eac3to - audio conversion tool

[honai]

No post in this thread from madshi for a while. Does this mean he sold the rights for eac3to to SlySoft for their new HD transcoding app? ;)

[rik1138]

@Madshi
I've discovered (well, I've been told about, technically) an issue with eac3to recognizing original, unused DTS-HD files. By unused, I mean the files created by the DTS-HD Master Audio Suite (the software that the studios use to actually create the DTS-HD MA audio). The DTS software creates files with an extra 140 bytes of header information at the beginning of the files. This information is stripped off when the DTS is used on a final disc, so if you rip DTS off a Blu-Ray disc, this data is not there. Thus, eac3to doesn't recognize a DTS file with this information... The header isn't vague, the first 8 characters are DTSHDHDR. It would be great if the software could be updated to recognize DTS files with this header information still attached.

I encoded a 7.1 test stream for someone else on the forum, and a member discovered the problem when trying to use eac3to on the file. The files are here:

http://www.sendspace.com/file/mz71xx

There's a DTS-HD High Resolution stream (I don't see this used very often, basically DTS that can go up to a 3mb bitrate), and a DTS-HD Master Audio stream included.

Thanks, Rik

[3ngel]

Hi,

just one question.

How i tell eac3to to use the maximum compression for a .flac destination encoding?

If it is not possible can a swith be added to do the thing?

Thank you very much

[Snowknight26]

How i tell eac3to to use the maximum compression for a .flac destination encoding?


It already does.

[3ngel]

No, because i've recompressed the flac from eac3to, with SoundOut avisynth plugin and the result is 500meg smaller, so i think the default is to use "medium" compression.

[Inspector.Gadget]

Quick question about eac3to and audio delay in Matroska files: The other day I ripped a DVD and then muxed it directly to Matroska without re-encoding, planning to encode it later. I checked the audio delay with DGIndex before muxing and applied the proper delay at mux. eac3to, when run without any options against the Matroska file, reports the audio delay as -301 ms. This sounds like too much of a delay, but off hand I can't remember what the audio delay reported by DGIndex was. Does anyone know how accurate eac3to's parsing of matroska delay values is?

Eduit: Mkvinfo -v shows that the audio starts 16ms after the video, which jogs my memory and I think was the delay I applied at muxing. What's going on here?

[Thunderbolt8]

No, because i've recompressed the flac from eac3to, with SoundOut avisynth plugin and the result is 500meg smaller, so i think the default is to use "medium" compression.perhaps the source was a 24-bit file and that plugin only compressed it to 16-bit?

[3ngel]

perhaps the source was a 24-bit file and that plugin only compressed it to 16-bit?

No, checked. 24bit in, 24 bit out.
81% from eac3to, 68% from SoundOut.

[StephenB]

How do I disable the 2nd pass? I have a file with discontinuities. Eac3to doesn't care and finishes processing. The result has minor picture and audio dropouts I can easily live with. However, eac3to erroneously sees an audio delay and starts a second pass. Afterwards the audio is no longer in sync from the first discon on. I am pretty sure this would not happen without the second pass. But that is automatically executed and I don't know how to prevent that.
Unfortunately the newest version without automatic 2nd pass (2.80) stops at the discons (even with -ignorediscon switch).

I would also like a switch that suppressed the second pass.

[Snowknight26]

No, because i've recompressed the flac from eac3to, with SoundOut avisynth plugin and the result is 500meg smaller, so i think the default is to use "medium" compression.

http://forum.doom9.org/showpost.php?p=1222771&postcount=7316

[saint-francis]

Got an issue with Terminator, the SaraConnor Chronicles. eac3to failed to detect an entire episode. There are three on there and eac3to only found 2. I managed to demux the m2ts file for the episode, but it lacked a chapter file.

PLAYLIST and CLIPINF folder can be found here. (http://www.mediafire.com/?sharekey=910d230ea6085fa036df4e8dca1419699b7d209600d73463b8eada0a1ae8665a)

Thanks madshi for all your work.

Disk 2 is even worse. It doesn't list any independent episodes.

CLIPINF and PLAYLIST files for disk 2 here (http://www.mediafire.com/?sharekey=910d230ea6085fa036df4e8dca14196954ce36e38a4b5ed3b8eada0a1ae8665a)

[3ngel]

http://forum.doom9.org/showpost.php?p=1222771&postcount=7316

The sense of that post is the opposite. Madshi doesn't want to add "less" compression, 'cause "speed is good enough".

In this sense i think Madshi didn't use a "Ultra" level to do a speed/compression balancing.

But in my case (and i think in other cases too at this point), 500 megs smaller are more than a weight difference.

[dev84]

Hi, simple question:


DTS Master Audio, 5.1 channels, 16 bits, 48khz
(core: DTS, 5.1 channels, 16 bits, 1509kbps, 48khz)
Decoding with ArcSoft DTS Decoder...


does it mean that source file is decoded as lossless or only as lossy core?
thx

[tebasuna51]

...does it mean that source file is decoded as lossless or only as lossy core?

ArcSoft decode DTS-MA as lossless.

With -libav only the core.

[TinTime]

The sense of that post is the opposite. Madshi doesn't want to add "less" compression, 'cause "speed is good enough".

In this sense i think Madshi didn't use a "Ultra" level to do a speed/compression balancing.

