eac3to - audio conversion tool

[yfed]

Well, I've tried a different program to convert DTS-MA to DTS (TsMuxer). That worked fine. Then I opened *.DTS in Adobe Audition and saved it as *.WAV.
Then did "eac3to input.wav output.ac3 -override -1 -16 -192 -48000" (since I don't really need DTS/high bitrate AC3 for 1965 mono sound)

I think there's something fishy about 1ch(mono) dts-ma. Maybe Arcsoft/Sonic couldn't handle it, although I have no idea what TsMuxer uses

[madshi]

Originally Posted by Henry The Ripper

The TrueHD to DTS isn't working in the 2.64 version. 2.60 that I have works fine.

The new version names the surround channels as BR/BL and Surcode gets confused cause it doesn't find SR/SL named channels and closes.

Will be fixed in the next build.

Originally Posted by himan2001

Can you implement a missing Option? :

-slowdown24 and -speedup24

for 24.000 -> 23.976 and 23.976 -> 24.000 conversions ;-) ?

It's already on my to do list.

Originally Posted by Atak_Snajpera

eac3to reports insane audio delay. DGIndex demuxes audio without problems.

Thanks for the sample, I'll have a look at this.

Originally Posted by Atak_Snajpera

Madshi could you add switch something like "-Automatically repeat audio demuxing if gaps are detected"?

Originally Posted by fib0by

I wouldn't mind if that was the default behavior, unless there's something non-obvious which would make that not the best choice.

Originally Posted by killa_kid

When eac3to finds gaps while doing a DTS (and i believe AAC, FLAC) it outputs as PCM and puts up the message about gaps being found before running the external program. I was wondering if you could add an option (maybe something like "-killOnGaps") that will not start that second part.

I'm planning to automate the gap removal for audio in a future build.

Originally Posted by va1aY

can this program create a FLAC file from six WAV files?

Multiple wav files are not supported as input yet. You need to use WaveWizard to combine those separate wav files into one.

Originally Posted by deathlord

I have an lpcm track (Crank blu-ray) where eac3to "sees" 7.1 channels, but only 5.1 of them seem to contain audio data.

A sample would be nice. Are you sure that the back channels are silent throughout the whole movie? I mean they may be silent for most of the movie, but maybe they're used only once in a while?

Originally Posted by yfed

Hi! I wanna report an error decoding DTS-HD MA:

Here are track specs: DTS Master Audio, 1.0 channels, 16 bits, 48khz (From "For A Few Dollars More")

Can I have a sample of the m2ts file, please? The first 50MB would be great. Maybe 20MB would also do...

Originally Posted by yfed

Well, I've tried a different program to convert DTS-MA to DTS (TsMuxer). That worked fine.

eac3to can do the same thing. Just use the "-core" switch. That will get rid of the DTS-HD data blocks.

Originally Posted by rica

Second pass:

[a02] The audio gaps/overlaps technically can't be removed from the TrueHD bitstream.
[a02] In order to remove them you'll have to transcode the audio to another format.

The question here is:
Shall i re-encode to another format after or before merging?

Ask eac3to to output *.wav or *.flac or *.pcm from the get go.

Originally Posted by xkodi

i assume that "white paper [1]" is the SpkrConfig5.doc and if so this isn't true at all, because KSAUDIO_SPEAKER_5POINT1_SURROUND is equal to 0x60F and that number is all over the document.

Yes, but the constant named "KSAUDIO_SPEAKER_5POINT1_SURROUND" is not mentioned at all, while the other 3 "KSAUDIO_SPEAKER" constants are mentioned. The doc reads as if there was no constant with the name of "KSAUDIO_SPEAKER_5POINT1_SURROUND".

Originally Posted by xkodi

in the "Windows Driver Development Kit" (DDK) there is a sample of simple WDM audio driver and that sample contains "KSAUDIO_SPEAKER_5POINT1_SURROUND", but according to the copyright notice it's from year 2000. the SpkrConfig5.doc is from year 2006.

