eac3to - audio conversion tool

[MuLTiTaSK]

@Snowknight26

stax76 is just demonstrating eac3to switches he did a great job creating a GUI for StaxRip check it out -> StaxRip 1.1.5.1 (bit.ly/cu2A2W)
http://img192.imageshack.us/img192/7747/ss20100217232155.png
http://img192.imageshack.us/img192/8356/ss20100217232121.png

[raquete]

using this .cue to burn with EACopy

REM GENRE Alternative
REM DATE 2005
REM DISCID EA12B813
REM COMMENT "ExactAudioCopy v0.99pb5"
PERFORMER "Coldplay"
TITLE "Greatest Hits (CD 1)"
FILE "Coldplay - Greatest Hits (CD 1).wav" WAVE
TRACK 01 AUDIO
TITLE "Square One"
PERFORMER "Coldplay"
...etc

i get 'error' on line 7(file type not supported) if i decompress .flac to .wav with eac3to
(eac3to "Coldplay - Greatest Hits (CD 1).flac" "Coldplay - Greatest Hits (CD 1).wav")

but when i decompress the .flac to .wav with adobe audition and using the same .cue, EACopy don't give error message and burn the media.

what can be?

[nurbs]

Sampling rate or bitdepth? Or maybe eac3to writes the wrong kind of wav (w64) by default. Would be helpful if you posted what eac3to has to say about the file.

[raquete]

Sampling rate or bitdepth? Or maybe eac3to writes the wrong kind of wav (w64) by default. Would be helpful if you posted what eac3to has to say about the file.

.flac Sampling rate 44100, bitdepth 16
command line to decompress with eac3to:
eac3to "Coldplay - Greatest Hits (CD 1).flac" "Coldplay - Greatest Hits (CD 1).wav"

loading the same .flac in adobe audition and saving as .wav 44100,16 works.

eac3to don't say anything against the .flac and about the .wav in the log, just decompress without advertences.

the only difference is in size of the .wav:
eac3to: 845.466.452 bytes
audition: 845.466.564 bytes

[Snowknight26]

Try with -simple.

[raquete]

Try with -simple.

i'm sorry, i don't understood...please, explain me what to do.

i forgot to post, here the eac3to log file:

eac3to v3.17
command line: eac3to "Coldplay - Greatest Hits (CD 1).flac" "Coldplay - Greatest Hits (CD 1).wav"
------------------------------------------------------------------------------
FLAC, 2.0 channels, 1:19:53, 16 bits, 864kbps, 44.1khz
Decoding FLAC...
Writing WAV...
Creating file "Coldplay - Greatest Hits (CD 1).wav"...
The original audio track has a constant bit depth of 16 bits.
eac3to processing took 2 minutes, 2 seconds.
Done.

[nurbs]

He means try:
eac3to "Coldplay - Greatest Hits (CD 1).flac" "Coldplay - Greatest Hits (CD 1).wav" -simple

[raquete]

He means try:
eac3to "Coldplay - Greatest Hits (CD 1).flac" "Coldplay - Greatest Hits (CD 1).wav" -simple

work, now EACopy get the file without problems,
:thanks: thank you both but...
i don't saw this option in eac3to help.

why is needed the option -simple?

[setarip_old]

@raquete

See this earlier post:

http://forum.doom9.org/showpost.php?p=1347703&postcount=9592

[raquete]

@ setarip_old

now is clever.
this option could be in the help file and be used as 'default'.

:thanks:

[setarip_old]

@raquete
this option could be in the help file and be used as 'default'.True - but it seems that the author is no longer supporting this program...

[raquete]

@raquete
True - but it seems that the author is no longer supporting this program...

a pity and :confused: i don't understand why he don't support as the program is great, full of features, fast(etc) and is used 'around the world' for lots of people in lots of forums.
i know some GUIs where eac3to was chosed to be used and is used...because works.
always i see questions and answers here and around about eac3to then, is still running, is good and deserve support but we can't discuss the developer tastes...

do you know the complete list of options/features?

