eac3to - audio conversion tool

[SimonH]

Hi,

I'm running the latest version of eac3to, v2.38 with freshly installed Sonic DTS Decoder 4.3.0.169 and Nero Essentials 7 with the Blu-ray/HD DVD Video Plug-in, and i'm just trying to convert a standard 5.1 channel DTS track to AC3 at 640kbps, but for some reason, eac3to just reports the DTS track details, and does nothing after that:

C:\Downloads\Applications\EAC3to>eac3to -test
Nero Audio Decoder (Nero 7 or older) works fine
Sonic Audio Decoder (4.3.0.169) works fine
Haali Media Splitter (2007-12-29) is installed
Surcode DTS Encoder doesn't seem to be installed
MkvToolnix (v2.2.0) is installed

C:\Downloads\Applications\EAC3to>eac3to c:\encode\output.dts c:\encode\output.ac3 -640 -nero
DTS, 5.1 channels, 2:08:32, 24 bits, 1536kbit/s, 48khz

C:\Downloads\Applications\EAC3to>

It does the exact same thing when using switches -nero -libav -sonic and taking the -640 switch off too. It seems to have stopped working since v2.35, I tried to convert a DTS file that worked on the old version, but it's refusing to do anything now on the current build.

I can't think for the life of me why it's doing this. Any ideas? Could you provide a link to v2.35 and i'll try again on that version?

[edo1080]

I read in the prog description that the -core option is used to extract core from DTS-HD master audio. Can this option be used also on Dolby Digital Plus eac3 files from HD DVD and extract dolby AC3 core of DD+, or the prog simply decodes DD+ and then re-encodes it in AC3 640kb max?

[Jaja1]

... I think the Haali splitter works just fine. The latest version has fixed some important issues. Unfortunately, the VC1 HD DVD's I've tried so far, still have the same problems as before. Short freezes or speed up from time to time, distorted audio. Remuxing to mkv using eac3to solves this, so I guess Haali's splitter hasn't improved on this point.

It makes me once more happy, that you made eac3to :)

[nautilus7]

http://www.sendspace.com/file/rxw6b1

Here is the evo file i can't find the subs.

[bmnot]

I read in the prog description that the -core option is used to extract core from DTS-HD master audio. Can this option be used also on Dolby Digital Plus eac3 files from HD DVD and extract dolby AC3 core of DD+, or the prog simply decodes DD+ and then re-encodes it in AC3 640kb max?

I don't think DD+ have core tracks. HD DVD set top players reeencode DD+ to DD or DTS for spdif, and I assume PowerDVD does the same.

[edo1080]

I think you are right. DD+ on HD DVD has no core because it's not necessary since every player has to decode it. By contrast BD has a core on DD+ (look at Dolby white paper on DD+) to be back compatible with DD only deocders which might be present on BD players. But since I'm interested to HD DVD tracks no core should be present. Anyway eac3to works woth -core option on a eac3 file and produces an AC3 track. I can't unterstand if it simply doesn't care about the -core option and decodes eac3 into ac3 or -core option does anything, it's not clear.

[quantum]

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.
I had a similar issue with a True HD track. To work around it, I added the delay as reported by eac3to in mkvmerge and remuxed. That solved it.

[EDIT]It turns out eac3to correctly handled the audio delay for this track. I didn't use the framecount parameter in my AVS script, so when two EVO's were joined, the frames were off and this caused audio sync issues at the halfway point of the movie.

[Chumbo]

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.
madshi, will you keep creating version release posts like these or will you just update the 1st post from now on? I only ask because there isn't one for 2.38. I find them useful because the thread is so large now, it's a nice way to keep track of issues post and pre certain releases by using your release posts. :)

[Henry The Ripper]

Originally Posted by 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?

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.

I do it because I have to demux all the tracks to get main audio and commentary. Or I can directly demux + transcode DTS and then demux all tracks to get the commentary.

But thanks, just worried because I can't test the channel order with my equipment.


Originally Posted by Henry The Ripper
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?

Well, I think they should tell the plugins etc. to start extracting to wavs and then to encode the file to dts?

[madshi]

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.
The reason for this behaviour is simple: There are simply no subtitles whatsoever in this first EVO file you sent me (which is only 95MB big). The first subtitle content often comes only after 150-200MB. That's why both EvoDemux and eac3to always read the first 300MB of an EVO source.

