1 - h264, 1080p24.h264"...
Creating file "movie - 2 - DTS Master Audio, 6.1 channels, 24 bits, 48khz.dtsma"...
Creating file "movie - 4 - AC3, 5.1 channels, 448kbps, 48khz.ac3"...
Creating file "movie - 3 - AC3, 5.1 channels, 448kbps, 48khz.ac3"...
Creating file "movie - 5 - AC3, 2.0 channels, 224kbps, 48khz.ac3"...
Creating file "movie - 6 - AC3, 2.0 channels, 224kbps, 48khz.ac3"...
This TS/M2TS file seems to be damaged (discontinuity).
This TS/M2TS file seems to be damaged (discontinuity).
Audio overlaps for 27 frames at playtime 1:44:05.
This TS/M2TS file seems to be damaged (discontinuity).
This doesn't seem to be a valid PES packet.

what can i do to demux this .m2ts file ?

[Chumbo]

The "unclean" detection for E-AC3 tracks is extremely reliable cause E-AC3 is protected by a CRC. So unless both eac3to and EvoDemux demux the track incorrectly, it is really damaged. Try running it through delaycut 1.3.
Thanks, will do.

May I make a suggestion please? Any chance you can change the logging to use the source folder and source file name and add .log.txt to the end? I use piping to log the info but then have to clean the piped file.

I know you create a log in the eac3to folder, but it's tedious to get it and sometimes running batch files with separate eac3to commands will only keep the last log entry since the previous commands' output log is overwritten. That means placing extra commands to rename the file and so on.

So...please, please, please...:)

[madshi]

Madshi, I see that you have replaced the haali source and splitter with your own. Any chance we could see that in a directshow version?
Not really.

If not any chance you could release the code for someone else to port?
Not sure about that. If a programmer asked for it, I'd think about that, depending on the cirumstances and conditions etc...

But honestly, the Haali splitter is not bad at all. It currently has some limitations (e.g. no TrueHD splitting), but apart from that for real time playback I think the Haali splitter works just fine. The latest version has fixed some important issues. So I'm not sure if it would be worthwhile to create a new DirectShow filter...

[madshi]

is there no other way to specify the directory for the demuxed files ?
I don't really see why it's necessary. Even if you don't use cmd.exe you should be able to call wine in such a way that you can define the "current directory" or "working directory" or however it's called in Linux.

..will a linux version from eac3to come ?
No. You just reported the the Windows version works fine with Wine... ;)

what can i do to demux this .m2ts file ?
You have 2 choices:

(1) Rerip the movie. Probably it's a damaged rip.
(2) If the movie is authored incorrectly (shit happens) you can use xport for demuxing. However, you'll have to live with damaged video/audio data then. No way around that.

[madshi]

May I make a suggestion please? Any chance you can change the logging to use the source folder and source file name and add .log.txt to the end? I use piping to log the info but then have to clean the piped file.

I know you create a log in the eac3to folder, but it's tedious to get it and sometimes running batch files with separate eac3to commands will only keep the last log entry since the previous commands' output log is overwritten. That means placing extra commands to rename the file and so on.

So...please, please, please...:)
Why in the source path? Wouldn't it make more sense to write the log file to the destination path? I mean the source path might not even be writable while the destination path always has to be writable. Of course if you ask eac3to to just check a file things are getting complicated because then there is no destination path. Well...

[earl]

You have 2 choices:

(1) Rerip the movie. Probably it's a damaged rip.
(2) If the movie is authored incorrectly (shit happens) you can use xport for demuxing. However, you'll have to live with damaged video/audio data then. No way around that.

you mean this -> http://www.videohelp.com/tools/xport ?
is there no way to demux my .m2ts file with eac3to ?

[madshi]

you mean this -> http://www.videohelp.com/tools/xport ?
Yes.

is there no way to demux my .m2ts file with eac3to ?
eac3to strictly wants clean rips. It refuses to work on damaged source files. xport is more relaxed and also allows damaged source files. Your source file seems to be damaged. So as I said in my previous comment, you can either rerip the movie to (hopefully) get a clean rip. Or you can use xport for demuxing. But if you do use xport on the damaged rip, obviously the demuxed files you'll get will contain some corruption.