[b66pak]

why is eac3to not respecting the M$ WAVE_FORMAT_EXTENSIBLE ChannelMask (http://www.microsoft.com/whdc/device/audio/multichaud.mspx)?

http://www.microsoft.com/whdc/device/audio/multichaud.mspx#E4C

http://www.microsoft.com/whdc/device/audio/multichaud.mspx#EKLAC


after decoding ac3/dts 5.1 channels to .wav i get the wrong channel order!

original file (.dts):

General
Complete name : audio.dts
Format : DTS
Format/Info : Digital Theater Systems
File size : 16.9 MiB
Duration : 1mn 33s
Overall bit rate : 1 510 Kbps

Audio
Format : DTS
Format/Info : Digital Theater Systems
Duration : 1mn 33s
Bit rate mode : Constant
Bit rate : 1 510 Kbps
Channel(s) : 6 channels
Channel positions : Front: L C R, Surround: L R, LFE
Sampling rate : 48.0 KHz
Resolution : 24 bits
Stream size : 16.9 MiB (100%)




the decoded .wav:

General
Complete name : audio.eac3to.wav
Format : Wave
File size : 77.3 MiB
Duration : 1mn 33s
Overall bit rate : 6 912 Kbps

Audio
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Unsigned
Codec ID : 00001000-0000-0100-8000-00AA00389B71
Codec ID/Hint : Microsoft
Duration : 1mn 33s
Bit rate mode : Constant
Bit rate : 6 912 Kbps
Channel(s) : 6 channels
Channel positions : Front: L, C, R, Middle: L, R, LFE should be Front: L, C, R, Surround: L, R, LFE
Sampling rate : 48.0 KHz
Resolution : 24 bits
Stream size : 77.3 MiB (100%)




File ........: audio.eac3to.wav
Size ........: 81017924 bytes

---------------------------------------------- Header Info
ChunkID .....: RIFF
RiffLength ..: 81017916
Container ...: WAVE
SubchunkID ..: fmt (Length: 40)
AudioFormat .: 65534 (WAVE_FORMAT_EXTENSIBLE)
NumChannels .: 6
SampleRate ..: 48000
ByteRate ....: 864000
BlockAlign ..: 18
BitsPerSample: 24
ValidBitsPS .: 24
MaskChannels : 1551 (FL FR FC LF SL SR) should be 63 (FL FR FC LF BL BR)
SubType .....: 1 (Integer)
SubchunkID ..: data (Length: 81017856)
Offset data .: 68
Duration ....: 93.77067 sec., (0h. 1m. 33.77067s.)
------------------------------------------------- End Info
_

[Snowknight26]

It's hard to tell if you really care or not when you say 'M$.'

[b66pak]

sorry...Microsoft WAVE_FORMAT_EXTENSIBLE ChannelMask...

a quick fix for the people who need it (for a 5.1 channel 24 bit@48000Hz .dts):

eac3to audio.dts stdout.raw | wavfix - audiofixed.wav -ignorelength -m 63 -i 24 -c 6 -s 48000

the result:

Front: L, C, R, Surround: L, R, LFE
MaskChannels : 63 (FL FR FC LF BL BR)

[Snowknight26]

Actually, a ChannelMask of 0x0000060F is recommended by Microsoft post XP SP2 for 5.1 channel WAVs. 0x0000003F is the old ChannelMask and is also compatible. Seems like it's merely a cosmetic difference.

[b66pak]

Actually, a ChannelMask of 0x0000060F is recommended by Microsoft post XP SP2 for 5.1 channel WAVs. 0x0000003F is the old ChannelMask and is also compatible. Seems like it's merely a cosmetic difference.


i am asking because i use apple's aacencoder (qtaacenc (http://tmkk.hp.infoseek.co.jp/qtaacenc/)) and with ChannelMask 1551 (0x0000060F) i get empty sorround channels (SL SR)...
_

L.E. this is now fixed in the new version of qtaacenc...
_

[Biggiesized]

Whenever I try to run an .mlp file through eac3to, I get an error claiming the program cannot determine the source file. I thought eac3to could decode the MLP file format?