I'm running the latest version of eac3to, v2.38 with freshly installed Sonic DTS Decoder 4.3.0.169 and Nero Essentials 7 with the Blu-ray/HD DVD Video Plug-in, and i'm just trying to convert a standard 5.1 channel DTS track to AC3 at 640kbps, but for some reason, eac3to just reports the DTS track details, and does nothing after that
Sorry, that's a new bug introduced by the subtitle support. Will be fixed by v2.39.

Could you provide a link to v2.35 and i'll try again on that version?
No. I generally do not repost old versions. I rather fix bugs in a new build.

I read in the prog description that the -core option is used to extract core from DTS-HD master audio. Can this option be used also on Dolby Digital Plus eac3 files from HD DVD and extract dolby AC3 core of DD+, or the prog simply decodes DD+ and then re-encodes it in AC3 640kb max?
The "-core" option extracts the 5.1 E-AC3 core from a 7.1 E-AC3 track. But it doesn't magically give you an AC3 core from an HD DVD E-AC3 track because HD DVD E-AC3 tracks don't have an AC3 core. However, if you use the ".ac3" extension for the filename, eac3to will transcode (= fully decode + reencode) the track to AC3 for you.

Unfortunately, the VC1 HD DVD's I've tried so far, still have the same problems as before. Short freezes or speed up from time to time, distorted audio.
Well, that's rather a problem with the EVO splitter and doesn't have much to do with VC-1 support. EVO files do not contain timestamp information for every video frame. As a result Haali's splitter has to "guess" the timestamps for most frames. If the guessing goes wrong, playback can show the problems you mentioned.

Older Haali versions showed major artifacts with some movies (e.g. POTO USA, Unforgiven, Million Dollar Baby, Silver Surfer Europe, several adult titles, ...). These problems are all fixed in the latest version. Also h264 seeking should work a lot better than before.

Remuxing to mkv using eac3to solves this
eac3to has the luxury to do offline remuxing. So eac3to can first try to use straight 23.976 timestamps. And only if that fails it has to try a different solution. Haali's filter as a realtime splitter might get into trouble if he used simple 23.976 timestamps without caring about the timestamps in the EVO file. That's where the difference is coming from, I guess...

madshi, will you keep creating version release posts like these or will you just update the 1st post from now on? I only ask because there isn't one for 2.38. I find them useful because the thread is so large now, it's a nice way to keep track of issues post and pre certain releases by using your release posts. :)
Check out the "changelog.txt" which is contained in every eac3to download.

2.37 and 2.38 don't start the encoding procedure with normal command line "eac3to source.pcm destination.dts"
That's a bug which will be fixed in v2.39.

[madshi]

eac3to v2.39 released

http://madshi.net/eac3to.zip

* simple audio transcoding was broken

[the_provider]

Greetings.

Today I encountered the first(...) DVD with dts audio file. I tried to convert it to ac3(5.1 ch.) but it didn't worked. When using EAC3to Conversion Utility I select some options that can be seen beneath:
eac3to.exe "C:\HEART\HEART T01 DELAY 0ms.dts" "C:\HEART\Z_HEART T01 DELAY 0ms.ac3" -448 -sonic -keepDialnorm

and the result was:
DTS-ES, 5.1 channels, 1:38:39, 16 bits, 1536kbit/s, 48khz
This audio conversion is not supported.
I replaced sonic with libav with no positive result.

I must mention that:
eac3to.exe "C:\HEART\HEART T01 DELAY 0ms.dts" "C:\HEART\Z_HEART T01 DELAY 0ms.ac3"
worked fine but due to the size I felt that I need to decrease the bitrate.

As I am new to using this tool my guess is that I missed something. I am willing to learn to use it properly that is the reason why I am kindly requesting a link to a User Manual or some hints to configure the conversion correctly.

Question: 378 and 320 kbit are not in the drop-down box, is it correct to assume that the conversion to this rates is not supported?

Best regards.

[Beastie Boy]

If
eac3to.exe "C:\HEART\HEART T01 DELAY 0ms.dts" "C:\HEART\Z_HEART T01 DELAY 0ms.ac3"
works OK, then use the command

eac3to.exe "C:\HEART\HEART T01 DELAY 0ms.dts" "C:\HEART\Z_HEART T01 DELAY 0ms.ac3" -448

Replace 448 with whatever bitrate you wish to use.

