eac3to - audio conversion tool

[ramicio]

No, Arcsoft is the only way to decode DTS-HD Master besides the DTS suite, and the rare and impossible to find Sonic decoder. I want eac3to to mix my stuff because it detects clipping, and it also uses the best possible processing. I also don't care for foobar2000 for anything but playback.

[hello_hello]

No, Arcsoft is the only way to decode DTS-HD Master besides the DTS suite, and the rare and impossible to find Sonic decoder.

Fair enough. I was just looking at ffdshow's settings and noticed the Pass-Through option for DTS HD, so wondered if that meant it could also decode it.

I want eac3to to mix my stuff because it detects clipping, and it also uses the best possible processing. I also don't care for foobar2000 for anything but playback.

I'm sure there's a rational logic behind all that.

[tebasuna51]

... It's these pesky 6.1 I come across now and then. I like eac3to because it detects clipping. Would it be possible for avisynth to turn out 6.1?
For 6.1 in monowavs you can use:
sl=WavSource("D:\tmp\SL.wav").ConvertAudioToFloat() # Load channels to mix
sr=WavSource("D:\tmp\SR.wav").ConvertAudioToFloat()
bc=WavSource("D:\tmp\BC.wav").ConvertAudioToFloat()
bl=MixAudio(sl, bc, 1.0, 0.7071) # Mix half power of BC into SL
br=MixAudio(sr, bc, 1.0, 0.7071) # Mix half power of BC into SR

fl=WavSource("D:\tmp\FL.wav").ConvertAudioToFloat() # Load the rest of channels
fr=WavSource("D:\tmp\FR.wav").ConvertAudioToFloat()
fc=WavSource("D:\tmp\FC.wav").ConvertAudioToFloat()
lf=WavSource("D:\tmp\lf.wav").ConvertAudioToFloat()

MergeChannels(fl, fr, fc, lf, bl, br) # Merge the 6 channels
Normalize() # Normalize to avoid clip
ConvertAudioTo24bit() # Change to desired precision

[wonkey_monkey]

From what I've read, eac3to should be able to extract a core AC3 track from a TrueHD file, but I get this:

M:\TMP\bd>eac3to bd.thd bd.ac3
TrueHD, 7.1 channels, 48kHz, dialnorm: -27dB
AC3 encoding doesn't support back channels. Will mix them into the surround.
Removing TrueHD dialog normalization...
Decoding with libav/ffmpeg...
Remapping channels...
Mixing surround channels...
Encoding AC3 <640kbps> with libAften...

which looks like it's completely re-encoding. Am I missing an option?

David

[ramicio]

That file doesn't have an AC3 core, it would be listed.

It would look like this:

d:\>\apps\eac3to\eac3to.exe file.thd
TrueHD/AC3, 7.1 channels, 48kHz
(embedded: AC3, 5.1 channels, 640kbps, 48kHz)

Along with my previous dilemma, I decided to just have 1.1.0.0 and 1.1.0.7 of the Arcsoft filters and flip-flop them depending on what I need.

And why does a line get added to the bottom of the code boxes?

[Midzuki]

Well, maybe Lord Mulder :devil: could want to write a "dtshdma2wavs.exe" :D , which would always pick the "best" Arcsoft DLL for each case :)

[Anakunda]

Hello any chance to add custom Vorbis encoder (oggenc2) support to eac3to ?

[Chumbo]

@madshi,
I was doing some batch extracts of audio from my media files when I noticed the following that I had not seen before.eac3to v3.24
command line: eac3to "file1.mkv" 2: "file1.ac3"
------------------------------------------------------------------------------
MKV, 1 video track, 1 audio track, 0:42:01, 24p /1.001
1: h264/AVC, 720p24 /1.001
2: AC3, ???, 5.1 channels, 448kbps, 48kHz, dialnorm: -25dB
I tried searching this thread but could not really find anything on the ???.

[tebasuna51]

...
2: AC3, ???, 5.1 channels, 448kbps, 48kHz, dialnorm: -25dB
Seems a country code not supported by eac3to.
Try with MediaInfo.