Originally Posted by xkodi

what will happen if the WDM audio driver knows about 0x60F, but doesn't know about 0x3f, which is the case with the MS sample driver from 2000, then maybe 0x3f won't work correctly and if the real WDM audio drivers are based on the old sample then that will explain why 0x60f is more better than 0x3f, but it doesn't make it more right.

It is very clear that all DTS, Dolby and Microsoft want 5.1 to be played through the surround speakers and not through the back speakers. That makes 0x60f clearly more right compared to 0x3f. The only thing speaking for 0x3f is that doc you posted and the fact that most software out there uses 0x3f. It's really a mess, just as you say. I think I will keep using 0x60f, unless we get convincing evidence that it makes problems. From my point of view 0x60f is correct. And MS confirmed that to me via email. Maybe 0x3f is correct, too (but obsolete/outdated), maybe not. Even MS seems to be confused about that...

Originally Posted by K1ngp1ng

there is only one valid multichannel setup which is described in ITU-R BS.775-1.
so 2/2/1 should be treated like 3/2/1 without center.

page 12: http://www.tonmeister.de/foren/surro...E_2002_v2a.PDF

Thanks! That's a good PDF. To be fair, this PDF doesn't seem to know the default speaker setup for 7.1. It doesn't even know that 6.1 exists at all. Probably the PDF was written before 6.1 and 7.1 were introduced for home cinema. But still the PDF clearly shows that 3/1, 2/2, 2/1 etc are all interpreted as 3/2 with some channels missing. That IMHO supports the new channel masks I'm using in the current eac3to version. Also the PDF clearly states that when duplicating the mono surround channel, volume should be lowered by 3db (or 50%), which is what the eac3to "-double7" option correctly does. So all is fine. The only thing currently wrong is that I still need to remap the channels when using 6.1 output without doubling the channels (due to the new channel mask I'm using).

[tebasuna51]

Originally Posted by madshi

It is very clear that all DTS, Dolby and Microsoft want 5.1 to be played through the surround speakers and not through the back speakers.

Of course, for Surround (110º) speakers but don't say Side (90º) speakers, is near than Back (150º) but isn't the same.

That makes 0x60f clearly more right compared to 0x3f. ... From my point of view 0x60f is correct. And MS confirmed that to me via email. Maybe 0x3f is correct, too (but obsolete/outdated), maybe not. Even MS seems to be confused about that...

Not problem for me if you accept 0x3f like valid 5.1 input in your soft. I will accept also 0x60f like valid 5.1 input in my soft
The order channel is the same then no problem.

- A 5.1 system must send the last channels to surround speakers without modify.

- A 6.1 system must extract the common part from the last channels to BackCenter speaker. The rest for the Left and Right Side/Surround/Back (you can use any name) speakers.

- A 7.1 system with Side (90º) and Back (150º) must send 0.9 from last channel to SideRight speaker and 0.4 from last channel to BackRight speaker (equivalent for Lefts)

The only thing currently wrong is that I still need to remap the channels when using 6.1 output without doubling the channels (due to the new channel mask I'm using).

Then, when play a 6.1 directly with ArcSoft or Sonic players we have a wrong mapping channels?
If you need remap a supported format maybe you aren't at right way.

I think this can be a big problem, here isn't a name problem, here the channel order is different.

And the flac behaviour? I have a flac Feature Request from 2008-04-22 without answer. Do you now any 'official' answer from flac developers?

[Atak_Snajpera]

I found another problem with .vob as a source. If I let eac3to to use gaps I loose audio synchronization. Audio demuxed without using gaps has perfect synchronization.
Source: http://www.mediafire.com/file/nnnfmnx3hnm/VTS_01_1.VOB

[madshi]

Originally Posted by tebasuna51

Of course, for Surround (110º) speakers but don't say Side (90º) speakers, is near than Back (150º) but isn't the same.

Side is much nearer to Surround than Back is...

Originally Posted by tebasuna51

Not problem for me if you accept 0x3f like valid 5.1 input in your soft.