Cheers, Beastie.

[Jaja1]

Well, that's rather a problem with the EVO splitter and doesn't have much to do with VC-1 support. I never noticed these specific problems with AVC, so I always understood it to be a VC1 problem

EVO files do not contain timestamp information for every video frame. As a result Haali's splitter has to "guess" the timestamps for most frames. If the guessing goes wrong, playback can show the problems you mentioned.
Okay, thanks for the explanation. I wonder then how standalone players work with EVO's. But that is another discussion.

eac3to has the luxury to do offline remuxing. So eac3to can first try to use straight 23.976 timestamps. And only if that fails it has to try a different solution. Haali's filter as a realtime splitter might get into trouble if he used simple 23.976 timestamps without caring about the timestamps in the EVO file. That's where the difference is coming from, I guess...Are you therefore saying that even in the unlikely event, that you make your splitter available as DS filter, your splitter will have the same problems? However, PDVD is able to split the stream without these problems. So it is possible.

[Chumbo]

...
Check out the "changelog.txt" which is contained in every eac3to download.
...
Oh I do, I just meant here on the site, the posts between the release posts are more easily followed. :)

[Rectal Prolapse]

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.

Any way I can contact him without subscribing to the matroska mailing list? I PM'd him details here already, but I don't know if he frequents doom9 anymore.

Thanks again!

[Chumbo]

Any way I can contact him without subscribing to the matroska mailing list? I PM'd him details here already, but I don't know if he frequents doom9 anymore.

Thanks again!
So you're aware, someone on the mailing list just brought this up, so it's in-the-know now. ;)

[Chumbo]

......* log file is copied to destination path (of first destination file)...
I, for one, really appreciate this. Thanks so much.

[Thunderbolt8]

thanks for the new version again!

I got a request, could you make eac3to handle at least corrupt TV sources (.ts files)? I understand that it makes sense to improve clean support for disc sources, as a disc can be re-ripped if there should be an error. but in most TV sources theres always at least 1 video or audio error and because of that then eac3to stops the whole process. unfortunately those files cant be recaptured, unless they repeat that movie sometime later on again.
the thing is when I encouter an error then i demux video and audio and try to correct the error (with delaycut for example), but afterwards ill end up with using eac3to again anyway trying to mux audio and video together, since I feel its most reliable. this process can be very time consuming though and it would be nice if eac3to can just handle smaller errors or problem like that progressive/interlace frame change at the beginning of a movie, so im not forced to get out of the remux loop. otherwise I unfortunately wouldnt be able to mux more than half tv sources I encouter with eac3to. thanks!

[Rectal Prolapse]

chumbo, thanks for the news! Um, but I really hope someone could duplicate the problem - I don't want to be the only one. :)

thunderbolt - I dunno, I think it would be difficult to handle partially-corrupted ts fiiles.

[laserfan]

I think it would be difficult to handle partially-corrupted ts fiiles.I agree. Thunderbolt8, try VideoReDo+ or TVSuite on your corrupt .ts file (it's free for 2 weeks full-function if you register your email address) i.e. run Quick Stream Fix on it. Other tools to try might be mpeg2repair and possibly HDTV2MPEG.

[Thunderbolt8]

thunderbolt - I dunno, I think it would be difficult to handle partially-corrupted ts fiiles.
well I cant really judge that, but I only know when you remux those files with haali for example it just accepts like any kind of input and mostly, if the problems are not more than usual corrupted frames, also stays in sync afterwards.

[BLKMGK]

eac3to v2.39 released

http://madshi.net/eac3to.zip

* simple audio transcoding was broken

Looks to be working again on 64bit Vista, thank you! I'll be trying my BD test again:cool:

[Greif]

madshi, can you update your first post to provide examples of how to demux chapters and subtitles. It will be easier for people to find.