[tebasuna51]

@Biggiesized
Work with all my mlp files.
Upload a sample to test.

[Biggiesized]

@Biggiesized
Work with all my mlp files.
Upload a sample to test.
http://rapidshare.com/files/354338838/track-01-02_1_-01-_L-R_-24-96000.mlp

[mrr19121970]

It's a longshot, I know as Madshi isn't active currently on this project. I'm in the process of converting my HD-DVDs (again) to Blu-Ray. This time I'm going the path E-AC3 to DTS-MA using "DTS-HD Master Audio Suite v2.0 (http://store.dts.com/)". Is there any chance that you can implement a similar 'hooking' to this application as you do with SurCode?

[Inspector.Gadget]

Why go lossy -> lossless? Makes no sense.

[TinTime]

Doesn't it make more sense from a quality point of view to convert it to a lossless format rather than another lossy one?

I mean, given that HD DVD DD+ has to be converted to something else for Blu-ray compatibility.

[Inspector.Gadget]

In a strictly technical sense, sure, but from the perceptual side? I don't transcode audio where I can get away with it, and like a lot of people I'd rather have lossless if the source itself is lossless, but once you're transcoding anyway, I doubt you the vast majority of people can tell the difference between DTS-MA and Dolby Digital @ 640 kbps.

[tebasuna51]

http://rapidshare.com/files/354338838/track-01-02_1_-01-_L-R_-24-96000.mlp
Yes eac3to don't recognize the track, ffmpeg can't decode it (lot of errors) and Foobar2000 crash when try play the file.

Can you play this file with any player?

[Biggiesized]

No, hence my problem. I get a strange, loud glitching noise when I try to play it in foobar2000 with the proper DVD-Audio component installed.

[mrr19121970]

Doesn't it make more sense from a quality point of view to convert it to a lossless format rather than another lossy one?

I mean, given that HD DVD DD+ has to be converted to something else for Blu-ray compatibility.

What would be a comparible DTS conversion for Blu-Ray format (as HD-DVD e-AC3 is not compatible)? DTS-HD I suppose is best (and not DTS-HA Master Audio as I though before).

Thanks for the reply.

[Inspector.Gadget]

DTS-MA is lossless. What I was saying to TinTime, though, was that transcoding from a lossy format to another lossy format seems more sensible to me provided that the destination file is transparent to the ear because there's no intrinsic value in retaining a lossless-from-the-master-copy format anymore. On the other hand, TinTime made the point that lossy->lossless will retain the mathematical quality of the original. This is essentially a philosophical debate, because you almost certainly won't be able to hear the difference between DTS-MA and 640kbps AC3.

[Snowknight26]

TinTime made the point that lossy->lossless will retain the mathematical quality of the original.

It won't though. Lossy to lossless is a lossy process.

[TinTime]

It won't though. Lossy to lossless is a lossy process.

...yes but the lossy source has to be decoded at some point before you can listen to it anyway. Whether that's at playback or offline beforehand doesn't really matter.

I'll approach the question differently.

mrr19121970 said that this is being done for Blu-ray compatibility. Presumably that means burning the result to a BD. In that case the audio codec choice should be straightforward - choose the highest quality you can that will fit on the blank disc. If there's room for DTS-HD MA then why not use it?

[Inspector.Gadget]

It won't though. Lossy to lossless is a lossy process.

You're of course correct. I should have said will retain MORE of the mathematical quality of the original because there's no data discarded for compression purposes.

[mrr19121970]

I'll just go E-AC3 to PCM. It's quick & easy. Space is not an issue.

On a side note, I noticed the following:

Converting to DTSHD-MA saved about 40% in space over the .WAV produced by eac3to. For some strange reason converting to DTSHD actually produced files larger than DTSHD-MA.

[jj666]

DTS-HD HR is CBR, DTS-HD MA is VBR, I suppose the compression is better that's all.

Cheers,

-jj-

[deado]

With the Lord of the Rings trilogy just announced as DTS-HD MA 6.1, I'm guessing it will be just like the Harry Potters and The Number 23 where it has a DTS-ES core and the MA information is not extracted as 6.1 by eac3to, but only played properly when decoded by a player such as the PS3.