But in my case (and i think in other cases too at this point), 500 megs smaller are more than a weight difference.

Madshi's post suggests he's using flac's --best compression level.

If you de-compress the flac files produced by eac3to and SoundOut are the resulting wav files identical?

[Thunderbolt8]

The sense of that post is the opposite. Madshi doesn't want to add "less" compression, 'cause "speed is good enough".but isn't "max compression" max compression already?

[dev84]

ArcSoft decode DTS-MA as lossless.

With -libav only the core.

Thank You

[DoomBot]

Any update on being able to extract forced subtitles only?

I would love to see this feature added it would make things so much easier.

[TinTime]

No, because i've recompressed the flac from eac3to, with SoundOut avisynth plugin and the result is 500meg smaller, so i think the default is to use "medium" compression.

I just ran two test encodes of a small 6 channel wav file. The first encoded to flac using eac3to and the second using the flac CLI encoder with the --best switch.

The resulting flac files were (very nearly) the same size. Perhaps there's a problem with SoundOut? It's not truncating the output file is it? I know that the flac CLI encoder can't cope with large files.

[rik1138]

@Madshi
I've discovered (well, I've been told about, technically) an issue with eac3to recognizing original, unused DTS-HD files. By unused, I mean the files created by the DTS-HD Master Audio Suite (the software that the studios use to actually create the DTS-HD MA audio). The DTS software creates files with an extra 140 bytes of header information at the beginning of the files. This information is stripped off when the DTS is used on a final disc, so if you rip DTS off a Blu-Ray disc, this data is not there. Thus, eac3to doesn't recognize a DTS file with this information... The header isn't vague, the first 8 characters are DTSHDHDR. It would be great if the software could be updated to recognize DTS files with this header information still attached.

I encoded a 7.1 test stream for someone else on the forum, and a member discovered the problem when trying to use eac3to on the file. The files are here:

http://www.sendspace.com/file/mz71xx

There's a DTS-HD High Resolution stream (I don't see this used very often, basically DTS that can go up to a 3mb bitrate), and a DTS-HD Master Audio stream included.

Follow-up to this:

There's more to this than I thought. Apparently the header info is different sized depending on the type of DTS file you've encoded... The header includes the 8 bytes following the string STRMDATA, that seems to be consistent with all the files I've encoded so far.

Rik

[Inspector.Gadget]

Anybody know whether the new Arcsoft 3 will continue to provide the necessary decoders for eac3to DTS decoding or whether the architecture has changed?

[laserfan]

Anybody know whether the new Arcsoft 3 will continue to provide the necessary decoders for eac3to DTS decoding or whether the architecture has changed?Yeah maybe have they made it impossible to use their filters outside of TMT...?

[kypec]

I just ran two test encodes of a small 6 channel wav file. The first encoded to flac using eac3to and the second using the flac CLI encoder with the --best switch.

The resulting flac files were (very nearly) the same size. Perhaps there's a problem with SoundOut? It's not truncating the output file is it? I know that the flac CLI encoder can't cope with large files.
I tried to encode via FLAC CLI (revision 1.2.1b alias the latest official release) with --no-padding together with the --best switch. Despite the fact that madshi has included only version 1.2.0 of libFLAC.dll with his eac3to package the only differences occur in the metadata tags at the beginning of files. Raw encoded audio data are bit-to-bit identical.

How to make FLAC CLI encodes bit-to-bit identical with eac3to
First download the latest FLAC library package (http://downloads.sourceforge.net/flac/flac-1.2.1-devel-win.zip) and replace libFLAC.dll in the eac3to folder with the downloaded one.
Run your CLI encoder with the following options:flac.exe --best --no-seektable --no-padding --tag=WAVEFORMATEXTENSIBLE_CHANNEL_MASK=0X0 --tag=VALID_BITS=16 --tag=HDCD=0 inputfile.wav
You will have to change the tag values accordingly for multichannel and/or different bitdepths, my example demonstrates ordinary 16-bit stereo WAV.

[StephenB]

When extracting from a VOB, EAC3TO 3.14 is frequently outputing duplicate subtitle tracks - for instance

Movie - 6 - Subtitle (DVD), 1857 captions.sup and
Movie - 6 - Subtitle (DVD).sup

The two files have identical contents.

The command line is
EAC3TO vts_01_1.vob+vts_01_2.vob+vts_01_3.vob+vts_01_4.vob+vts_01_5.vob+vts_01_6.vob Movie.mkv -seekToIFrames -lowPriority

[3ngel]

My test was done on a 24bit 5.1 2 hours audio. Pheraps the difference takes place on the long run and high def files?

[Audionut]

Hi madshi.

Could you please take a look at this file.

http://d01.megashares.com/?d01=72e4388

It's recorded from DTV. It has 5.1 channels during main broadcast and switches to 2.0 channel during the ad breaks.
Converting to flac results in the audio during the ad breaks being high pitched.

[TinTime]

My test was done on a 24bit 5.1 2 hours audio. Pheraps the difference takes place on the long run and high def files?

I've just tried it using a 2hr+ 5.1 24 bit source and don't find any significant difference in size between flacs produced by eac3to and soundout.

Are you sure the audio is not being corrupted somewhere in your avisynth / soundout chain? If you convert your target flacs back to wav are they identical?