:thanks:

[Greif]

I'm getting the following message:

eac3to.exe FEATURE_1.EVO+FEATURE_2.EVO 2: chapters.txt 3: video.mkv 4: audio.ac3 9: subtitles.sup -448 -libav
EVO, 1 video track, 5 audio tracks, 3 subtitle tracks, 1:47:43
"Main Movie"
1: Joined EVO file
2: Chapters, 20 chapters with names
3: VC-1, 1080p24 /1.001 (16:9)
4: E-AC3, English, 5.1 channels, 1536kbit/s, 48khz, dialnorm: -27dB
5: E-AC3, Spanish, 5.1 channels, 768kbit/s, 48khz, dialnorm: -27dB
6: E-AC3, French, 5.1 channels, 768kbit/s, 48khz, dialnorm: -27dB
7: E-AC3, English, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB
8: E-AC3, English, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB
9: Subtitle, English
10: Subtitle, Spanish
11: Subtitle, French
Creating file "chapters.txt"...
Extracting primary video track...
Muxing video to Matroska...
Extracting audio track number 4...
Removing dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
Getting "Nero Audio Decoder 2" instance failed.

I am specifying the libav parameter so why would it want to try to decode with Nero? Possible bug?

Update: It works if I do the audio on its own.

[BLKMGK]

Madshi, doing the BD now - so far the video looks good! One thing I noticed that is different with M2TS files vs EVO is that when done eac3to isn't reporting a framecount. With EVO files I have used this in an AVIsynth file with the impression it was necessary to maintain audio synch. Am I mistaken? At this point not yet sure what I will put in my AVIsynth file to feed meGUI but I'm working on it :)

[Chumbo]

I'm getting the following message:

eac3to.exe FEATURE_1.EVO+FEATURE_2.EVO 2: chapters.txt 3: video.mkv 4: audio.ac3 9: subtitles.sup -448 -libav
EVO, 1 video track, 5 audio tracks, 3 subtitle tracks, 1:47:43
"Main Movie"
1: Joined EVO file
2: Chapters, 20 chapters with names
3: VC-1, 1080p24 /1.001 (16:9)
4: E-AC3, English, 5.1 channels, 1536kbit/s, 48khz, dialnorm: -27dB
5: E-AC3, Spanish, 5.1 channels, 768kbit/s, 48khz, dialnorm: -27dB
6: E-AC3, French, 5.1 channels, 768kbit/s, 48khz, dialnorm: -27dB
7: E-AC3, English, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB
8: E-AC3, English, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB
9: Subtitle, English
10: Subtitle, Spanish
11: Subtitle, French
Creating file "chapters.txt"...
Extracting primary video track...
Muxing video to Matroska...
Extracting audio track number 4...
Removing dialog normalization...
Decoding with DirectShow (Nero Audio Decoder 2)...
Getting "Nero Audio Decoder 2" instance failed.

I am specifying the libav parameter so why would it want to try to decode with Nero? Possible bug?

Update: It works if I do the audio on its own.
I believe you need to add the parameters in the right place. Your -448 and -libav are for the audio transcoding which is #4. So you need to put those for #4, e.g.,
eac3to.exe FEATURE_1.EVO+FEATURE_2.EVO 2: chapters.txt 3: video.mkv 4: audio.ac3 -448 -libav 9: subtitles.sup

[Chumbo]

Madshi, doing the BD now - so far the video looks good! One thing I noticed that is different with M2TS files vs EVO is that when done eac3to isn't reporting a framecount. With EVO files I have used this in an AVIsynth file with the impression it was necessary to maintain audio synch. Am I mistaken? At this point not yet sure what I will put in my AVIsynth file to feed meGUI but I'm working on it :)
I believe the frame count is only returned if your source is VC-1. You shouldn't have any sync issues with other sources like mpeg2 and AVC/h264.

[BLKMGK]

Hrm, I'm pretty sure I've seen it in some H.264 HD-DVD. I'll double check later with Transformers maybe, right now I'm proceeding without it and am not using a Graph file either. Thanks!

[pbjr]

Madshi,

I just want to start off by saying thank you so much for eac3to!! Great work! Just like everyone else, I really appreciate all your work.

Now that I got that out of the way... :)

I converted Constantine's TrueHD to Flac, but the channel mapping is wrong when I playback through mplayer.

What is the default channel order for TrueHD? I tried searching the web but didn't find much. Is it L,R,C,LFE,Ls,Rs?

As a goof I tried converting to WAVs, just to see this output.