[Chumbo]

Why in the source path? Wouldn't it make more sense to write the log file to the destination path? I mean the source path might not even be writable while the destination path always has to be writable. Of course if you ask eac3to to just check a file things are getting complicated because then there is no destination path. Well...
Good grief, yes. I meant destination. Ugh. Thanks for catching that. :)

I guess if there's no destination path, you can always fall back on where eac3to is or provide a log file switch where we can enter specifically the path and name of the log file. That would at least save you the headache of having to do too many if-this-or-that checks.

...things are getting complicated...
That cracked me up madshi. Like the stuff you've done isn't complicated enough, to hear you say the log option is complicated was pretty funny. ;) Just incase someone misunderstands my jest here, I hope you understand the spirit in which I'm saying this, i.e., I have a lot of respect for the things mashi's accomplished in this program and all its complexities.

[EDIT] btw, I got my hands on delaycut 1.3 and did run the ddp file through it to fix the CRC errors and eac3to was able to transcode after it was fixed. Also, I was only able to demux it with EVOdemux. Do you think it's worth allowing eac3to to demux tracks even if they're "dirty" since in some cases, like this one, the track can be fixed?

[moshmothma]

Not sure about that. If a programmer asked for it, I'd think about that, depending on the cirumstances and conditions etc...

But honestly, the Haali splitter is not bad at all. It currently has some limitations (e.g. no TrueHD splitting),

Madshi, I don't know how difficult it would be to do this but there are a couple of reasons....

1. You are far more in tune with HDDVD and Bluray codecs and concerns. You are also better plugged into this community.
2. Haali is not exactly the most responsive agent in the world ;) I don't fault him (and in fact am very grateful for his filters) but it is frustrating communicating with him.
3. It is always good to have an alternative and competition. Different perspectives promote growth.

These are mainly selfish but say a lot about you and your efforts. I hope you will consider this request.

[bmnot]

Was just testing eac3to's blu-ray detection abilities.

Loaded in Queen Rock Montreal and it said for the DTS track
3: DTS Hi-Res, English, 5.1 channels, 24 bits, 3072kbit/s, 48khz

This is incorrect, the frequency is 96kHz.

[madshi]

Do you think it's worth allowing eac3to to demux tracks even if they're "dirty" since in some cases, like this one, the track can be fixed?
I already posted some time before that maybe I'll add demuxing support (but probably not transcoding support) for dirty tracks some day.

Madshi, I don't know how difficult it would be to do this but there are a couple of reasons....

1. You are far more in tune with HDDVD and Bluray codecs and concerns. You are also better plugged into this community.
2. Haali is not exactly the most responsive agent in the world ;) I don't fault him (and in fact am very grateful for his filters) but it is frustrating communicating with him.
3. It is always good to have an alternative and competition. Different perspectives promote growth.

These are mainly selfish but say a lot about you and your efforts. I hope you will consider this request.
I'm sorry, but my answer is still no. In the commercial world competition can be a motivator for more development. In the freeware world competition often is a waste of resources IMO. I don't have the ambition to "beat" other programmers. And I will only duplicate other programmers' efforts only if I could really make a noticable improvement - and if I'm interested in that improvement myself. Most of my motivation for developing eac3to is just to satisfy my own needs. But personally I'm (more or less) satisfied with Haali's splitter for real time playback. So that's why I don't plan on my writing my own DirectShow splitter...

Was just testing eac3to's blu-ray detection abilities.

Loaded in Queen Rock Montreal and it said for the DTS track
3: DTS Hi-Res, English, 5.1 channels, 24 bits, 3072kbit/s, 48khz

This is incorrect, the frequency is 96kHz.
As has been said before, there's no documentation whatsoever available on DTS-HD. Because of that eac3to just reads out the parameters of the DTS core. In some cases the core parameters differ from the DTS-HD parameters. But don't worry, if you decode this track with the Sonic Audio Decoder, you'll correctly get 96khz. So it's really just a very minor problem.

[Rectal Prolapse]

I actually believe with moshmothma's sentiments - Haali, with all the great work he's done, isn't as responsive as we'd like (for many legitimate reasons, for sure!). LPCM playback from m2ts is still broken, and has been broken for nearly a year, is one example.

With that said, I'm hoping that the TS splitter in MPC could be modified for proper m2ts playback - the Haali Splitter I think does "too much" and that makes it much harder to get fixes incorporated and tested.

I'm sure Haali is busy with commercial products as well (isn't he involved with MKV implementations on the various media center hardware out there?).

As a programmer, I would be very interested in knowing how the Nero and Sonic decoders can be manipulated into providing the proper output - no need for you madshi to release source I think. :) Such information would be useful for creating a splitter that can interface with these decoders while preserving quality.