[tebasuna51]

...
It's recorded from DTV. It has 5.1 channels during main broadcast and switches to 2.0 channel during the ad breaks.
Converting to flac results in the audio during the ad breaks being high pitched.

Use DelayCut to split the ac3 source.

[narshorn]

Sorry to bother, some eac3to -decodeHdcd questions ;

Any reason to why eac3to with -decodeHdcd switch gets 6 dB gain lower than standard decoding ?
If so, is it OK to add +6dB as a switch while doing conversions ?
Madshi, Is there a possibility to add -deEmphasis switch option to eac3to ? :)

Best regards,

narshorn

narshorn

[madshi]

Thanks for adding support for 24.975fps. However, when I test on a blank 100 sec audio clip, EAC3TO seems to use 25fps in place of 24.975fps...
Will be fixed in the next build.

The exact FPS for -24.975 option is 25/1.001 = 24.975025
Argh, why did nobody tell me before!? It makes things much simpler...

Any reason why one of my 1080i60 streams makes eac3to output 'xxxxxx fields' while the other one 'xxxxxx frames'?
Yes, because some 1080i60 streams are encoded using fields while other 1080i60 streams are encoded using frames.

Also, would be nice if when demuxing audio, if eac3to can't correct a, say, -9ms delay, it would add -9ms to the file name.
This might be troublesome. All the GUIs out there don't expect the file names to have such a delay added to them. So if I did that, it would probably break all GUIs. Also I'm not sure if it's really worth it. I mean e.g. 9ms are really small...

I've checked the problematic subtitle stream again with 2.13, but the subtitle doesn't display, as with DTS fields set to zero. Only with the DTS entries from tsMuxeR it is displayed.
Argh, ok, will have to do a full DTS fix then. That might take a while...

Ive just tried to demux Body Of Lies bluray and get these errors.(log attatched)
The bluaray is not a seemless branching movie and is a single m2ts. Also running latest anydvd.

edit I think maybe the disc has mastering errors do to the fact that at the times posted by eac3to as errors, there is glitches in video and audio from the original disk. This is a new movie and is not scratched in any way. I will just return for another
If you run into such a problem, always try using the AnyDVD HD built in ripping tool to copy the disc to harddisk first and then run eac3to on the harddisk folder. If that also fails, the disc might be faulty.

What is this ???? v02 The video framerate is correct, but rather unusual. ???????

C:\BT>eac3to "Transporter 2 2005 1080p BluRay MPEG-4 AVC DTS-HD 5.1-Mike" 1)
M2TS, 1 video track, 2 audio tracks, 1 subtitle track, 15.052p
1: Chapters, 12 chapters
2: h264/AVC, 1080p24 (16:9)
v02 The video framerate is correct, but rather unusual.
Most Blu-Rays are 24/1.001. Yours is 24/1.000. That is unusual. eac3to just posts this warning to alert you that the framerate used by this Blu-Ray is strange. You can ignore the warning. But be aware that you might run into stuttering issues when playing back this Blu-Ray.

madshi, how hard would it be to implement reading from the other formats that libavcodec handles? Would be nice to be able to decode Vorbis, or MP3, AAC, and more with eac3to. Wishful thinking on my part?
libav MP3 decoding is already supported. Vorbis and AAC are not. The AAC decoder is not feature complete yet, so I'm waiting until it is. Vorbis might be added sooner or later...

depending on the sountrack being 1.5mbit DTS core or FLAC(from DTSHD in eac3to), the sound levels are much different ?!
I can hardly hear dialogs in DTSHD FLAC and need to increase the center channel to like 140%

I've got all DRC disabled in AC3Filter and it's the same story w/ ffdshow/sonic 4.2 for DTS core, so what gives ?!

it happens in all the DTSHD>FLAC tracks I've built, and forcing dialog normalization in ea3to doesn't do anything(same CRC whether it's enabled or not)

OTOH transcoding LPCM in FLAC gives proper sound levels, identical to DTS core :o

if anyone's got a clue, I'd love to hear it. thanks!
Maybe your DTS decoder is not working properly? Don't know...

5.0 channel DTS can't seem to be converted to AC3.
AC3 encoding is only supports with 5.1 and 2.0 right now. Other formats will be supported in a future version.

If I feed eac3to a duff audio file it reports "The format of the source file could not be detected." but the return code is 0. Could it be changed to non-zero please?
Will be implemented in the next build.

Having an issue with a specific movie - The Game Plan. The video appears to rip just fine but the audio track is a mess :(
IIRC this movie had several playlists. The main playlist has messed up audio. Try some of the other playlists. At least one of them works fine. This is a mastering problem, not an eac3to bug.

[mrr19121970]

This might be troublesome. All the GUIs out there don't expect the file names to have such a delay added to them. So if I did that, it would probably break all GUIs. Also I'm not sure if it's really worth it. I mean e.g. 9ms are really small...



Clown_BD (http://clownbd.techxt.com/) handles this....

see here (http://forum.slysoft.com/showpost.php?p=181531&postcount=499)

[a06] Creating file "C:\Temp\Batman Begins\Audio_6_English.THD+AC3"...