Creating file "audio.24bit.L.wav"...
Creating file "audio.24bit.C.wav"...
Creating file "audio.24bit.LFE.wav"...
Creating file "audio.24bit.R.wav"...
Creating file "audio.24bit.SR.wav"...
Creating file "audio.24bit.SL.wav"...

Is that the sequence order eac3to is seeing it? I'm going to try another TrueHD (300) in the mean time.

Thanks,
- PBjr

[Chumbo]

Hrm, I'm pretty sure I've seen it in some H.264 HD-DVD. I'll double check later with Transformers maybe, right now I'm proceeding without it and am not using a Graph file either. Thanks!
Maybe you saw the frame count in EVOdemux since it's an HD DVD? ;) Yeah, you can use DGAVCDec for h264 instead of a graph.

[woah!]

Maybe you saw the frame count in EVOdemux since it's an HD DVD? ;) Yeah, you can use DGAVCDec for h264 instead of a graph.

EVO, 2 video tracks, 4 audio tracks, 2 subtitle tracks, 0:50:34
"Where No Man Has Gone Before"
1: Chapters, 6 chapters without names
2: h264/AVC, 1080p24 /1.001 (16:9)
3: h264/AVC, 480p24 /1.001 (15:11), -17ms
4: TrueHD, English, 5.1 channels, 48khz, dialnorm: -27dB, -84ms
5: AC3, English, 2.0 channels, 192kbit/s, 48khz, -84ms
6: AC3, Spanish, 2.0 channels, 192kbit/s, 48khz, dialnorm: -30dB, -84ms
7: E-AC3, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB, -84ms
8: Subtitle, English
9: Subtitle, Spanish
Creating file "Y:\movie.txt"...
Extracting primary video track...
Extracting audio track number 4...
Removing dialog normalization...
Muxing video to Matroska...
Applying RAW/PCM delay...
Encoding AC3...
Extracting subtitle track number 8...
Creating file "Y:\movie.ac3"...
Creating file "Y:\movie.sup"...
Added fps value to MKV header.
Video track 2 contains 72733 frames.
Video track 3 contains 90871 frames.
eac3to processing took 7 minutes, 58 seconds.
Done.

came from the log so it does seem to do h264 no???

[umaximus]

@madshi

Is it safe to join seamless branching titles by 'copy b method.. Will the future version of eac3to be able to mux the audio without gaps/overlaps on already joined files? Or should we wait with any kind of joining. How come there is no sync issues with ''copy /b'' & tsRemux on seamless branching discs (i have made several titles like ratt, deja vu, cars, etc, with no playback problems)?

[edo1080]

I tested to extract the AC3 core from a .eac3 DD+ stream demuxed from an HD DVD (Blades of Glory) unsing the -core switch and it works. Could this mean the DD+ has a DD AC3 core on HD DVD too?

E-AC3, 5.1 channels, 1:33:09, 768kbit/s, 48khz, dialnorm: -27dB
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
Encoding AC3...
Creating file "C:\The Bourne Blu Ray\Audio.ac3"...
eac3to processing took 11 minutes, 12 seconds.
Done.


This is what the log reports, anyway it seems decoding first with nreo decoder and then encoding; but how can eac3to encode if it has no built-in encoder?

[nautilus7]

@madshi

Is it safe to join seamless branching titles by 'copy b method.. Will the future version of eac3to be able to mux the audio without gaps/overlaps on already joined files? Or should we wait with any kind of joining. How come there is no sync issues with ''copy /b'' & tsRemux on seamless branching discs (i have made several titles like ratt, deja vu, cars, etc, with no playback problems)?When joining multiple .m2ts files there's a chance a small audio piece (usually a frame) to exist in both previous and next .m2ts files. Suppose we have AC3 audio (which has the longest frame of all audio types - 32 ms), this can result in max 16ms (half frame) de-sync in each join point. Multiply by the .m2ts files and here it is the problem. For example if a movie consists of 10 files (9 joints) and AC3 audio , if all parts overlap by a frame, you can have a max of 9x16ms= 144ms de-sync. In other audio types frame duration is a lot sorter (~5ms) which means the problem is tiny and can't be noticed. Of course there's a chance that audio frames don't overlap, so no problem at all.

