eac3to - audio conversion tool

[dade49]

Originally Posted by stinman

The only way to to extract a TrueHD atmos stream now is to run a MKV thru MakeMKV and out put the TrueHD atmos to flac, then the flac can be converted to wavs.

This is much simpler. Extract the TrueHD Atmos stream to a multi-channel wav via ffmpeg, and then run that wav through eac3to to dts or whatever.

ffmpeg -i 00000.m2ts -map 0:1 extract.wav
eac3to extract.wav extract.dts

[tebasuna51]

Post about ffmpeg moved to ffmpeg syntax thread.

Please don't use this thread for ffmpeg discussions.

[Arm3nian]

I'm experimenting with the dolby headphone wrapper plugin that's available with foobar, so I'm remapping channels with eac3to to see the effects. My source file is the song from the end of the first hobbit movie (7.1 channel dts-hd ma). I downmix to 5.1 and then analyze the channels with the foobar oscilloscope. The default output of eac3to is L,R,C,LFE,SL,SR, and is verified by the waveforms. The loudest is the center, followed by the left and right, then the two surrounds, and the LFE is near non existent, but is there.

http://imgur.com/DeIVxyd

I did the same thing but remapped the channels to L,R,SL,SR,LFE,C. As can be seen in the oscilloscope, the center is still the loudest (6th waveform), followed by the left and right (first and second waveforms), but the LFE somehow has more activity now and is louder (5th waveform). The SL and SR are closely related in activity and amplitude to each other, just like before remapping, but have lower activity and amplitude than the LFE... and also have lower amplitude than before the remap.

http://imgur.com/mTc9GgA

What's going on? Does eac3to apply different mix levels while disregarding which channel is what and going just based on sequential order, or is foobar doing something.

[tebasuna51]

Originally Posted by Arm3nian

...I did the same thing but remapped the channels to L,R,SL,SR,LFE,C.

Please put the log or at least the command line used to do this.

Does eac3to apply different mix levels while disregarding which channel is what and going just based on sequential order,

Of course, the channel order define which channel is what.

or is foobar doing something.

I don't know what you do with foobar.

[stax76]

Hi everybody,

I've improved StaxRip's eac3to support quite much lately, maybe a few people here have some interest to look at the latest release:

http://forum.doom9.org/showthread.ph...91#post1707791

[Arm3nian]

Originally Posted by tebasuna51

Please put the log or at least the command line used to do this.
Of course, the channel order define which channel is what.

hobbitsongchannel6mix - Log.txt
So if the channels are not in the default order when remapped then they get mixed at wrong levels?

[tebasuna51]

eac3to input71.dt output.wav -down6
finish with:

FL' = FL       (0)
FR' = FR       (1)
FC' = FC       (2)
LF' = LF       (3)
SL' = BL + SL  (4 + 6)
SR' = BR + SR  (5 + 7)

But with remap: eac3to input71.dt output.wav -down6 -0,1,4,5,3,2
finish with:

FL' = FL       (0)
FR' = FR       (1)
FC' = BL       (4)
LF' = BR       (5)
SL' = LF + SL  (3 + 6)
SR' = FC + SR  (2 + 7)

Seems first remap the first 6 channel and after mix the Side channels .

Maybe you need 2 pass, first to only -down6 and after remap with -0,1,4,5,3,2

[Arm3nian]

Originally Posted by tebasuna51

eac3to input71.dt output.wav -down6
finish with:

FL' = FL       (0)
FR' = FR       (1)
FC' = FC       (2)
LF' = LF       (3)
SL' = BL + SL  (4 + 6)
SR' = BR + SR  (5 + 7)

But with remap: eac3to input71.dt output.wav -down6 -0,1,4,5,3,2
finish with:

FL' = FL       (0)
FR' = FR       (1)
FC' = BL       (4)
LF' = BR       (5)
SL' = LF + SL  (3 + 6)
SR' = FC + SR  (2 + 7)

Seems first remap the first 6 channel and after mix the Side channels .

Maybe you need 2 pass, first to only -down6 and after remap with -0,1,4,5,3,2

Yeah two passes works as expected. Madshi can change it to were the order of the commands matter; as of right now it always remaps first. I personally don't care though, just experimenting.

[kevmitch]

This seems to be a semi-popular topic, but I don't think my specifc question has previously been answered.

With lord of the rings 6.1 dts-hd-ma audio

C:\> eac3to lotr.mkv
MKV, 1 video track, 5 audio tracks, 5 subtitle tracks, 2:07:41, 24p /1.001
1: h264/AVC, English, 1080p24 /1.001 (16:9)
2: DTS Master Audio, English, 6.1 channels, 24 bits, 48kHz
   (core: DTS-ES, 6.1 channels, 24 bits, 1509kbps, 48kHz)
   "Surround 6.1"
...

I decode it to flac (arcsoft 1.1.0.0)

C:\> eac3to lotr.mkv 1:lord_of_the_rings.flac

What order will the tracks be in the flac file for example if I open it in audacity?