At the moment I'm busy with another project, so the above is just an example of what someone may request without actually getting source from madshi.

[madshi]

I actually believe with moshmothma's sentiments - Haali, with all the great work he's done, isn't as responsive as we'd like (for many legitimate reasons, for sure!). LPCM playback from m2ts is still broken, and has been broken for nearly a year, is one example.
Broken LPCM playback is really bad, I agree with that. However, Haali has been responsive to me when I reported bugs with exact information what exactly is wrong and how to fix it. His latest version fixes several important bugs I reported to him.

[DeepBeepMeep]

I think we shouldn't under estimate the work to create a good Direct Show splitter. Such a splitter must be capable to estimate the duration of a video file (even with damaged files), support fast seeking, support various sub formats (e.g for transport streams: ATSC, DVB-T, M2TS, etc...).

It seems the latest Haali filter has fixed most of past reported bugs (beside the LPCM one). I think if madshi has some free time, It would be great if before anything else he could spend it on supporting multiparts m2ts files and muxing directly video and audio tracks into an mkv file.

[hero88]

I think i have found a bug:

I wanted to convert track 3 TRUEHD to PCM and I accidentally wrote :3uk.pcm instead of 3: uk.pcm

Then eac3to does what it should but creates a file called ":3uk.pcm" which is an invalid filename in windows, so you cannot find it afterwards, it does create the file as the diskspace was reduced by 3 gig afterwards

This is input to / output from eac3to

eac3to.exe 00000.m2ts :3uk.pcm
M2TS, 1 video track, 4 audio tracks
1: VC-1, 1080p24 /1.001 (16:9)
2: AC3, English, 5.1 channels, 640kbit/s, 48khz, dialnorm: 27dB
3: TrueHD/AC3, English, 5.1 channels, 48khz, dialnorm: -27dB
Extracting audio track number 2...
Removing dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
Disabling DRC for Nero (E-)AC3 decoding...
DirectShow reports 5.1 channels, 24 bits, 48khz
Swapping endian...
Remapping channels...
Creating file ":3uk.pcm"...
eac3to processing took 13 minutes, 24 seconds.
Done.

Suggestion: a small check of filename to ensure it doesn't contain invalid characters like " ? / \ < >

[nautilus7]

And it would be even better if madshi could make directshow decoders instead of a splitter. Such thing isn't available yet. :cool:

[madshi]

Then eac3to does what it should but creates a file called ":3uk.pcm" which is an invalid filename in windows, so you cannot find it afterwards, it does create the file as the diskspace was reduced by 3 gig afterwards
Actually it's a valid filename. It creates an NTFS substream for the current folder... ;) See here:

http://www.ntfs.com/ntfs-multiple.htm

[hero88]

Ahh, you live you learn

Any idea on how to get to it? So I can delete it :)

[BLKMGK]

It seems like we're hearing lots of folks wanting Madshi to go one direction or another with his code and his time. I for one would like to simply say thank you for the code he has provided and say that I think he's moving in a direction that will continue to keep *me* pretty happy.

My primary motivation for using eac3to has been to convert HD-DVD disks to a format I can compress and then playback on my HTPC using Linux (XBMC). Now that HD-DVD is dead <sob> Madshi has begun to spend some time on the M2TS format and he's made progress. Without his work on the EVO format conversion of my library, which has grown thanks to price cuts, would've been much much much harder. His tool does much of the heavy lifting for me and for that I am very thankful. M2TS still cause me grief but I'm confident that as time passes his tool and others will mature, thankfully I've only got just the one BD disk to deal with right now. Prices being what they are I have plenty of time to hone my technique!

Anyway, I recognize this is a spare time labor of love for Madshi but his work has made my life easier and in the midst of folks asking him to go this way or that I wanted to make sure that he knows that many have greatly appreciated everything he has done so far. Yeah, there are directions I'd like to see him go too but I'm patient and if the need is great enough it'll be filled. Thankfully stuff I need seems to be pretty basic. <shrug> These formats and the ability to rip them are pretty new, I'm thrilled to be able to manipulate them to the extent I can.

[madshi]

Any idea on how to get to it? So I can delete it :)
Just delete the directory from which you called eac3to. That should also get rid of the NTFS sub stream.