Hope i made it clear enough to understand it. :)

[nautilus7]

I tested to extract the AC3 core from a .eac3 DD+ stream demuxed from an HD DVD (Blades of Glory) unsing the -core switch and it works. Could this mean the DD+ has a DD AC3 core on HD DVD too?



This is what the log reports, anyway it seems decoding first with nreo decoder and then encoding; but how can eac3to encode if it has no built-in encoder?
No, no core inside. eac3to uses aften ac3 encoder that comes with the package. The -core switch you type is just ignored.

[edo1080]

Thanks nautilus for this explaination. It's what I needed.

[the_provider]

If ...
works OK, then use the command

eac3to.exe "C:\HEART\HEART T01 DELAY 0ms.dts" "C:\HEART\Z_HEART T01 DELAY 0ms.ac3" -448

Replace 448 with whatever bitrate you wish to use.


Thank you for your swift reply but when desired bitrate is mentioned it doesn't work.
I also tried to mention explicitly the filter and I got the same result in all 3 attempts (default, sonic, libav).

In fact the ONLY combination that works is:

C:\HEART>"C:\Program Files\eac3to\eac3to.exe" "C:\HEART\HEART T01 DELAY 0ms.dts" "C
:\HEART\HEART T01 DELAY 0ms.ac3" -libav

Any attempt to set bitrate or other filter returns a
"This audio conversion is not supported." message.
I have installed ffdshow filter package.

Any other ideas?

Thank you in advance.
Best regards.

[nautilus7]

Do you use the keymakers gui? It has a bug and it doesn't put a space between the output filename and the -448. Switch to the alternate gui (eac3to and more gui - see separate thread in the audio section) or use the cmd version.

Best decoder for dts is sonic. If you have it let it decode instead of selecting any other. Don't use the -keepdialnorm option. There's no point.

[crazydane]

So I ripped this title using AnyDVD and kept the following 2 files:

00000.m2ts 15,964,734 KB
00006.m2ts 17,032,122 KB

I then joined them with the copy /b command, and proceeded to demux the video & audio with eac3to (soundtrack converted to flac).

All went well and I then used mkvmerge as usual to create my final .mkv container.

The movie plays back flawlessly in mpc and wmp11 and there are no sync issues. BUT, I ended up with the alternate ending version as opposed to the theatrical one. I have watched both versions via my stand-alone BD player and I don't really mind ending up with the alternate ending version for my backup.

My question is, for titles with alternate endings (I think it is only the last couple of chapters on this title), how the heck can one pick one over the other?

As I recall, I think there was a 00003.m2ts file at around 2GB or so, which probably contained the theatrical ending and somehow when playing this version on a standalone player, the 00003.m2ts file gets swapped in near the end?

Sorry for the slightly off topic question, but I'd love for eac3to be able to deal with alternate ending movies like this someday!

And Madshi, that you for this wonderfull tool! I have been ripping my BD titles using only AnyDVD, eac3to and mkvmerge with no issues with any of the 30 or so titles I have done so far!

[SoToS]

please also GUI renew!

[nautilus7]

So I ripped this title using AnyDVD and kept the following 2 files:

00000.m2ts 15,964,734 KB
00006.m2ts 17,032,122 KB

I then joined them with the copy /b command, and proceeded to demux the video & audio with eac3to (soundtrack converted to flac).

All went well and I then used mkvmerge as usual to create my final .mkv container.

The movie plays back flawlessly in mpc and wmp11 and there are no sync issues. BUT, I ended up with the alternate ending version as opposed to the theatrical one. I have watched both versions via my stand-alone BD player and I don't really mind ending up with the alternate ending version for my backup.

My question is, for titles with alternate endings (I think it is only the last couple of chapters on this title), how the heck can one pick one over the other?

As I recall, I think there was a 00003.m2ts file at around 2GB or so, which probably contained the theatrical ending and somehow when playing this version on a standalone player, the 00003.m2ts file gets swapped in near the end?

Sorry for the slightly off topic question, but I'd love for eac3to be able to deal with alternate ending movies like this someday!

And Madshi, that you for this wonderfull tool! I have been ripping my BD titles using only AnyDVD, eac3to and mkvmerge with no issues with any of the 30 or so titles I have done so far!No no no no...