[tebasuna51]

Hello any chance to add custom Vorbis encoder (oggenc2) support to eac3to ?
You can use the external encoder with:
eac3to input stdout.wav | OggEnc2 -q 3 --ignorelength -o output.ogg -

[Chumbo]

Seems a country code not supported by eac3to.
Try with MediaInfo.
I had done that already and didn't notice any difference. I only ask because normally I don't see anything in the ??? position, i.e.,
2: AC3, 5.1 channels, 448kbps, 48kHz, dialnorm: -25dB, 17ms
Aside from items like duration, file size, etc. MediInfo reports identical results for files with the ??? and without.
Audio
ID : 2
Format : AC-3
Format/Info : Audio Coding 3
Mode extension : CM (complete main)
Codec ID : A_AC3
Duration : 42mn 10s
Bit rate mode : Constant
Bit rate : 448 Kbps
Channel(s) : 6 channels
Channel positions : Front: L C R, Side: L R, LFE
Sampling rate : 48.0 KHz
Bit depth : 16 bits
Compression mode : Lossy
Delay relative to video : 17ms
Stream size : 135 MiB (12%)
Language : English
I don't see any problems caused by this but I was mostly curious as to what the ??? position is supposed to be reporting.

[tebasuna51]

...I only ask because normally I don't see anything in the ??? position,...
No?
Maybe your mkv's don't have the language field filled:
eac3to v3.24
command line: "D:\Programa\eac3to\eac3to.exe" "D:\Test\cono3.mkv"
------------------------------------------------------------------------------
MKV, 1 video track, 2 audio tracks, 1 subtitle track, 0:00:22, 25p
1: h264/AVC, English, 720p25
2: DTS, English, 5.1 channels, 24 bits, 755kbps, 48kHz
3: AC3, German, 5.1 channels, 448kbps, 48kHz
4: Subtitle (SRT), Spanish

[Chumbo]

No?
Maybe your mkv's don't have the language field filled:
eac3to v3.24
command line: "D:\Programa\eac3to\eac3to.exe" "D:\Test\cono3.mkv"
------------------------------------------------------------------------------
MKV, 1 video track, 2 audio tracks, 1 subtitle track, 0:00:22, 25p
1: h264/AVC, English, 720p25
2: DTS, English, 5.1 channels, 24 bits, 755kbps, 48kHz
3: AC3, German, 5.1 channels, 448kbps, 48kHz
4: Subtitle (SRT), Spanish
MediaInfo seems to have no issues picking up the language as English and I always set the language on the audio stream specifically to eng using mkvtoolnix when creating my MKV files. Like I said, it's not a big deal, I was just curious about the ???.

[sneaker_ger]

Mkvmerge (in recent versions) doesn't write the language field if set to "English" and eac3to doesn't recognize that. That's why eac3to doesn't display any language for English tracks if muxed with such a version. I don't know where the "???" comes from. Look at the value of the field "Language" in mkvinfo - maybe you can spot a difference between files with "" (empty) and "???".

[elmarikon]

Hi guys!

I don't know if this has been asked before, but how can I prevent eac3to from downsampling wavs when I come from a 24bit, 48KHz, 6ch pcm file, trying to extract mono wav files?
I don't know why, but for some reason eac3to from time to time thinks it has to reduce the bit depth to 16bit...

I have tried to define all values, like eac3to in-6ch.pcm out\out.wavs -48000 -24 -6 -little -override (even tried -down24)... still receving 16bit wavs, but I do need 24bit.

Wht can I do, and why s it doing that anyway?

l the best and thanx for the apart from that great tool!

all the best

swk

[Snowknight26]

Your input file is 16-bit then.

[ramicio]

Eac3to detects when a 24-bit file has the 8 least significant bits filled with zeroes (which is easily accomplished by taking a 16-bit file and simply saving it as 24-bit) and drops it to 16 bits. Keeping it 24 bits would be pointless and a waste of space. When eac3to detects that the file is between 17 and 23 bits, it will keep it 24.

[Anakunda]

You can use the external encoder with:
eac3to input stdout.wav | OggEnc2 -q 3 --ignorelength -o output.ogg -

this doesnot work for me :(
"D:\media\eac3to\eac3to.exe" "I:\video\title00 ID3 Lossless.dts" -110ms -normalize -down6 -progressnumbers -log="I:\video\title00 ID3 Lossless_Output_eac3to.txt" | D:\media\OGG\aoTuV\oggenc2.exe -q 5.0 --ignorelength -o "I:\video\title00 ID3 Lossless_Output.ogg" -

Any idea?