It seems like we're hearing lots of folks wanting Madshi to go one direction or another with his code and his time.
Fortunately I'm stubborn... ;)

M2TS still cause me grief
Is there any grief besides those Blu-Ray titles with multiple parts? That's the only problem I'm aware of.

in the midst of folks asking him to go this way or that I wanted to make sure that he knows that many have greatly appreciated everything he has done so far.
Thanks.

[madshi]

eac3to v2.37 released

http://madshi.net/eac3to.zip

* added HD DVD chapter support
* added HD DVD subtitle demuxing support
* added pre-freeze detection for Haali Matroska Muxer bug
* invalid characters are removed from file names now
* log file is copied to destination path (of first destination file)

Edit: Fixed file path handling problem in v2.38.

[nautilus7]

Thanks for update, but i don't know how to use subs and chapters. Are they extracted only with -demux option? I used an evo file, but there wasn't a option in the track list for those.

[BLKMGK]

Chapter support, excellent I had someone asking me about that the other day and had to admit to being clueless about it, I'll tell them it's there! :p

BD disks, indeed the multiple versions that can be had on a single disk are an issue. It would be nice if the various paths could be presented like the languages are when info is gotten about a set of files. That said, the one BD disk I've got only has like 2 or 3 files so figuring out it's path was quite easy unlike some others I've heard about that are the stuff of nightmares with 20 or more pieces.:eek:

When I do HD-DVD it's simple, I use eac3to to pull vid out and put a vid stream into a MKV. meGUI then compresses the MKV further with x264, I combine the compressed file and AC3 audio with MKVmerge (I use -libav and 640), and then I'm done. No synch issues or other problems it just works.

When I tried to do much the same with M2TS files, admittedly a few weeks ago, things were not so smooth and the path slightly different. The resulting file had it's time displayed incorrectly and long seeks crashed my player - jumping to advance in the video could also crash it. At the end of the file the screen went black before the video was done and I just got audio. I do not know where the problem lies, my player is certainly suspect as it's under heavy development (http://xbmc.org/forum/forumdisplay.php?f=52&order=desc). I seem to recall using some other tools in the mix too unlike what I do with HD-DVd so the issue could have been there too.

Tell you what, I will redo the BD video attempting to use much the same process that I use with HD-DVD today (same AVIsynth script for instance) and the latest eac3to and let you know how it goes. As it happens I REALLY like the video in question (The Patriot) and would love to be able to watch it all the way through! (lol)

Edit: Uh Oh! Madshi, the new version of eac3to will NOT run under 64bit Vista. When I attempt to start it I get a message telling me it's not compatible - never seen this before!
D:\Video\eac3to>eac3to
This version of D:\Video\eac3to\eac3to.exe is not compatible with the version of Windows you're running. Check your computer's system information to see whether you need a x86 (32-bit) or x64 (64-bit) version of the program, and then contact the software publisher.

[Henry The Ripper]

When I demuxed PCM from m2ts, it says "Remapping channels", then when encoding to DTS it's again "Remapping channels"??

So is the result the incorrect order?

2.37 and 2.38 don't start the encoding procedure with normal command line "eac3to source.pcm destination.dts"

[Encoder888]

@madshi

Just curious, are you planning on including support for Blu-Ray chapters at some point too? No rush, just wondering :) Thanks.

[Denner]

I am having problems with TOP GUN (HD-DVD), eac3to reports an audio delay of -84ms, an does its thing when re-encoding the Dolby True HD audio to FLAC.

But when I mux audio and video to MKV, the audio is out of sync (the audio is to early) and stays out of sync for the whole movie.

The HD-DVD plays fine both from the HD-DVD and from my harddrive.

This is the first time I am having audio sync problems with a FLAC file created by eac3to and the muxed in to an MKV file.

Have any of you re-encodet TOP GUN to MKV and if so did you experience any problem doing so?

[Denner]

SORRY double post :-(

[madshi]

Thanks for update, but i don't know how to use subs and chapters. Are they extracted only with -demux option? I used an evo file, but there wasn't a option in the track list for those.
Chapters are only available if the VPLST000.XPL file is there. Subtitle tracks don't need that file, but eac3to can only list the subtitle track languages if the VPLST000.XPL is there. Handling chapters and subtitles works the same way as audio and video files. They're listed with a track number, they're demuxed when using the "-demux" option and you can also manually handle them. Example:

EVO, 1 video track, 3 audio tracks, 8 subtitle tracks, 1:35:56
"mainMovie"
1: Joined EVO file
2: Chapters, 12 chapters with names
3: VC-1, 1080p24 /1.001 (16:9)
4: DTS, French, 5.1 channels, 16 bits, 1536kbit/s, 48khz
5: DTS Master Audio, English, 5.1 channels, 16 bits, 48khz
6: DTS, German, 5.1 channels, 16 bits, 1536kbit/s, 48khz
7: Subtitle, French
8: Subtitle, English
9: Subtitle, German
Here you could e.g. do this:

eac3to source.evo 2: chapters.txt 9: German.sup

Edit: Uh Oh! Madshi, the new version of eac3to will NOT run under 64bit Vista. When I attempt to start it I get a message telling me it's not compatible - never seen this before!
Strange. Is that v2.37 or v2.38?