A_AC3, "C:\Temp\Batman Begins\Audio_6_English DELAY 200ms.THD+AC3", timeshift=200ms, lang=eng

[madshi]

Following on from this I've been trying to work out why eac3to doesn't like some (but not all) of my input files. I used eac3to to create a wav file from a pcm input:
eac3to v3.14
command line: "D:\Vtemp\programs\eac3to\eac3to.exe" "test.pcm" "E:\Vtemp\test1.wav" -6 -24 -little -96000 -override -log=1.txt
------------------------------------------------------------------------------
RAW/PCM, 5.1 channels, 1:38:26, 24 bits, 13824kbps, 96khz
Reading RAW/PCM...
Writing WAV...
Creating file "E:\Vtemp\test1.wav"...
The original audio track has a constant bit depth of 24 bits.
Caution: The WAV file is bigger than 4GB. <WARNING>
Some WAV readers might not be able to handle this file correctly. <WARNING>
eac3to processing took 9 minutes, 41 seconds.
Done.

If I then run eac3to on the output wav file I get:
eac3to v3.14
command line: "D:\Vtemp\programs\eac3to\eac3to.exe" "E:\Vtemp\test1.wav" -log=t1.txt
------------------------------------------------------------------------------
The format of the source file could not be detected. <ERROR>

I've been trying to find out why this file does not work and other apparently similar ones do and the difference seems to be the amount of silence at the beginning of the file. This file contains 11866608 bytes = 6.86725 seconds of silence at the beginning. I tried editing the first six samples to non-zero values and it's then detected by eac3to.
This was caused by the HDCD decoder not being able to lock unto the stream due to the many zeroes. Will be fixed in the next build.

Hey, I got a 20gig m2ts file containing several episodes in it. I want to demux the video and audio separately to mkv and flac.

Can I split m2ts files by chapters? If so, how?
Currently not in any automatic way. You could try splitting the m2ts file manually by using a hexeditor or by using some other external tools.

Just wanted to post the following issue that I found using eac3to (v3.14). The sample file that I have used can be found in the link here below.

http://rapidshare.de/files/46162759/extract.h264.html

- The sample is a part of a h264 file extracted directly from blu-ray using eac3to
- When performing the following set of commands on this sample:

eac3to extract.h264 extract.mkv
eac3to extract.mkv 1: extractfrommkv.h264


then you will notice that the result file "extractfrommkv.h264" is different from the original "extract.h264".

Looking a bit more in detail there is a sequence of 162 bytes added at specific places in the file, starting with the following 34 bytes:

00 01 27 64 00 29 AC 7B 01 E0 08 9F 97 FF 00 01 00 01 10 00 00 3E 90 00 0B B8 08 40 00 00 00 01 28 EA

The sequence is the same throughout the file and removing it will give back the original "extract.h264".

The issue has been introduced as of v3.07 (v3.06 works fine), and I think it happens only with AVC/H.264 encoded files with 2 Reframes. Did the test on 4 different films (directly from blu-ray), all with 2 reframes, and they all had the issue, i.e. a lot of artifacts during very fast-moving scenes.
Did some other tests with VC1, MPEG2 and higher-frame H.264 encoded files and did not find a difference between the files before and after the 2 eac3to steps provided here above.
Thanks for the report. Will be fixed in the next build. FWIW, the MKV file is just fine. It's just the MKV h264 extraction which produced incorrect results.

I have problem when i try to demux TsMuxer created Blu-Ray.

M2TS, 1 video track, 1 audio track, 1 subtitle track, 1:24:34, 24p /1.001
1: Chapters, 16 chapters
2: h264/AVC, 1080p23.975 (16:9)
3: AC3, English, 5.1 channels, 640kbps, 48khz
4: Subtitle (PGS), English
v02 The video bitstream is encoded in a non-standard framerate.
v02 The video bitstream framerate field doesn't seem to match the timestamps.

but that is not true, i encode to H264 with x264 and input avs have parametar fps=23.976, all programs and players reconize as 23.976, aslo i tryed with checking changefps in TsMuxer and selecting 24000/1001, but still eac3to report message.
Using "fps=23.976" is wrong. You should encode with 24/1.001 instead. tsMuxeR's conversion to 24/1.001 was buggy, that's why eac3to didn't like that, either. This problem is fixed in the latest tsMuxeR version, though.

Second question is: why eac3to when extracting chapters not include final time anymore?
I've done that by user request! What would you need the final time for?

Any reason to why eac3to with -decodeHdcd switch gets 6 dB gain lower than standard decoding ? If so, is it OK to add +6dB as a switch while doing conversions ?
eac3to is just writing to disk what the HDCD decoder outputs. Of course you can change volume, but please be aware that changing volume involves floating point math, so basically the result will not be 20bit integer, anymore. Instead it will be 64bit floating point. You can then dither that down to whatever bitdepth you want. eac3to by default dithers down to 24bit.

Here's an easy request:

When I convert a 5.1 stream to WAV files, it names the Surround channels .SL.wav and .SR.wav. The encoding software I use for Dolby and DTS expect .LS.wav and .RS.wav (the software will auto-find all the files in their correct channel locations if they are named like this, and it seems to be the industry standard...) Any chance of swapping those letters in a future update? :)
And next week someone else uses a different encoding software which expects "SL" and "SR" or something completely different and I have to change names again? I don't think there's a fixed standard on how to name these files. If there was such a standard, I'd be happy to make use of it. But as long as different encoders expect different names, I'll not adjust to one specific encoder.