I get the error: Please specify the source and dest files first and then the options.ERROR: Input file "(stdin)" is not a supported format

[sneaker_ger]

You have to specify "stdout.wav" as a fake output file, like in tebasuna51's example. Try:
"D:\media\eac3to\eac3to.exe" "I:\video\title00 ID3 Lossless.dts" stdout.wav -110ms -normalize -down6 -progressnumbers -log="I:\video\title00 ID3 Lossless_Output_eac3to.txt" | D:\media\OGG\aoTuV\oggenc2.exe -q 5.0 --ignorelength -o "I:\video\title00 ID3 Lossless_Output.ogg" -

[Anakunda]

If I specify fake output file then the output will go to file and not stdout right?

I tried it now:
"D:\media\eac3to\eac3to.exe" "I:\video\title00 ID3 Lossless.dts" $$fake.wav -110ms -normalize -down6 -progressnumbers -log="I:\video\title00 ID3 Lossless_Output_eac3to.txt" | D:\media\OGG\aoTuV\oggenc2.exe -q 5.0 --ignorelength -o "I:\video\title00 ID3 Lossless_Output.ogg" -

and still got error:

ERROR: Input file "(stdin)" is not a supported format


Sorry I missed the name thankyou!!!

[sneaker_ger]

No. "stdout.wav" is a reserved name which will let eac3to pipe to stdout instead of writing to a file.

[laserfan]

@tebasuna51 you wrote (a LONG time ago!)
Use the external encoder Aften with eac3to STDOUT:

eac3to 3.dtsma stdout.wav | Aften -b 320 - 3.ac3

But when I used this with a 3ch DTS that MediaInfo says is "Front: L C R", I got an ac3 file that is "Front: L, R and Side: C" which had the center channel coming-out from the surrounds! But next I looked at aften's help and tried this:

eac3to 3ch.dts stdout.wav | aften -b 576 -chconfig 3/0 - 3ch.ac3

and I think it sounds right, and MediaInfo says this output ac3 file is now "Front: L C R" just as the DTS! :)

I just wanted to thank you for having steered me in the right direction (all the way from 2009)!

:thanks:

[tebasuna51]

You are welcome.

Yes, with 'pipe' sintax you can add all Aften parameters.
To know the all the parameters:

Aften -longhelp

[nibus]

madshi, do you have any plans to incorporate qtaacenc.exe, similar to neroaacenc.exe? Piping works fine but it would be pretty handy to have a built in switch to use qtaacenc.

[Midzuki]

How about looking for mkvmerge, etc., in the PATH, not in the Registry :rolleyes:
[C:\]
=>eac3to -test

eac3to (v3.24) is installed
Nero Audio Decoder (Nero 6 or older) doesn't seem to be installed
http://www.nero.com/eng/store-blu-ray.html
CAUTION: You need Nero 7. Nero 8 won't work with eac3to.
ArcSoft DTS Decoder (1.1.0.1) works fine
Sonic Audio Decoder (3.24.0.0) doesn't seem to be installed
Haali Matroska Muxer doesn't seem to be installed
http://haali.net/mkv
Nero AAC Encoder (1.5.4.0) is installed
Surcode DTS Encoder doesn't seem to be installed
http://www.surcode.com
MkvToolnix doesn't seem to be installed
http://www.bunkus.org/videotools/mkvtoolnix

[C:\]
=>

Also, fancy "background colors" in a CLI window are annoying :( Don't assume

1) everybody has their monitors (and eyes) calibrated in the same way as yours;

2) everybody sets cmd.exe to use white text on a black background;

At least Mplayer gives me an option to turn OFF :cool: the stupid "feature":

## Fancy Color Messages
msgcolor=no

[tebasuna51]

madshi, do you have any plans to incorporate qtaacenc.exe, similar to neroaacenc.exe? Piping works fine but it would be pretty handy to have a built in switch to use qtaacenc.

Is not the same than neroaacenc (standalone encoder), to work with qtaacenc you need install QuickTime.

To incorporate a encoder you need support for all parameters of the encoder.