When I demuxed PCM from m2ts, it says "Remapping channels", then when encoding to DTS it's again "Remapping channels"??

So is the result the incorrect order?
Why are you doing this in two steps? You can directly demux + transcode to DTS in one step! Anyway, don't worry about the multitude of "Remapping channels" messages. Usually the end result is just fine. Of course it would make sense to double check, just to be sure.

2.37 and 2.38 don't start the encoding procedure with normal command line "eac3to source.pcm destination.dts"
What do v2.37 and v2.38 do?

Just curious, are you planning on including support for Blu-Ray chapters at some point too? No rush, just wondering :) Thanks.
I'm not sure right now whether Blu-Ray chapters are stored. I'm planning to add Blu-Ray subtitle demuxing. But I'm not sure if it's easily possible to extract chapters.

I am having problems with TOP GUN (HD-DVD), eac3to reports an audio delay of -84ms, an does its thing when re-encoding the Dolby True HD audio to FLAC.

But when I mux audio and video to MKV, the audio is out of sync (the audio is to early) and stays out of sync for the whole movie.

The HD-DVD plays fine both from the HD-DVD and from my harddrive.

This is the first time I am having audio sync problems with a FLAC file created by eac3to and the muxed in to an MKV file.

Have any of you re-encodet TOP GUN to MKV and if so did you experience any problem doing so?
There is a small number of titles where audio is not in sync. Bourne Supremecy is another example. I don't know why this happens. But usually it's easy to enough to correct the audio sync. I'd suggest playing the resulting MKV file with MPC HC. You can then at runtime correct audio sync by using the "+" and "-" keys in the num block. Once you found the correct audio delay, you can use eac3to to apply that delay to the FLAC track (e.g. "eac3to source.flac dest.flac +100ms") and finally mux the corrected FLAC track to the final MKV file.

[bmnot]

Good update.

What files do I need for chapter support? For a while I was removing all files except for the main .evos for a few titles. I'd like to make mkvs from them now. I can rerip the files needed for chapter support if needed.

[bmnot]

I'm not sure right now whether Blu-Ray chapters are stored. I'm planning to add Blu-Ray subtitle demuxing. But I'm not sure if it's easily possible to extract chapters.

BDEdit can extract chapters to a text file from the .mpls playlists.

[BLKMGK]

Madshi the link (http://forum.doom9.org/showthread.php?p=1122525#post1122525) said 1.37 but checking the file on a 32bit XP system is says 1.38. Was something done differently on this binary?

[Yraen]

madshi,

Are there any plans to make chapters and subtitles list when you list the contents of a folder? Or just when listing the contents of the .evos? I ask this because the gui doesn't "see" the chapters and subtitles until it is too late.


eac3to v2.38
command line: "D:\editing apps\eac3to\2.38\eac3to.exe" "H:\HD\BOURNEIDENTITY\"
------------------------------------------------------------------------------
1) FEATURE_1.EVO+FEATURE_2.EVO, 1:58:09
"Main Movie"
- VC-1, 1080p (16:9)
- VC-1, 480p (4:3)
- E-AC3, English, 5.1, 48khz
- E-AC3, French, 5.1, 48khz
- E-AC3, English, 2.0, 48khz, "Director's Commentary"


eac3to v2.38
command line: "D:\editing apps\eac3to\2.38\eac3to.exe" "H:\HD\BOURNE_ULTIMATUM_NA_V2\" 1) 4: "N:\test\audio.ac3" -640 2: "N:\test\video.mkv"
------------------------------------------------------------------------------
EVO, 2 video tracks, 5 audio tracks, 3 subtitle tracks, 1:55:18
"Main Movie"
1: Joined EVO file
2: Chapters, 20 chapters with names
3: VC-1, 1080p24 /1.001 (16:9)
4: VC-1, 480p30 /1.001 (3:2)
5: TrueHD, English, 5.1 channels, 48khz, dialnorm: -27dB
6: E-AC3, English, 5.1 channels, 384kbit/s, 48khz, dialnorm: -27dB, -17ms
7: E-AC3, French, 5.1 channels, 384kbit/s, 48khz, dialnorm: -27dB, -17ms
8: E-AC3, English, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB, -17ms
"Commentary"
9: E-AC3, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB, -4ms
10: Subtitle, English, "SDH"
11: Subtitle, French
12: Subtitle, French, "Forced"
This video conversion is not supported.