Of course I do!

Originally Posted by tebasuna51

Then, when play a 6.1 directly with ArcSoft or Sonic players we have a wrong mapping channels?

ArcSoft/Sonic output a different channel order, but the channel mask they're using (0x13f) fits to that channel order.

Originally Posted by tebasuna51

If you need remap a supported format maybe you aren't at right way.

So ArcSoft and Sonic must always be right? You don't know how many hacks I had to add to eac3to to make the decoders work correctly. Both ArcSoft and Sonic decoders make a lot of (different) mistakes. Especially with samples with funny channel configurations.

Originally Posted by tebasuna51

I think this can be a big problem, here isn't a name problem, here the channel order is different.

Yes, I'm aware of that. That's the main reason why eac3to used to double the 7th channel by default! Because with 7.1 there's no confusion. With 6.1 there's room for confusion.

Originally Posted by tebasuna51

And the flac behaviour? I have a flac Feature Request from 2008-04-22 without answer. Do you now any 'official' answer from flac developers?

Check out this thread:

http://www.hydrogenaudio.org/forums/...howtopic=49182

Search for "WAVEFORMATEXTENSIBLE_CHANNEL_MASK". This new tag is already implemented in the current eac3to build! But not yet in madFlac...

[madshi]

Originally Posted by Atak_Snajpera

I found another problem with .vob as a source. If I let eac3to to use gaps I loose audio synchronization. Audio demuxed without using gaps has perfect synchronization.

Thanks, will check this out, too.

[madshi]

Originally Posted by tebasuna51

Of course, for Surround (110º) speakers but don't say Side (90º) speakers, is near than Back (150º) but isn't the same.

Just double checked this. You are wrong!!

See here:

http://www.audioholics.com/tweaks/sp...ssential-guide
http://www.smartbuyspeakers.com/speakersetup
http://forum.ecoustics.com/bbs/messa...79/129023.html

There's no difference between "side" and "surround" speakers. And there's no difference of the placement of side/surround speakers between 5.1, 6.1 and 7.1. They are always at 90°-110°.

[tebasuna51]

Originally Posted by madshi

Just double checked this. You are wrong!!

See here:

http://www.audioholics.com/tweaks/sp...ssential-guide
http://www.smartbuyspeakers.com/speakersetup
http://forum.ecoustics.com/bbs/messa...79/129023.html

There's no difference between "side" and "surround" speakers. And there's no difference of the placement of side/surround speakers between 5.1, 6.1 and 7.1. They are always at 90°-110°.

Why I wrong? Using your samples:
My surround speakers in room with 5.1 are at 110º, correct?
My side speakers in room with 7.1 are at 90º and the back at 135º, correct?
Are typical locations. With less speakers we need cover more space. And now the difference between surround-side is 20º and surround-back is 25º.

We need be a little more flexible, I say you surround in near to side than back, but you never can convince me that is the same.

[madshi]

Originally Posted by tebasuna51

My surround speakers in room with 5.1 are at 110º, correct?

Anywhere between 90° and 110° is correct for the surround=side speakers in a 5.1, 6.1 or 7.1 setup.

Originally Posted by tebasuna51

My side speakers in room with 7.1 are at 90º and the back at 135º, correct?

Anywhere between 90° and 110° is correct for the surround=side speakers in a 7.1, 6.1 or 7.1 setup.

--------------------------------

The point is: There's no difference in recommended surround/side speaker placement between 5.1, 6.1 and 7.1. That is what Dolby says. That is what the first 10 hits about speaker placement at google say. So for all intends and purposes we can say "surround = side". If you don't agree with that then you seem to be alone with your opinion, at least according to what a google search told me...

Do you have a link to an article (from a well respected source) which claims that surround speakers should be placed differently compared to side speakers?

[Jeff Flowerday]

Just got this message converting "There Will Be Blood" TrueHD -> pcm

[libav] End of stream indicated
[libav] Lossless check failed - expected 0, calculated 21
[a02] This audio track has a constant bit depth of 24 bits.
Video track 1 contains 227909 frames.
eac3to processing took 18 minutes, 19 seconds.
Done.