If 'pipe' work you can use any new/not-common parameters without a new eac3to version.
I use the 'pipe' method also for Aften.

[laserfan]

You are welcome.

Yes, with 'pipe' sintax you can add all Aften parameters.
To know the all the parameters:

Aften -longhelp
I missed that one too--will have to look at longhelp.

Many thanks T51. :)

EDIT: BTW where is the latest binary? I have two copies presently on my C: drive, both of which say "Date modified 1/19/2008 8:51 AM" and 272,896 bytes, and the -version info says only

Aften: A/52 audio encoder
Version SVN
(c) 2006-2007 Justin Ruggles, Prakash Punnoor, et al.

:confused:

[nibus]

Is not the same than neroaacenc (standalone encoder), to work with qtaacenc you need install QuickTime.

To incorporate a encoder you need support for all parameters of the encoder.

If 'pipe' work you can use any new/not-common parameters without a new eac3to version.
I use the 'pipe' method also for Aften.

Right, I do understand the difference between them. Maybe it would be easier to incorporate QuickTime AAC from within eac3to.

Of course I have had issues with QuickTime and large multichannel files. It usually quits partway through them.

[tebasuna51]

EDIT: BTW where is the latest binary?
The latest binaries I know:
http://code.google.com/p/wavtoac3encoder/downloads/list

[laserfan]

The latest binaries I know:
http://code.google.com/p/wavtoac3encoder/downloads/list
Thanks! I'm gonna try the x64 version.

[nurbs]

Is there a way to make eac3to able to decode DTS Express mono tracks?
6: DTS Express, English, 1.0 channels, 16 bits, 96kbps, 48kHz
I have Arcsoft 1.1.0.7 installed and that apparently isn't able to cope with such tracks. Would either upgrading the Arcsoft decoder or installing the Nero Decoder which I still have lying around somewhere fix my problem?
If that wouldn't work is there some other software I can use to decode the track to PCM?

Thanks in advance.

[nibus]

Is there a way to make eac3to able to decode DTS Express mono tracks?
6: DTS Express, English, 1.0 channels, 16 bits, 96kbps, 48kHz
I have Arcsoft 1.1.0.7 installed and that apparently isn't able to cope with such tracks. Would either upgrading the Arcsoft decoder or installing the Nero Decoder which I still have lying around somewhere fix my problem?
If that wouldn't work is there some other software I can use to decode the track to PCM?

Thanks in advance.

I tried Arcsoft 1.1.0.0, 1.1.0.7, and 1.1.0.8 and none of them properly decoded mono dts-express. I believe 1.1.0.8 would say it was decoding, but the output file was just silent and too short. Nero wouldn't even start it.

[xkodi]

I tried Arcsoft 1.1.0.0, 1.1.0.7, and 1.1.0.8 and none of them properly decoded mono dts-express. I believe 1.1.0.8 would say it was decoding, but the output file was just silent and too short. Nero wouldn't even start it.

if my memory serves well somewhere in this thread there are samples of mono dts-express files, which i've tried before and the conclusion was that you need to use 'Nero' decoder for such files, because only the 'Nero' decoder can decode them correctly.

[nurbs]

Alright, I checked the Nero decoder and the results were interesting. The output was a 30 second long file where the audio fades in and out. The content of that file is fine, I can understand what the people are saying, but being only 30 seconds long it is of course nowhere near the 2 hours of the source and so not usable.

I also updated Arcsoft to 1.1.0.8 and tried to decode the track and the result plays way too fast, so now I'll try to stretch the track and see if that produces usable results.

If someone has a recommendation for a software where I can slow down the track by an arbitrary factor I would appreciate it. I'll report back if I find a working solution. :thanks:

[nurbs]

I took several tries, but I managed to find the right factor I needed to stretch the audio and was able align it to the video correctly. Certainly not a pretty solution, but it works if you invest some time in it. I still hope a better solution comes up soon :)

[Thunderbolt8]

I have a problem remuxing the south pacific BD (BBC, german BD, 1080i50, 2xDTS-HD Hi-res) to .mkv (video and english DTS-HD Hi-res track only included)

problem seems to be that either the video track then is too short or one audio track too long.

the .m2ts plays fine for its total running time of 1:06:17 with both audio tracks.

but when I remux to .mkv, the length of the video track suddenly becomes 0:49:19 for the English DTS-HD track and the video track, while the runtime of the German DTS-HD track (just as a comparison) still is 1:06:17.
when remuxed, both seem to be in sync despite their short runtime, but the overall runtime of the remux is still 1:06:17 and playback basically stops (if you try to skip a bit forward, otherwise it continues silently without changing of the last frame and without sound) at 0:49:19 and nothing happens any more.

so how come the runtimes of the video and the english DTS-HD track are different now from that of the german track and the runtime of the .m2ts?