I'm trying to use EAC3TO to process an 8 bit PCM file. The goal is to resample it to 48 kHz and transcode it.
8bit handling is buggy right now. Will have to work on that. But it's low on my priority list.

arcsoft totalmedia theatre 3 is out, does anyone know if there could be benefits for eac3to from it?
You could try:

(1) Does the DTS decoder still work for eac3to?
(2) Does it handle those 7.1 DTS-MA tracks with funny speaker placement correctly? That was the only thing that could be improved. Apart from this the old ArcSoft DTS decoder was already pretty perfect.

How do I disable the 2nd pass? I have a file with discontinuities. Eac3to doesn't care and finishes processing. The result has minor picture and audio dropouts I can easily live with. However, eac3to erroneously sees an audio delay and starts a second pass.
Erroneously? How do you know that?

Afterwards the audio is no longer in sync from the first discon on. I am pretty sure this would not happen without the second pass.
How do you know that? You could try pressing "Ctrl+C" when the 2nd pass starts. Depending on how the 2nd pass is executed, this might get you what you want. Eventually eac3to has already deleted first pass results, though. Depends on the circumstances...

What is changed in V3.06 that can explain this behavior? So in short, eac3to can perfectly read the contents of a BD, but it can not read the m2ts files (using CrossOver on Mac OS-X).
Do you have this problem when reading directly from the physical BD? Or when reading a folder structure? Newer eac3to versions use memory mapped file read for physical drive access because that was *a lot* faster. For folder reading the same logic should be used as before, though.

Is there any way to demux a PIP video and its associated soundtrack and keep them in-sync?

Whenever I have attempted to do this, the audio is always wildly out of sync. I'm assuming that either the video or audio is continuous, whilst the corresponding audio or video isn't: often there are long stretches of blank video between the PIP segments.
I've never really tried. But as you say, I think PIP video sometimes has big (very VERY big) areas where there's no video content at all, while secondary audio usually is continuous. That of course makes things very difficult. However, I imagine eac3to should detect the gaps in the video stream and try to repair them when muxing to MKV?

[madshi]

Got an issue with Terminator, the SaraConnor Chronicles. eac3to failed to detect an entire episode. There are three on there and eac3to only found 2. I managed to demux the m2ts file for the episode, but it lacked a chapter file.
As far as I can see, the 3 episodes are 4.m2ts, 12.m2ts and 13.m2ts. Strange enough, there is one playlist for 4.m2ts and one for 13.m2ts, but none for 12.m2ts. So that's the simple reason why eac3to "failed to detect" one episode. eac3to does not detect anything, it just lists the available playlists. If there's no playlist for an episode, naturally eac3to will not list it.

Does this mean he sold the rights for eac3to to SlySoft for their new HD transcoding app? ;)
I don't think they'd like Delphi code... ;)

I've discovered (well, I've been told about, technically) an issue with eac3to recognizing original, unused DTS-HD files. By unused, I mean the files created by the DTS-HD Master Audio Suite (the software that the studios use to actually create the DTS-HD MA audio). The DTS software creates files with an extra 140 bytes of header information at the beginning of the files. This information is stripped off when the DTS is used on a final disc, so if you rip DTS off a Blu-Ray disc, this data is not there. Thus, eac3to doesn't recognize a DTS file with this information... The header isn't vague, the first 8 characters are DTSHDHDR. It would be great if the software could be updated to recognize DTS files with this header information still attached.

I encoded a 7.1 test stream for someone else on the forum, and a member discovered the problem when trying to use eac3to on the file. The files are here:

There's a DTS-HD High Resolution stream (I don't see this used very often, basically DTS that can go up to a 3mb bitrate), and a DTS-HD Master Audio stream included.

[...]

Follow-up to this:

There's more to this than I thought. Apparently the header info is different sized depending on the type of DTS file you've encoded... The header includes the 8 bytes following the string STRMDATA, that seems to be consistent with all the files I've encoded so far.
Interesting. It seems there's a whole file structure. It's very easy to understand, though, so it should be easy adding support for that. Well, "easy" means "not difficult". It does cost some time, though, so it may have to wait a while.

How i tell eac3to to use the maximum compression for a .flac destination encoding?
It already uses max compression by default.

No, because i've recompressed the flac from eac3to, with SoundOut avisynth plugin and the result is 500meg smaller
Then your SoundOut setup is probably broken and does some damage to the audio data. If you want to be sure, you can upload the first 50MB of the eac3to created FLAC file and the first 50MB of the SoundOut created FLAC file. I'll have a look at them then.

Quick question about eac3to and audio delay in Matroska files: The other day I ripped a DVD and then muxed it directly to Matroska without re-encoding, planning to encode it later. I checked the audio delay with DGIndex before muxing and applied the proper delay at mux. eac3to, when run without any options against the Matroska file, reports the audio delay as -301 ms. This sounds like too much of a delay, but off hand I can't remember what the audio delay reported by DGIndex was. Does anyone know how accurate eac3to's parsing of matroska delay values is?

Eduit: Mkvinfo -v shows that the audio starts 16ms after the video, which jogs my memory and I think was the delay I applied at muxing. What's going on here?
Is it possible that the video stream doesn't start with a proper sequence header? eac3to always calculates from the first *displayable* video frame.

I would also like a switch that suppressed the second pass.
If you can find a good reason for that?