Just wondering what it means and what effect does it have on the end result?

[killa_kid]

Originally Posted by Jeff Flowerday

Just got this message converting "There Will Be Blood" TrueHD -> pcm

[libav] End of stream indicated
[libav] Lossless check failed - expected 0, calculated 21
[a02] This audio track has a constant bit depth of 24 bits.
Video track 1 contains 227909 frames.
eac3to processing took 18 minutes, 19 seconds.
Done.

Just wondering what it means and what effect does it have on the end result?

having ripped the same movie and gotten the same (or similar) error, it meant nothing to me The movie came out gorgeous and the audio was spot on

[fib0by]

It looks like chapter track detection may not work properly in some cases.

There are 4 titles on this disk. Titles 1 and 2 are the movie. Titles 3 and 4 are the movie with "video commentary" (images from production rendered on top of the movie, pretty silly). The length of the video is the same for all titles - after all, it's the same movie.

eac3to seems to detect the chapters list just fine for titles 3 and 4, but not for 1 and 2. I can't quite believe that there are no chapters for 1 and 2 since that is the actual movie that most people would watch.

Even if I run eac3to directly on the .m2ts files, it's the same - it can't find any chapters for 00008.m2ts but it does find them for 00011.m2ts

In this particular case, I can probably extract the chapters from title 3 and use them with the video from title 1, but if 1 actually has chapters but they are not detected, it's probably a bug that needs to be squished.

E:\work>eac3to E:\hd\blah
1) 00010.mpls, 00008.m2ts, 1:53:18
   - VC-1, 1080p24 /1.001 (16:9)
   - DTS Master Audio, English, multi-channel, 48khz
   - DTS, English, stereo, 48khz

2) 00011.mpls, 00008.m2ts, 1:53:18
   - VC-1, 1080p24 /1.001 (16:9)
   - DTS Master Audio, English, multi-channel, 48khz
   - DTS, English, stereo, 48khz

3) 00013.mpls, 00011.m2ts, 1:53:18
   - VC-1, 1080p24 /1.001 (16:9)
   - DTS, English, multi-channel, 48khz

4) 00012.mpls, 00011.m2ts, 1:53:18
   - VC-1, 1080p24 /1.001 (16:9)
   - DTS, English, multi-channel, 48khz

E:\work>eac3to E:\hd\blah 1)
M2TS, 1 video track, 2 audio tracks, 2 subtitle tracks, 1:53:18
1: VC-1, 1080p24 /1.001 (16:9)
2: DTS Master Audio, English, 7.1 channels, 24 bits, 48khz
3: DTS, English, 2t channels, 24 bits, 256kbps, 48khz
4: Subtitle (PGS), English
5: Subtitle (PGS), Spanish

E:\work>eac3to E:\hd\blah 2)
M2TS, 1 video track, 2 audio tracks, 2 subtitle tracks, 1:53:18
1: VC-1, 1080p24 /1.001 (16:9)
2: DTS Master Audio, English, 7.1 channels, 24 bits, 48khz
3: DTS, English, 2t channels, 24 bits, 256kbps, 48khz
4: Subtitle (PGS), English
5: Subtitle (PGS), Spanish

E:\work>eac3to E:\hd\blah 3)
M2TS, 1 video track, 1 audio track, 2 subtitle tracks, 1:53:18
1: Chapters, 21 chapters
2: VC-1, 1080p24 /1.001 (16:9)
3: DTS, English, 5.1 channels, 24 bits, 1536kbps, 48khz
4: Subtitle (PGS), English
5: Subtitle (PGS), Spanish

E:\work>eac3to E:\hd\blah 4)
M2TS, 1 video track, 1 audio track, 2 subtitle tracks, 1:53:18
1: Chapters, 21 chapters
2: VC-1, 1080p24 /1.001 (16:9)
3: DTS, English, 5.1 channels, 24 bits, 1536kbps, 48khz
4: Subtitle (PGS), English
5: Subtitle (PGS), Spanish