(am using arcsoft 1.1.0.1 but not sure if that matters for demuxing only; LAV audio decoder at playback)

edit: the English track is 16-bit and gets patched to 24-bit when being demuxed (the german track already is 24-bit) so maybe thats why the different runtime? but still, I guess that doesnt change why the video is only 49:19 as .mkv and plays till 1:06:17 just fine inside the .m2ts file.

[Snowknight26]

eac3to fails to convert audio data-less WAVs:

eac3to v3.24
command line: eac3to.exe C:\temp\test_audacity.wav C:\temp\test_eac3to.wav
------------------------------------------------------------------------------
WAV, 5.1 channels, 16 bits, 4234kbps, 44.1kHz
Reading WAV...
Original audio track: no audio data.
The WAV writer didn't receive the format information. <ERROR>
Aborted at file position 44. <ERROR>

[tebasuna51]

And what is the problem?

[Richard1485]

I would like to confirm something. I am transcoding an AC-3 file (direct from the source) to another AC-3 file. Eac3to reports a delay of 48ms. I take it that eac3to applied this delay properly even though the file is being slowed down. Here is the relevant portion of the log. (I know that 48ms is only a little more than a frame and is probably not noticeable, but I'd like to know for sure because of the slowdown and in case I am in the same situation with greater audio delays.)

2: AC3, English, 5.1 channels, 384kbps, 48kHz, dialnorm: -27dB, 48ms
[a02] Decoding with libav/ffmpeg...
[a02] Remapping channels...
[a02] Applying RAW/PCM delay...
[a02] Changing FPS from 25.000 to 23.976...
[a02] Encoding AC3 <448kbps> with libAften...
[a02] Creating file "E:\ST.1.TMP\ST.1.The.Motion.Picture.1080p.h264.dd51.ac3"...
[a02] The last (E-)AC3 frame is incomplete and thus gets skipped. <WARNING>

[Thunderbolt8]

And what is the problem?this :p I have a problem remuxing the south pacific BD (BBC, german BD, 1080i50, 2xDTS-HD Hi-res) to .mkv (video and english DTS-HD Hi-res track only included)

problem seems to be that either the video track then is too short or one audio track too long.

the .m2ts plays fine for its total running time of 1:06:17 with both audio tracks.

but when I remux to .mkv, the length of the video track suddenly becomes 0:49:19 for the English DTS-HD track and the video track, while the runtime of the German DTS-HD track (just as a comparison) still is 1:06:17.
when remuxed, both seem to be in sync despite their short runtime, but the overall runtime of the remux is still 1:06:17 and playback basically stops (if you try to skip a bit forward, otherwise it continues silently without changing of the last frame and without sound) at 0:49:19 and nothing happens any more.

so how come the runtimes of the video and the english DTS-HD track are different now from that of the german track and the runtime of the .m2ts?

(am using arcsoft 1.1.0.1 but not sure if that matters for demuxing only; LAV audio decoder at playback)

edit: the English track is 16-bit and gets patched to 24-bit when being demuxed (the german track already is 24-bit) so maybe thats why the different runtime? but still, I guess that doesnt change why the video is only 49:19 as .mkv and plays till 1:06:17 just fine inside the .m2ts file.

[tebasuna51]

I would like to confirm something. I am transcoding an AC-3 file (direct from the source) to another AC-3 file. Eac3to reports a delay of 48ms. I take it that eac3to applied this delay properly even though the file is being slowed down. Here is the relevant portion of the log. (I know that 48ms is only a little more than a frame and is probably not noticeable, but I'd like to know for sure because of the slowdown and in case I am in the same situation with greater audio delays.)