Any update on being able to extract forced subtitles only?
It's still on my to do list... :)

When extracting from a VOB, EAC3TO 3.14 is frequently outputing duplicate subtitle tracks - for instance

Movie - 6 - Subtitle (DVD), 1857 captions.sup and
Movie - 6 - Subtitle (DVD).sup

The two files have identical contents.
Can you please post a full eac3to log file (in "[ code ] [ / code ]" brackets, please)?

Is there a possibility to add -deEmphasis switch option to eac3to ? :)
What would that switch do exactly?

Clown_BD handles this....
Nice!

[madshi]

eac3to v3.15 released

http://madshi.net/eac3to.zip

* "24.975" is now interpreted as "25.000/1.001"
* Blu-Ray "sup" are demuxed with DTS set to 0 again, proper fix will come later
* fixed: error code not set for "source file format could not be detected"
* fixed: audio resampling from/to 24.975 didn't work properly
* fixed: WAV files beginning with lots of zeroes were sometimes not accepted

[evdberg]

Do you have this problem when reading directly from the physical BD? Or when reading a folder structure? Newer eac3to versions use memory mapped file read for physical drive access because that was *a lot* faster. For folder reading the same logic should be used as before, though.
I am always reading from a folder on HDD. However, CrossOver maps folders on my Mac HDD to a device in Windows. For example, if I read from '/Volumes/HardDisk1/BD/Movie' it is mapped to 'D:\BD\Movie'. I am pretty sure the file mapping is causing the problem. Maybe you can fallback to normal IO if filemapping fails?

[Thunderbolt8]

thanks for the update :D

[princenigma]

* Blu-Ray "sup" are demuxed with DTS set to 0 again, proper fix will come later

Can someone explain this please ?
Does it means we should wait the proper fix of eac3to to demux sup file ?

[madshi]

I am always reading from a folder on HDD. However, CrossOver maps folders on my Mac HDD to a device in Windows. For example, if I read from '/Volumes/HardDisk1/BD/Movie' it is mapped to 'D:\BD\Movie'. I am pretty sure the file mapping is causing the problem. Maybe you can fallback to normal IO if filemapping fails?
If I knew how to detect that? How does your eac3to log file look like?

* Blu-Ray "sup" are demuxed with DTS set to 0 again, proper fix will come later

Can someone explain this please ?
Does it means we should wait the proper fix of eac3to to demux sup file ?
If you want to convert those sup files to SRT, there's no problem. If you want to remux them into a m2ts structure, then with some movies the current sup files produced by eac3to won't display correctly on some standalone players.

[princenigma]

If you want to remux them into a m2ts structure, then with some movies the current sup files produced by eac3to won't display correctly on some standalone players.

Thanks .

Are PowerDVD and TMT display correctly?

[DoomBot]

So if i have already demuxed the subtitles and when the problem gets fixed can i run the demuxed subs back through eac3to and be fixed then or do i have to go back to the blu-ray it self and do it again?

[laserfan]

If you want to remux them into a m2ts structure, then with some movies the current sup files produced by eac3to won't display correctly on some standalone players.I wondered why a couple of my re-muxed .sups didn't display.

Many thanks madshi for the update--your attention to detail is VERY MUCH APPRECIATED! :)

[evdberg]

If I knew how to detect that? How does your eac3to log file look like?
I just get an error message like this:
Error reading file "D:\BD\Movie\BDMV\STREAM\00001.m2ts".

[narshorn]

[/QUOTE]


What would that switch do exactly?


[/QUOTE]

Some rare CDs have pre-emphasis. Extracting them in secure mode with EAC leaves the file emphasised. eac3to could have a switch to
get proper gain figure at high frequencies when processing the wavs.

WaveEmph110 does the job, but a switch in eac3to could be a nice feature, although rare.

Regards,

narshorn

[napalm-187]

hi

i have buyed arcsoft TMT3 eac3 dosnet work with this decoder
is there any chance to runs this audiodecoder?

zhx

[Snowknight26]

Glad to see you back in action madshi.

Small feature request while you're still around. Would it be possible to add a -down7 switch that makes 6.1 channel output from 8 (7.1) channel soures?

[StephenB]

Originally Posted by StephenB
When extracting from a VOB, EAC3TO 3.14 is frequently outputing duplicate subtitle tracks - for instance

Movie - 6 - Subtitle (DVD), 1857 captions.sup and
Movie - 6 - Subtitle (DVD).sup

The two files have identical contents.
Can you please post a full eac3to log file (in "[ code ] [ / code ]" brackets, please)?
This is from a script that automatically called EAC3TO again if it detected a gaps file