Is there any way that eac3to can convert such tracks to 6.1 FLAC? Or are we simply limited to 5.1 for those?

[Inspector.Gadget]

deado - are you using the Arcsoft DTS decoder?

[deado]

deado - are you using the Arcsoft DTS decoder?

Yes I am, version 1.1.0.7. I just tried with Terminator 2 Skynet and eac3to only detects 5.1 channels, not 6.1. So there seems to be no way to properly extract the 6.1?

[Snowknight26]

If eac3to detects it as having 5.1 channels, it's 5.1 channel audio. The ArcSoft decoder isn't used until you actually start decoding the audio.

[deado]

Hmm well that's false advertising stating 6.1 all over the back cover... that goes for both Lionsgate (T2 Skynet) and Warner (HP 1& 2 ultimate, Number 23, LOTR trilogy).

They should say 5.1 ES not 6.1. X-men 3 seems to be pretty much the only BD which is marked as 6.1 which is actually encoded as such.

Here I thought I might have been able to watch LOTR trilogy in 8-channel FLAC, rather than 6-channel FLAC :(

[tebasuna51]

Hmm well that's false advertising stating 6.1 all over the back cover... that goes for both Lionsgate (T2 Skynet) and Warner (HP 1& 2 ultimate, Number 23, LOTR trilogy).

They should say 5.1 ES not 6.1. X-men 3 seems to be pretty much the only BD which is marked as 6.1 which is actually encoded as such.

Here I thought I might have been able to watch LOTR trilogy in 8-channel FLAC, rather than 6-channel FLAC :(
There are two kinds of DTS-ES, and eac3to report like:

1) DTS-ES, 6.1 channels
Called DTS-ES discrete, and the Back Center is encoded digitally like a independent channel.
ArcSoft decode this like 6.1 channels

2) DTS-ES, 5.1 channels
Here the Back Center channel is encoded in Back Left/ Back Right channels in analog mode. This is DTS ES-matrixed.
When you send this stream (SPDIF or HDMI) to a receiver DTS-ES compliant you can obtain a 6.1 output.
ArcSoft decode this like 5.1 wav, but the Back Center channel is still matrixed in Back channels.
When you send this wav (or Flac decoded) to a receiver like PCM 5.1 (only by HDMI) you can obtain a phantom Back Center channel.

[deado]

There are two kinds of DTS-ES, and eac3to report like:

1) DTS-ES, 6.1 channels
Called DTS-ES discrete, and the Back Center is encoded digitally like a independent channel.
ArcSoft decode this like 6.1 channels

2) DTS-ES, 5.1 channels
Here the Back Center channel is encoded in Back Left/ Back Right channels in analog mode. This is DTS ES-matrixed.
When you send this stream (SPDIF or HDMI) to a receiver DTS-ES compliant you can obtain a 6.1 output.
ArcSoft decode this like 5.1 wav, but the Back Center channel is still matrixed in Back channels.
When you send this wav (or Flac decoded) to a receiver like PCM 5.1 (only by HDMI) you can obtain a phantom Back Center channel.

Hm, damn well I'm just outputting over analog with madFLAC -> ffdshow -> Reclock WASAPI so is there anything in ffdshow that can do the same thing as receivers do to obtain that rear channel?

[tebasuna51]

Hm, damn well I'm just outputting over analog with madFLAC -> ffdshow -> Reclock WASAPI so is there anything in ffdshow that can do the same thing as receivers do to obtain that rear channel?
over analog? I don't know sound cards with 6.1 output (only 5.1 or 7.1)

AFAIK ffdshow can't extract the back center channel.

[deado]

over analog? I don't know sound cards with 6.1 output (only 5.1 or 7.1)

AFAIK ffdshow can't extract the back center channel.

It's 7.1... but I was using the eac3to -double7 command to double the rear channel on the couple of 7-channel FLAC files I have (X-Men 3 and Crash Australian version), to get 8-channel FLAC.