I have verified that the flac is bit-for-bit identical (including channel order) to the makemkv decode results. This post on the makemkv forum says that the channel order should therefore be

L R C LFE Ls Rs Cs

However, if I produce my own 7 channel flac file (with libsoundfile/audiolab), I find that downmixing it with eac3to

eac3to my_7_channel.flac my_7_channel.stereo.flac -downstereo

emperically uses the following matrix

[[ 1.      0.      0.7071  0.      0.7071  1.      0.    ]
 [ 0.      1.      0.7071  0.      0.7071  0.      1.    ]]

which suggests that it is interpeting my file to have the order

L R C LFE Cs Ls Rs

Is this discrepancy due to differing channel mask between my file and that created by eac3to? Or is my first conclusion incorrect that eac3to places the centre surround channel last in the flac when decoding from dts-hd-ma?

[nevcairiel]

The "official" FLAC 6.1 order is L R C LFE Cb Ls Rs .. unfortunately 6.1 is annoying and both orders are commonly used.
See here: http://xiph.org/flac/format.html#frame_header (scroll down to <4> channel assignment)

7 channels: front left, front right, front center, LFE, back center, side left, side right

The MakeMKV post you referenced is old, since FLAC did add official 7 and 8 channel orders now.
However, it may be nice and write metadata into the FLAC headers to indicate the channel order, which eac3to should hopefully read.

[tebasuna51]

Originally Posted by kevmitch

This post on the makemkv forum says that the channel order should therefore be

L R C LFE Ls Rs Cs

And the info in the post is wrong.

I tested MakeMKV decoding to wav a DTS-MA 6.1 and output the standard 6.1 channel order (the same than eac3to output):

FL FR FC LF BC SL SR

where
BC = Back Center
SL = Side Left
SR = Side Rigth

Talking about channel order please use always the channel names and order defined here to avoid confusions.

The channel names and order inside a encoded format can be differents but, to transcode, we needs always decode to standard WAV order and channels, to allow the new encoder convert the audio to their internal order and channels.

Standard order for 7.1 is (like I put in my precedent post):

FL FR FC LF BL BR SL SR

[Thunderbolt8]

regarding 2.1 DTS-HD MA tracks: Does the arcsoft bug (LFE channel cannot be decoded) still exist in each arcsoft version? Or is there a version which is not affected by the bug?

when I play such a track with MPC-HC & LAV Audio with arcsoft capabilities, I wont hear the LFE channel then? what happens when I transform the 2.1 track with eac3to, using the arcsoft decoder, e.g. into wavs. will the LFE channel then be processed incorrectly? but when I just demux the track nothing happens?

[stax76]

M2TS, 1 video track, 2 audio tracks, 5 subtitle tracks, 24p /1.001
1: Chapters, 24 chapters
2: h264/AVC, 1080p24 /1.001 (16:9)
3: DTS Master Audio, English, 5.1 channels, 24 bits, 48kHz
(core: DTS, 5.1 channels, 24 bits, 1509kbps, 48kHz)
4: AC3 Surround, English, 2.0 channels, 320kbps, 48kHz
5: Subtitle (PGS), English
6: Subtitle (PGS), Swedish
7: Subtitle (PGS), Norwegian
8: Subtitle (PGS), Danish
9: Subtitle (PGS), Finnish
a03 The ArcSoft and Sonic decoders don't seem to work, will use libav instead.
a03 The libav DTS decoder doesn't decode the full DTS-HD information.
Creating file "V:\Blurays and DVDs\Saving Mr Banks\Saving Mr Banks_Chapters.txt"...
v02 Extracting video track number 2...
a03 Extracting audio track number 3...
a03 Extracting DTS core...
a03 Decoding with libav/ffmpeg...
a03 Reducing depth from 64 to 32 bits...
a03 Encoding AAC <0.50> with NeroAacEnc...
v02 Creating file "V:\Blurays and DVDs\Saving Mr Banks\Saving Mr Banks.h264"...
a03 0:21:12 The source file seems to be damaged (sync byte missing).
v02 0:21:13 The source file seems to be damaged (sync byte missing).
v02 0:21:13 The source file seems to be damaged (discontinuity).
a03 0:21:12 The source file seems to be damaged (discontinuity).
a03

Can somebody tell me what could be the reason for this error?

Currently it's not possible that GUIs can read colored eac3to output, because of this I don't have the full error message.

[Snowknight26]

The reason is probably that "the source file seems to be damaged."

[tebasuna51]

Originally Posted by Thunderbolt8

regarding 2.1 DTS-HD MA tracks: Does the arcsoft bug (LFE channel cannot be decoded) still exist in each arcsoft version? Or is there a version which is not affected by the bug?

Using dtsdecoderdll v1.1.0.0 the LFE is decoded without problems.

[GCRaistlin]

Does the DRC bug still exist in 3.27?

[tebasuna51]

Originally Posted by GCRaistlin

Does the DRC bug still exist in 3.27?

Seems you talk about the NERO DRC bug

Now the default decoder for AC3 is libav instead Nero for that problem.

[GCRaistlin]

tebasuna51, using libav leads to another issue, doesn't it?

[nevcairiel]

Originally Posted by GCRaistlin

tebasuna51, using libav leads to another issue, doesn't it?

Those issues have long been fixed.

[r0lZ]

Another question about the Arcsoft DTS decoder. Is it better to use it when decoding "regular" 5.1 DTS tracks? Or is it useful only for the recent 7.1 evolutions? AFAIK, a decoder that is able to decode a specific format (without bugs) should give exactly the same result than another decoder, and therefore there is no advantage to use the slow Arcsoft decoder when decoding a DTS tracks that the libav decoder can also decode properly. Am I right?