eac3to v3.14
command line: "c:\VID2EVA\tools\eac3to\eac3to.exe" "c:\VID2EVA\Videos\Vid2eva_0.vob"+"c:\VID2EVA\Videos\Vid2eva_1.vob"+"c:\VID2EVA\Videos\Vid2eva_2.vob"+"c:\VID2EVA\Videos\Vid2eva_3.vob"+"c:\VID2EVA\Videos\Vid2eva_4.vob"+"c:\VID2EVA\Videos\Vid2eva_5.vob"+"c:\VID2EVA\Videos\Vid2eva_6.vob"+"c:\VID2EVA\Videos\Vid2eva_7.vob"+"c:\VID2EVA\Videos\Vid2eva_8.vob"+"c:\VID2EVA\Videos\Vid2eva_9.vob"+"c:\VID2EVA\Videos\Vid2eva_10.vob"+"c:\VID2EVA\Videos\Vid2eva_11.vob"+"c:\VID2EVA\Videos\Vid2eva_12.vob"+"c:\VID2EVA\Videos\Vid2eva_13.vob"+"c:\VID2EVA\Videos\Vid2eva_14.vob"+"c:\VID2EVA\Videos\Vid2eva_15.vob"+"c:\VID2EVA\Videos\Vid2eva_16.vob"+"c:\VID2EVA\Videos\Vid2eva_17.vob" "c:\VID2EVA\Videos\Vid2eva.mkv" -seekToIFrames -lowPriority
------------------------------------------------------------------------------
VOB, 1 video track, 3 audio tracks, 3 subtitle tracks, 2:00:55
1: Joined VOB file
2: MPEG2, 480p24 /1.001 (4:3) with pulldown flags
3: AC3 Surround, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB, 24ms
4: AC3, 5.1 channels, 448kbps, 48khz, dialnorm: -27dB, 24ms
5: AC3 Surround, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB, 24ms
6: Subtitle (DVD)
7: Subtitle (SRT)
8: Subtitle (DVD)
[v02] Extracting video track number 2...
[a03] Extracting audio track number 3...
[a04] Extracting audio track number 4...
[a05] Extracting audio track number 5...
[s07] Extracting subtitle track number 7...
[a03] Removing AC3 dialog normalization...
[a04] Removing AC3 dialog normalization...
[a05] Removing AC3 dialog normalization...
[v02] Muxing video to Matroska...
[a03] Applying (E-)AC3 delay...
[a04] Applying (E-)AC3 delay...
[a05] Applying (E-)AC3 delay...
[a03] A remaining delay of -8ms could not be fixed.
[a04] A remaining delay of -8ms could not be fixed.
[a05] A remaining delay of -8ms could not be fixed.
[a04] Creating file "c:\VID2EVA\Videos\Vid2eva - 4 - AC3, 5.1 channels, 448kbps, 48khz.ac3"...
[a03] Creating file "c:\VID2EVA\Videos\Vid2eva - 3 - AC3 Surround, 2.0 channels, 192kbps, 48khz.ac3"...
[a05] Creating file "c:\VID2EVA\Videos\Vid2eva - 5 - AC3 Surround, 2.0 channels, 192kbps, 48khz.ac3"...
[s08] Extracting subtitle track number 8...
[s06] Extracting subtitle track number 6...
[s06] Creating file "c:\VID2EVA\Videos\Vid2eva - 6 - Subtitle (DVD).sup"...
[s08] Creating file "c:\VID2EVA\Videos\Vid2eva - 8 - Subtitle (DVD).sup"...
[v02] Video has a gap of 7 frames at playtime 1:57:07. <WARNING>
[v02] The MKV file was created without making use of the gap/overlap information. <WARNING>
[v02] Please check whether audio is in sync. If it is in sync everything is fine. <WARNING>
[v02] Otherwise ask eac3to to repeat the muxing. It will then automatically make <WARNING>
[v02] use of the detailed gap/overlap information. <WARNING>
[s07] Creating file "c:\VID2EVA\Videos\Vid2eva - 7 - Subtitle (SRT).srt"...
Added fps value to MKV header.
Video track 2 contains 174101 frames.
Subtitle track 6 contains 1857 captions.
Subtitle track 8 contains 1421 captions.
eac3to processing took 10 minutes, 11 seconds.
Done.