But anything which isn't originally discrete 7-channel and only 6-channel with ES it seems I'm out of luck :(

[Snowknight26]

Hmm, after audio overlaps are detected, it seems that eac3to no longer runs audio encoding concurrently. Say extract and convert two DTS-HD MA tracks to FLAC and both of them have audio overlaps, you'd expect eac3to to run the 2nd pass, fix the overlaps, and do the conversion of both at the same time, but the 2nd track's conversion begins only after the first one has finished. :|

[utenteanonimo64]

Sorry if this has been covered already but this thread is overwhelming to search.


eac3to v3.17
command line: "c:\program files (x86)\eac3to\eac3to.exe" track01.dts "01 - .flac"
------------------------------------------------------------------------------
DTS-96/24, 5.1 channels, 0:03:12, 24 bits, 1510kbps, 96khz
The ArcSoft and Sonic decoders don't seem to work, will use libav instead.
Decoding with libav/ffmpeg...
Remapping channels...
Reducing depth from 64 to 24 bits...
Encoding FLAC with libFlac...
Creating file "01 - .flac"...
eac3to processing took 24 seconds.
Done.


The FLAC file I get is 48KHz not 96KHz. Is this normal? Why? (I suppose because only the core DTS is decoded). Do I need one of the external decoders to decode the full DTS stream and maintain 96KHz?

Thanks.

Appendix: I have downloaded and installed a trial copy of ArcSoft TotalMedia. eac3to can't find the DTS decoder even after adding the ArcSoft decoder directory to the Windows PATH variable (which by the way it is not what the first post says but it's "C:\Program Files (x86)\ArcSoft\TotalMedia Theatre 3\Codec".
Any help on this additional problem?

[TinTime]

The FLAC file I get is 48KHz not 96KHz. Is this normal? Why? (I suppose because only the core DTS is decoded). Do I need one of the external decoders to decode the full DTS stream and maintain 96KHz?

Yes. You need to use the ArcSoft or Sonic decoder for full 96/24 decoding.

[utenteanonimo64]

Yes. You need to use the ArcSoft or Sonic decoder for full 96/24 decoding.

OK Thanks. So now I know I need to make this ArcSoft decoder work. Any suggestions? I am using Windows 7 64bit and I installed the trial version of TotalMedia Theatre 3. As far as I can see the DTS decoder is in this folder:

C:\Program Files (x86)\ArcSoft\TotalMedia Theatre 3\Codec

and

dtsdecoderdll.dll is version 1.1.0.7 which according to madshi's instructions should be compatible with eac3to.

As I said I have added the above directory to PATH enviroment variable and I have also copied all dts dlls into the eac3to directory. But if I run "eac3to -test" eac3to cannot find any external decoder. Is it because TMT is a trial version?

Any help is appreciated.

[TinTime]

Can't help you there I'm afraid - it just worked on my system (WinXP) without any fiddling about.

Try looking through this thread (http://forum.doom9.org/showthread.php?t=148324).

[b66pak]

look at this:

http://forum.doom9.org/showthread.php?t=148324
_

[utenteanonimo64]

look at this:

http://forum.doom9.org/showthread.php?t=148324
_

Thanks! I followed the instructions in this message:

http://forum.doom9.org/showthread.php?p=1255699#post1255699

and it worked.

:thanks:

[madshi]

A while back you mentioned that you were working or had implemented a plugin system for eac3to with support for plugging custom filters. Would that suit my needs (I'm a programmer myself)? Any documentation for it?
It's not ready yet.

i have the Freedom bluray set and i would like to remux the english versions but eac3to can't do it correctly, as the playlists use only parts of certain m2ts files. it's the only bluray where i've seen such playlists.
I was aware of that this in theory is possible, but I haven't seen such a Blu-Ray myself yet. Currently eac3to does not support this kind of Blu-Ray. I'd have to have such a full Blu-Ray myself, then I could add support for that. But honestly, just one Blu-Ray doing this is not enough to motivate me to work on implementing such a complicated thing.

any chance to update the lavc lib ? there are EAC3 Spectral & AAC LATM patches floating around in the FFmpeg-dev ML...
I should really do that, but I'm a bit afraid, because it will be quite a bit of work and I'll need to retest all libav decoders etc. Will come sooner or later, but not right now...