I am Legend Blu-ray is authored by retards. EACH one of the 2 files you mention (00000.m2ts & 00006.m2ts) contain the WHOLE movie, but with different endings (2 scenes are also different btw). The disc doesn't make any use of seamless branching as it should. It's ridiculously authored. I don't know exactly how you can make one file with the 2 versions of the movie (that doesn't contain common parts twice, obviously).

[crazydane]

You're right nautilus. My "I Am Legend.mkv" is 26,154,776 KB and when I checked it out before I just watched part of the beginning and then dragged the slider towards the end and watched it. I bet if I move the slider to the middle, I'll see the ending credits of the theatrical version, followed by the beginning of the alternate version.

You're right though, I can't believe they put 2 full version of the movie on there. What a waste!

[Chumbo]

EVO, 2 video tracks, 4 audio tracks, 2 subtitle tracks, 0:50:34
"Where No Man Has Gone Before"
1: Chapters, 6 chapters without names
2: h264/AVC, 1080p24 /1.001 (16:9)
3: h264/AVC, 480p24 /1.001 (15:11), -17ms
4: TrueHD, English, 5.1 channels, 48khz, dialnorm: -27dB, -84ms
5: AC3, English, 2.0 channels, 192kbit/s, 48khz, -84ms
6: AC3, Spanish, 2.0 channels, 192kbit/s, 48khz, dialnorm: -30dB, -84ms
7: E-AC3, 2.0 channels, 192kbit/s, 48khz, dialnorm: -27dB, -84ms
8: Subtitle, English
9: Subtitle, Spanish
Creating file "Y:\movie.txt"...
Extracting primary video track...
Extracting audio track number 4...
Removing dialog normalization...
Muxing video to Matroska...
Applying RAW/PCM delay...
Encoding AC3...
Extracting subtitle track number 8...
Creating file "Y:\movie.ac3"...
Creating file "Y:\movie.sup"...
Added fps value to MKV header.
Video track 2 contains 72733 frames.
Video track 3 contains 90871 frames.
eac3to processing took 7 minutes, 58 seconds.
Done.

came from the log so it does seem to do h264 no???
That's great. The last couple avc titles I put into mkv containers did not show the frames in the log. I think I was using versions 2.36 and older of eac3to at the time. I wish I had the log, but I no longer have it. I remember because I specifically looked for the frame count. The last one was Hitchhiker's Guide.

[umaximus]

When joining multiple .m2ts files there's a chance a small audio piece (usually a frame) to exist in both previous and next .m2ts files. Suppose we have AC3 audio (which has the longest frame of all audio types - 32 ms), this can result in max 16ms (half frame) de-sync in each join point. Multiply by the .m2ts files and here it is the problem. For example if a movie consists of 10 files (9 joints) and AC3 audio , if all parts overlap by a frame, you can have a max of 9x16ms= 144ms de-sync. In other audio types frame duration is a lot sorter (~5ms) which means the problem is tiny and can't be noticed. Of course there's a chance that audio frames don't overlap, so no problem at all.

Hope i made it clear enough to understand it. :)

ay yes, i kinda know that but thanks for detailed info.. my concern is if the eac3to could detect this gaps/overlaps, made by joining via copy/b, and fix them.. also, why TSremux doesnt have this problems, or better, how TSremux deals with this de-sync joint points?

[nautilus7]

I don't really know if eac3to can correct them. I don't think it can. I don't think also that tsremux detects those problems. It's just that the de-sync (if any - remember, it's possible that no de-sync at all to exist) is very small and can't be noticed.

[wildchild22]

Not sure if this is the right place to post this. I am thinking of buying a popcorn hour and as far as I can tell it supports lpcm but not any new hd-dvd or blu-ray hd audio formats. Is it possible to convert dolby digital + to lpcm ?

I am thinking about ripping my hd-dvds and turning all the audio into lpcm so I dont loose quality.

( I have used eac3to to convert dolby digital + to ac3-640 before and its great just not sure if it can do all the conversions to lpcm)

[nautilus7]

Of course that's possible, but space consuming. Does it support AC3 and DTS? The quality loss isn't that significant. It's your choice.

[wildchild22]

it does support ac3 and dts. Is this command line correct for lpcm

eac3to whateverfile.dd+ whateverfile.pcm

[nautilus7]

Yes it is.