eac3to v3.14
command line: "c:\VID2EVA\tools\eac3to\eac3to.exe" "c:\VID2EVA\Videos\Vid2eva_0.vob"+"c:\VID2EVA\Videos\Vid2eva_1.vob"+"c:\VID2EVA\Videos\Vid2eva_2.vob"+"c:\VID2EVA\Videos\Vid2eva_3.vob"+"c:\VID2EVA\Videos\Vid2eva_4.vob"+"c:\VID2EVA\Videos\Vid2eva_5.vob"+"c:\VID2EVA\Videos\Vid2eva_6.vob"+"c:\VID2EVA\Videos\Vid2eva_7.vob"+"c:\VID2EVA\Videos\Vid2eva_8.vob"+"c:\VID2EVA\Videos\Vid2eva_9.vob"+"c:\VID2EVA\Videos\Vid2eva_10.vob"+"c:\VID2EVA\Videos\Vid2eva_11.vob"+"c:\VID2EVA\Videos\Vid2eva_12.vob"+"c:\VID2EVA\Videos\Vid2eva_13.vob"+"c:\VID2EVA\Videos\Vid2eva_14.vob"+"c:\VID2EVA\Videos\Vid2eva_15.vob"+"c:\VID2EVA\Videos\Vid2eva_16.vob"+"c:\VID2EVA\Videos\Vid2eva_17.vob" "c:\VID2EVA\Videos\Vid2eva.mkv" -seekToIFrames -lowPriority
------------------------------------------------------------------------------
VOB, 1 video track, 3 audio tracks, 3 subtitle tracks, 2:00:55
1: Joined VOB file
2: MPEG2, 480p24 /1.001 (4:3) with pulldown flags
3: AC3 Surround, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB, 24ms
4: AC3, 5.1 channels, 448kbps, 48khz, dialnorm: -27dB, 24ms
5: AC3 Surround, 2.0 channels, 192kbps, 48khz, dialnorm: -27dB, 24ms
6: Subtitle (DVD)
7: Subtitle (SRT)
8: Subtitle (DVD)
[v02] Video gap description file detected, will be used for muxing...
[v02] Extracting video track number 2...
[a03] Extracting audio track number 3...
[a04] Extracting audio track number 4...
[a05] Extracting audio track number 5...
[s07] Extracting subtitle track number 7...
[a03] Removing AC3 dialog normalization...
[v02] Muxing video to Matroska...
[a04] Removing AC3 dialog normalization...
[a05] Removing AC3 dialog normalization...
[a03] Applying (E-)AC3 delay...
[a04] Applying (E-)AC3 delay...
[a05] Applying (E-)AC3 delay...
[a03] A remaining delay of -8ms could not be fixed.
[a04] A remaining delay of -8ms could not be fixed.
[a05] A remaining delay of -8ms could not be fixed.
[a04] Creating file "c:\VID2EVA\Videos\Vid2eva - 4 - AC3, 5.1 channels, 448kbps, 48khz.ac3"...
[a03] Creating file "c:\VID2EVA\Videos\Vid2eva - 3 - AC3 Surround, 2.0 channels, 192kbps, 48khz.ac3"...
[a05] Creating file "c:\VID2EVA\Videos\Vid2eva - 5 - AC3 Surround, 2.0 channels, 192kbps, 48khz.ac3"...
[s08] Extracting subtitle track number 8...
[s06] Extracting subtitle track number 6...
[s06] Creating file "c:\VID2EVA\Videos\Vid2eva - 6 - Subtitle (DVD).sup"...
[s08] Creating file "c:\VID2EVA\Videos\Vid2eva - 8 - Subtitle (DVD).sup"...
[s07] Creating file "c:\VID2EVA\Videos\Vid2eva - 7 - Subtitle (SRT).srt"...
Added fps value to MKV header.
Video track 2 contains 174101 frames.
Subtitle track 6 contains 1857 captions.
Subtitle track 8 contains 1421 captions.
eac3to processing took 11 minutes, 24 seconds.
Done.

After this completed, the directory contained:

03/26/2009 04:42 PM 174,273,024 Vid2eva - 3 - AC3 Surround, 2.0 channels, 192kbps, 48khz.ac3
03/26/2009 04:42 PM 406,637,056 Vid2eva - 4 - AC3, 5.1 channels, 448kbps, 48khz.ac3
03/26/2009 04:42 PM 174,273,024 Vid2eva - 5 - AC3 Surround, 2.0 channels, 192kbps, 48khz.ac3
03/26/2009 04:30 PM 3,126,648 Vid2eva - 6 - Subtitle (DVD), 1857 captions.sup
03/26/2009 04:42 PM 3,126,648 Vid2eva - 6 - Subtitle (DVD).sup
03/26/2009 04:42 PM 204,990 Vid2eva - 7 - Subtitle (SRT).srt
03/26/2009 04:30 PM 2,338,892 Vid2eva - 8 - Subtitle (DVD), 1421 captions.sup
03/26/2009 04:42 PM 2,338,892 Vid2eva - 8 - Subtitle (DVD).sup
03/26/2009 04:42 PM 2,768 Vid2eva - Log.txt
03/26/2009 04:42 PM 5,461,622,129 Vid2eva.mkv
03/26/2009 04:30 PM 16 Vid2eva.mkv.gaps

[SamNZDat]

Welcome home madshi!

I'm trying to serve a live 16-bit 44.1 FLAC stream via Tribler, but so far, so stumped. Tribler uses Python to convert live input (FLAC in my case) to .mpegts, but thus far, none of my .mpegts output is readable. There is one thing I should mention about the FLAC I use as input: VLC can play it as a live stream (from another instance of VLC) but foobar2000 can only play it once it has been saved as a finished file.

Any tips on how to proceed? madshi, especially would appreciate your weighing in here.

[deathlord]

Erroneously? How do you know that?


How do you know that? You could try pressing "Ctrl+C" when the 2nd pass starts. Depending on how the 2nd pass is executed, this might get you what you want. Eventually eac3to has already deleted first pass results, though. Depends on the circumstances...

Eac3to says there is an audio delay of 600ms after the discontinuity. After the second pass, eac3to should have corrected that, so audio should be in sync. However, it is not, it is out of sync by something that looks like approx. 600 ms.
So I conlude that if eac3to had not "corrected" the "delay", it would still be in sync.

I was going to try Ctrl+C, but I thought I would ask first because I don't like it. Maybe I'm too late and the files are already changed and might not work any longer.

I assume it would be easy for you to add an undocumented switch "-no2pass" or similar.
If it means a lot of trouble for you, please forget about it.

[laserfan]

I'm confused about the handling of Dolby Digital, Dolby Digital Plus (aka E-AC3), and TrueHD. If I encounter a track like this:

3: TrueHD/AC3, English, 5.1 channels, 48khz
(embedded: AC3, 5.1 channels, 640kbps, 48khz, dialnorm: -30dB)

and I don't want the TrueHD track, do I convert to "ordinary AC3" by using the -core option, or -640 option, or neither e.g. "eac3to 3: audio.ac3" ? :confused:

[TinTime]

Use the last option you mentioned, i.e...

"eac3to 3: audio.ac3"