A aesthetic bug I noticed is that eac3to reports the dialnorm setting different between Dolby and DTS. For a TrueHD track it might say -27dB (the DN setting), but for a DTS-MA track it says -4dB (the DN offset). It's unlikely the DTS-MA track was actually set 27dB lower than standard.
I've researched this. The AC3 and DTS specs are very kinda the opposite of each other in terms of dialnorm:

AC3: The dialnorm value tells us how much headroom there is between 100% volume and speech volume.
DTS: The dialnorm value tells us how much the decoder should lower volume.

Because of that a -27dB AC3 dialnorm value equals a -4dB DTS dialnorm value. Now of course eac3to could change the display of the dialnorm values, so that AC3 and DTS match. But I'm not sure if I should do that? What do you guys think?

I was little bit inaccurate, libsox is library under LGPL, so it can be used
Oh, didn't know that - thanks! However, right now I don't feel the need to use sox, as eac3to's current resampling algorithm seems to work pretty well...

Any News about the libav Lossless check failed on TrueHD Tracks with libavcodec ?
What news do you expect? Do a search in this thread. All has been explained.

What is the status of the PGS subtitle issue regarding Decoder Time Stamp? Thanks.
Ehm, not fixed yet... :(

i noticed that eac3to isn't capable of removing dialnorm from dts master tracks. is this a known issue?
eac3to can remove dialnorm from the core. However, eac3to at this point in time can not modify the DTS-HD data blocks, because there's no full DTS-HD specification available. I wouldn't know what to change and where. To my best knowledge, the Sonic DTS decoder ignores dialnorm, anyway. Not sure about ArcSoft, though...

Due to Miramax giving lossless honours to the English dub only on their Hero blu-ray, I've remuxed the original Mandarin DTS-ES track from the R3 DVD with an x264 encode of the blu-ray video. However, I needed to add 7900ms of delay for the A/V to sync up and when I put the DTS track through eac3to for this the resulting track has 7900ms of 'noise' at the beginning instead of silence.
eac3to can not "create" new silent DTS frames which are needed for this delay. Instead eac3to simply repeats the very first frame of the audio track as often as necessary to achieve the wanted delay. Normally a DVD audio track begins silent, so this usually works. If your DVD audio track begins non-silent, then eac3to's way of delaying isn't silent, but results in ugly noise, sadly. If there was a free DTS encoder available, I could solve this, but there isn't. For AC3 I'm using Aften to do the audio delay properly and guaranteed-noise-free.

delaycut can do this better, but since it doesn't have a DTS encoder, either, it simply uses a default silent DTS frame - which doesn't have the DTS-ES flag set, of course.

I'm an old-school GNU/Linux command line junkie. Running eac3to inside
screen (http://en.wikipedia.org/wiki/GNU_Screen), the output comes out like this: [...]
eac3to was not written with Linux in mind. I've zero knowledge about Linux. But I thought Linux was quite good at emulating Windows? If so, why can't it even simulate colored command line output? eac3to isn't doing anything special, it's using the official win32 APIs for colored command line output. IMHO you should ask the Linux guys (wine?) to fix the problem.

I'd also love to have an option to not play sound.
Just delete or rename the wav files.

apparently, wine devs
don't want apps to detect wine (http://forum.winehq.org/viewtopic.php?p=25906&sid=4a6025570ec9cb97aa5e4d952cf6ad88)
Great! So ask them to fix the problem with colored command line output. Done. Next?

I think I found bug.
eac3to 00001.m2ts 1: chapters.txt 3: audio1.m4a -no2ndpass 6: audio2.m4a -quality=0.6 -no2ndpass 9: english.sup
It starts, spawns 2 nero encoders, processes whole input and then one Nero finishes, but 2nd does not and stays idle.
I'll add that to my to do list.