E:\work>eac3to E:\hd\blah\BDMV\STREAM\00008.m2ts
M2TS, 1 video track, 2 audio tracks, 2 subtitle tracks, 1:53:18
1: VC-1, 1080p24 /1.001 (16:9)
2: DTS Master Audio, English, 7.1 channels, 24 bits, 48khz
3: DTS, English, 2t channels, 24 bits, 256kbps, 48khz
4: Subtitle (PGS), English
5: Subtitle (PGS), Spanish

E:\work>eac3to E:\hd\blah\BDMV\STREAM\00011.m2ts
M2TS, 1 video track, 1 audio track, 2 subtitle tracks, 1:53:18
1: Chapters, 21 chapters
2: VC-1, 1080p24 /1.001 (16:9)
3: DTS, English, 5.1 channels, 24 bits, 1536kbps, 48khz
4: Subtitle (PGS), English
5: Subtitle (PGS), Spanish

madshi, what do you want me to do to help you troubleshoot it? Upload a 50MB sample of 00008.m2ts on a website somewhere? Let me know.

[nautilus7]

The chapter list is stored in the .mpls files (these are the playlist files). In your case 00010.mpls and 00011.mpls. You can check whether there are chapters when playing the disc with powerdvd or in your standalone player. You can also check and extract chapters with BDedit.

[rack04]

Maybe someone can help me out with the Blu-ray version of "Terminator The Sarah Connor Chronicles Season 1 Disc 1". When eac3to analyzes the disc structure it comes up with the following format:



I'm not sure which 3 titles to select to get the 3 individual episodes.

[bigdog660]

Looks like 2) 3) and 4) judging by the playing time.

[rebkell]

Originally Posted by rack04

Maybe someone can help me out with the Blu-ray version of "Terminator The Sarah Connor Chronicles Season 1 Disc 1". When eac3to analyzes the disc structure it comes up with the following format:



I'm not sure which 3 titles to select to get the 3 individual episodes.

Actually, I think you are going to have to break them out, there is no episode 2 on the disk menu. I think the first line 00004.m2ts is episode 1, 00012.m2ts is episode 2 and 00013.m2ts is episode 3.

Playlist 2) is an extended version of episode 1.
You can't actually play episode 2 from the menu, it's not even an option. It's a strange menu, you pop in the disk and it starts playing the 2hr+ selection which I assume is all three episodes and is probably 1) ..

Ok, 4 is the first episode which is 46:40
12 is episode 2 which is 44:05
13 is episode 3 which is 43:47
15,16 and 17 are a special feature which combined equals 39:08

It's a really strange disk.

[nurbs]

rebkell is right. 00012.m2ts is the second(?) episode. I was worried about that myself when I encoded the disk, but there is nothing missing.

[madshi]

Originally Posted by fib0by

It looks like chapter track detection may not work properly in some cases.

There are 4 titles on this disk. Titles 1 and 2 are the movie. Titles 3 and 4 are the movie with "video commentary" (images from production rendered on top of the movie, pretty silly). The length of the video is the same for all titles - after all, it's the same movie.

eac3to seems to detect the chapters list just fine for titles 3 and 4, but not for 1 and 2. I can't quite believe that there are no chapters for 1 and 2 since that is the actual movie that most people would watch.

Even if I run eac3to directly on the .m2ts files, it's the same - it can't find any chapters for 00008.m2ts but it does find them for 00011.m2ts

In this particular case, I can probably extract the chapters from title 3 and use them with the video from title 1, but if 1 actually has chapters but they are not detected, it's probably a bug that needs to be squished.