When you decode the source the audio delay is applied to the uncompressed audio exactly to 48 ms.
After the slowdown is applied, then you have a exact sync.

BTW, if you want mix the audio to other video maybe the delay is different.

[Richard1485]

When you decode the source the audio delay is applied to the uncompressed audio exactly to 48 ms. After the slowdown is applied, then you have a exact sync.

BTW, if you want mix the audio to other video maybe the delay is different.

:thanks: That's what I needed to know. (I'm not muxing it to different video: it's the same video with a slowdown. Basically I'm converting to NTSC.) :)

[tebasuna51]

this :p

Sorry but I don't understand how are related the two problems.

"eac3to fails to convert audio data-less WAVs"
A WAV without data is only the header and is nothing to convert.

What is the size of 'test_audacity.wav'?
If is greater than a few bytes maybe you have a corrupt header without a DATA chunk.

"edit: the English track is 16-bit and gets patched to 24-bit when being demuxed"
I can't understand. If you only demux the DTS you can't change anything, only DTS-MA have bitdeph, don't exist other DTS 16bits or 24bits. Please put the log when you extract audios from m2ts.

[Thunderbolt8]

they arent, but you asked what the problem is so I pointed to my problem :)

[tebasuna51]

Then put the eac3to log extracting the audios.

[Thunderbolt8]

eac3to v3.24
command line: eac3to 3) D:\E06.mkv
------------------------------------------------------------------------------
M2TS, 1 video track, 2 audio tracks, 1:06:17, 50i
1: Chapters, 10 chapters
2: VC-1, 1080i50 (16:9)
3: DTS Hi-Res, German, 5.1 channels, 24 bits, 2814kbps, 48kHz
(core: DTS, 5.1 channels, 24 bits, 1509kbps, 48kHz)
4: DTS Hi-Res, English, 5.1 channels, 16 bits, 2046kbps, 48kHz
(core: DTS, 5.1 channels, 16 bits, 1509kbps, 48kHz)
Creating file "D:\E06 - Chapters.txt"...
[v02] Extracting video track number 2...
[a03] Extracting audio track number 3...
[a04] Extracting audio track number 4...
[a04] Patching bitdepth to 24 bits...
[v02] Muxing video to Matroska...
[a03] Creating file "D:\E06 - 3 - DTS Hi-Res, German, 5.1 channels, 24 bits, 2814kbps, 48kHz.dtshr"...
[a04] Creating file "D:\E06 - 4 - DTS Hi-Res, English, 5.1 channels, 16 bits, 2046kbps, 48kHz.dtshr"...
Added fps value (25) to MKV header.
Video track 2 contains 73977 frames.
eac3to processing took 5 minutes, 10 seconds.
Done.

doesnt seem to help much though, as there is nothing wrong you could see. all I can say now is the runtime of the video .mkv file is 49:19, same for the german audio track and the english audio track is 1:06:17.

funny thing is when muxing english track and video together, the final file still has a runtime of 1:06:17 (as the english track), but video finishes at 49:19 and still is completely in sync with the audio. still, there is then ~15 minutes of nothing at the end of the file (still last frame and no sound)

[tebasuna51]

Yes, don't help too much.
If you extract the 'core' work fine the mkv?
Maybe a problem with DTS-HR and mkv, I never test this.

[Thunderbolt8]

its the same. but anyway, that wouldnt explain why the runtime of the video is now shorter as well (as said, the .m2ts play fine till the end of 1:0617)

its strange that eng track has a longer runtime on the one hand, but is as long as the video when both are remuxed together. is there some form of speedup involved?

[me7]

I installed the trail version of the latest ArcSoft TotalMedia Theater to get access to their DTS-MA decoder, but eac3to fails to find the decoder. Do I need some additional configuration?

[73ChargerFan]

[code]eac3to v3.24
command line: eac3to 3) D:\E06.mkv
------------------------------------------------------------------------------
M2TS, 1 video track, 2 audio tracks, 1:06:17, 50i
1: Chapters, 10 chapters
2: VC-1, 1080i50 (16:9)


Perhaps try using tsMuxeR to demux the m2ts into elementary streams first. I had difficulty remuxing 1080i60 extras from a bd a year ago, and madshi told me eac3to sometimes didn't work with those sources.