[madshi]

What files do I need for chapter support?
VPLST000.XPL

BDEdit can extract chapters to a text file from the .mpls playlists.
Ok, then I should be able to do that, too (sooner or later)...

Madshi the link (http://forum.doom9.org/showthread.php?p=1122525#post1122525) said 1.37 but checking the file on a 32bit XP system is says 1.38. Was something done differently on this binary?
v2.38 is the same as v2.37 with just one more bugfix. The link is always the same and always points to the latest release.

Are there any plans to make chapters and subtitles list when you list the contents of a folder? Or just when listing the contents of the .evos? I ask this because the gui doesn't "see" the chapters and subtitles until it is too late.
I could show both chapters and subtitles in the "folder content" listing. I had decided not to because there are often dozens of subtitle tracks available which would make the "folder content" listing extremely long. Because of that I thought it would make more sense to only list the most important tracks in the "folder content" listing. This listing is not 100% reliable, anyway. eac3to is just listing what the XPL and VTI files say without ever actually reading any EVO file in. If the XPL/VTI files are incorrect, also the "folder content" listing will be incorrect. Futhermore (and very important!) sometimes the order of the tracks listed in the XPL/VTI files do not match the order of the files in the EVO files. So it's kind of dangerous to use the order of the tracks from the XPL/VTI file for demuxing/remuxing. Also often the XPL/VTI do not list sub video and audio tracks which are contained in the EVO files. Because of these reasons the "folder content" listing does not even try to display any track numbers. The listing is only meant to give a quick overview of what titles are available for a HD DVD movie.

[nautilus7]

Chapters are only available if the VPLST000.XPL file is there. Subtitle tracks don't need that file, but eac3to can only list the subtitle track languages if the VPLST000.XPL is there. Handling chapters and subtitles works the same way as audio and video files. They're listed with a track number, they're demuxed when using the "-demux" option and you can also manually handle them.As expected, but subtitles don't work with the evo file i tried (The Shining). I did test again... 1st.evo from that movie don't list any subs, while 2nd.evo does, as well as 1st.evo+2nd.evo combination. EVOdemux shows exactly the same behavior with the 2 evo files. I 'll upload a sampple.

Also 2 things i noticed...

To get chapters list you need to load 1st+2nd evo files, and of course the xpl to be present (makes sense)

When running eac3to to a hd dvd folder to list all titles among with each streams, the subs aren't listed there. Only when loading evo files. Is this on purpose? EDIT: already answered!

[Yraen]

VPLST000.XPL
I could show both chapters and subtitles in the "folder content" listing. I had decided not to because there are often dozens of subtitle tracks available which would make the "folder content" listing extremely long. Because of that I thought it would make more sense to only list the most important tracks in the "folder content" listing. This listing is not 100% reliable, anyway. eac3to is just listing what the XPL and VTI files say without ever actually reading any EVO file in. If the XPL/VTI files are incorrect, also the "folder content" listing will be incorrect. Futhermore (and very important!) sometimes the order of the tracks listed in the XPL/VTI files do not match the order of the files in the EVO files. So it's kind of dangerous to use the order of the tracks from the XPL/VTI file for demuxing/remuxing. Also often the XPL/VTI do not list sub video and audio tracks which are contained in the EVO files. Because of these reasons the "folder content" listing does not even try to display any track numbers. The listing is only meant to give a quick overview of what titles are available for a HD DVD movie.

Thanks for the reply. I know how to work around that limitation, I just wanted to be sure that it wasn't going to change right after I changed my code. :)

I wasn't aware that the track listing in the XPL/VTi might be incorrect. It worked on all of my hd-dvds, I set up a massive batch conversion in the gui and let it run until it finished, and according to the logs I got the tracks I wanted. I guess I just got lucky. :)

[nautilus7]

Same here. All xpl files listed the right tracks, but the order could be wrong. Madshi has a point.