E:\work>eac3to E:\hd\blah
1) 00010.mpls, 00008.m2ts, 1:53:18
   - VC-1, 1080p24 /1.001 (16:9)
   - DTS Master Audio, English, multi-channel, 48khz
   - DTS, English, stereo, 48khz

3) 00013.mpls, 00011.m2ts, 1:53:18
   - VC-1, 1080p24 /1.001 (16:9)
   - DTS, English, multi-channel, 48khz

E:\work>eac3to E:\hd\blah 1)
M2TS, 1 video track, 2 audio tracks, 2 subtitle tracks, 1:53:18
1: VC-1, 1080p24 /1.001 (16:9)

E:\work>eac3to E:\hd\blah 3)
M2TS, 1 video track, 1 audio track, 2 subtitle tracks, 1:53:18
1: Chapters, 21 chapters
2: VC-1, 1080p24 /1.001 (16:9)

madshi, what do you want me to do to help you troubleshoot it? Upload a 50MB sample of 00008.m2ts on a website somewhere? Let me know.

Don't need any m2ts files in this case. You can upload the playlists listed in the eac3to log (*.mpls) instead.

Originally Posted by Atak_Snajpera

eac3to reports insane audio delay. DGIndex demuxes audio without problems.

Checked the VOB. Actually eac3to is absolutely right. The audio begins to play only after about 1 minute. Before that there's no audio in the VOB. The reason why things didn't work as expected when using eac3to is that there's a video gap of 1 minute in the stream, too. And MPEG2 video gap/overlap fixing in eac3to v2.64 was broken. Basically this VOB looks like this:

[a few stupid video frames] [...no data for 1 minute...] [here comes the real data]

Since both audio and video have about 1 minute missing data, straightly demuxing things without correcting audio/video sync kind of works (which is what DGIndex probably does), although strictly spoken it's not correct. If you play the original VOB file you'll notice that it's playing 1 minute of black screen without audio. If you play the video/audio streams demuxed by DGIndex you'll notice that this 1 minute of black silence is gone. Now if you use eac3to v2.65 and let eac3to fix the video and audio gaps/delays, the final MKV+AC3 will play just like the VOB with correct audio sync and 1 black silence minute in the beginning.

Originally Posted by Atak_Snajpera

I found another problem with .vob as a source. If I let eac3to to use gaps I loose audio synchronization. Audio demuxed without using gaps has perfect synchronization.

Again eac3to detects things correctly (there's an overlap of about 1 second in both video and audio), but the fixing of the video overlap was broken in v2.64. With v2.65 letting eac3to fix audio+video overlap will result in correct audio sync. The resulting MKV+AC3 will be about 1 second shorter if you let eac3to fix the overlaps.

[madshi]

eac3to v2.65 released

http://madshi.net/eac3to.zip

* automatic channel remapping for 6.1 tracks with wrong channel mask
* automatic channel remapping for ArcSoft DTS decoder 6.1 tracks
* fixed: TrueHD -> Surcode encoding didn't work, anymore
* fixed: MPEG2 + h264 video gap/overlap removal didn't work properly

[nautilus7]

Originally Posted by madshi

Checked the VOB. Actually eac3to is absolutely right. The audio begins to play only after about 1 minute. Before that there's no audio in the VOB. The reason why things didn't work as expected when using eac3to is that there's a video gap of 1 minute in the stream, too. And MPEG2 video gap/overlap fixing in eac3to v2.64 was broken. Basically this VOB looks like this:

[a few stupid video frames] [...no data for 1 minute...] [here comes the real data]

Since both audio and video have about 1 minute missing data, straightly demuxing things without correcting audio/video sync kind of works (which is what DGIndex probably does), although strictly spoken it's not correct. If you play the original VOB file you'll notice that it's playing 1 minute of black screen without audio. If you play the video/audio streams demuxed by DGIndex you'll notice that this 1 minute of black silence is gone. Now if you use eac3to v2.65 and let eac3to fix the video and audio gaps/delays, the final MKV+AC3 will play just like the VOB with correct audio sync and 1 black silence minute in the beginning

I 've came across a few dvds that DGIndex reports a huge delay with. neuron2 suggests skipping a few GOPs at the beginning because they don't contain valid timecodes.