eac3to - audio conversion tool

[LigH]

^ omitting the space between "-" and "core".

[Thunderbolt8]

there hasnt been any activity in the dcadec department on github for 2 months now. can we conclude from that the development of the decoder is finished and everything is working?

[hello_hello]

Originally Posted by laz1989

Nop.Try already.
I have been thinking it s possible to be an error from DirectShow so i changed the permision to nero, nothing.
SO my question is : At that command line input.thd ouput.ac3 , it's working at someone?

The way I understand TrueHD (someone correct me if I'm wrong) there is no "core" as there is with DTS-HD. Even though it only appears as a single audio stream, there's two independent streams, the TrueHD audio and the fall-back, lossy AC3. Which you extract appears to depend on the output extension.

The -core parameter only applies to DTS-HD, according to this:
https://en.wikibooks.org/wiki/Eac3to...nd_Line_Syntax

I don't use eac3to via the command line much. I tend to use the HD Streams Extractor built into MeGUI instead (or there's a standalone version here). But I tried a few extractions and checked the command line MeGUI was using. The same TrueHD+AC3 stream was selected each time.

When extracting both streams as a single file with a thd+ac3 extension:
"C:\Program Files\MeGUI\tools\eac3to\eac3to.exe" "E:\Video\" 1) 3:"D:\F1_T3_Audio - English.thd+ac3" -progressnumbers

TrueHD only:
"C:\Program Files\MeGUI\tools\eac3to\eac3to.exe" "E:\Video\" 1) 3:"D:\F1_T3_Audio - English.thd" -progressnumbers

AC3 only:
"C:\Program Files\MeGUI\tools\eac3to\eac3to.exe" "E:\Video\" 1) 3:"D:\F1_T3_Audio - English.ac3" -progressnumbers

Or when opening the m2ts file directly, rather than let the HD Streams Extractor open the whole disc and find the appropriate mpls file, or whatever it does.

"C:\Program Files\MeGUI\tools\eac3to\eac3to.exe" "E:\Video\BDMV\STREAM\00007.m2ts" 3:"D:\T3_Audio - English.ac3" -progressnumbers

eac3to v3.29
command line: "C:\Program Files\MeGUI\tools\eac3to\eac3to.exe" "E:\Video\BDMV\STREAM\00007.m2ts" 3:"D:\T3_Audio - English.ac3" -progressnumbers
------------------------------------------------------------------------------
M2TS, 1 video track, 3 audio tracks, 1:54:39, 50i
1: Chapters, 21 chapters
2: h264/AVC, 1080i50 (16:9)
3: TrueHD/AC3, English, 5.0 channels, 96kHz
(embedded: AC3, 5.0 channels, 640kbps, 48kHz)
4: RAW/PCM, English, 2.0 channels, 16 bits, 48kHz
5: AC3, English, 2.0 channels, 224kbps, 48kHz
[a03] Extracting audio track number 3...
[a03] Extracting AC3 stream...
[a03] Creating file "D:\T3_Audio - English.ac3"...
Video track 2 contains 171972 frames.
eac3to processing took 7 minutes, 31 seconds.
Done.

When I tried each extension, the thd+ac3 file was 5.7GB, the thd file was 5.2GB and the AC3 was 525MB, so that seems right.

Hopefully the info above will help you out.

[turab]

There is this DTS-ES Matrix track that I want to encode to AAC, but I want to know if it's possible to preserve the channel information. Does AAC even support a 6.1 channel configuration?

[tebasuna51]

@turab
Yes, AAC support 6.1 channel configuration, but a DTS-ES Matrix have only 5.1 discrete channels, the Back Center channel is mixed in surround channels.

AFAIK AAC headers don't have any flag to inform the player about that.
But I think than is not necesary, you can recode your DTS to a 5.1 AAC and, if you have a 5.1 audio speakers you can listen the Back Center like a fantom channel, if you have a 6.1 audio speakers your receiver/amplifier maybe can extract the Back Center channel for you.

There are a option, not recommended, to convert a 6.1 Matrix in 6.1 discrete channels:
1) Decode 6.1 Matrix to 6 wavs with eac3to
2) Use SL and SR wavs to create a stereo wav (with Sox, WaveWizard or any audio editor like Audacity)
3) Use CenterCutGUI to create 3 channels
4) Make a 6.1 wav replacing SL/SR with the 3 new channels and encode to AAC

[heerschop]

Originally Posted by tebasuna51

@turab
There are a option, not recommended, to convert a 6.1 Matrix in 6.1 discrete channels:
1) Decode 6.1 Matrix to 6 wavs with eac3to
2) Use SL and SR wavs to create a stereo wav (with Sox, WaveWizard or any audio editor like Audacity)
3) Use CenterCutGUI to create 3 channels
4) Make a 6.1 wav replacing SL/SR with the 3 new channels and encode to AAC

With CenterCut I have created a center.wav and a sides.wav from the combined stereo SL-SR.wav.
- Do you use the center.wav as the center back channel?
- How do you create the wavs for replacing the SL and SR channel? Is this accomplished by converting the stereo center.wav to a center-left.wav and center-right.wav?

greetz

[Nebudchanezzer]

Originally Posted by LigH

^ omitting the space between "-" and "core".

True, thats how it goes sometimes using a android-pad to write on.

[tebasuna51]

Originally Posted by heerschop

With CenterCut I have created a center.wav and a sides.wav from the combined stereo SL-SR.wav.
- Do you use the center.wav as the center back channel?

Yes, center.wav is now BC.wav

- How do you create the wavs for replacing the SL and SR channel? Is this accomplished by converting the stereo center.wav to a center-left.wav and center-right.wav?

Nope, use eac3to to split sides.wav to wavs and rename the outputs FL-FR to SL-SR.

[turab]

Originally Posted by tebasuna51

@turab
Yes, AAC support 6.1 channel configuration, but a DTS-ES Matrix have only 5.1 discrete channels, the Back Center channel is mixed in surround channels.

AFAIK AAC headers don't have any flag to inform the player about that.
But I think than is not necesary, you can recode your DTS to a 5.1 AAC and, if you have a 5.1 audio speakers you can listen the Back Center like a fantom channel, if you have a 6.1 audio speakers your receiver/amplifier maybe can extract the Back Center channel for you.

There are a option, not recommended, to convert a 6.1 Matrix in 6.1 discrete channels:
1) Decode 6.1 Matrix to 6 wavs with eac3to
2) Use SL and SR wavs to create a stereo wav (with Sox, WaveWizard or any audio editor like Audacity)
3) Use CenterCutGUI to create 3 channels
4) Make a 6.1 wav replacing SL/SR with the 3 new channels and encode to AAC

Thank you. So when it's encoded to 5.1 AAC, the audio can still be decoded to 6.1 (if the right software/hardware exists). I was thinking that maybe DTS-ES streams have some side information that's needed for matrix decoding that would get lost in the process. If not, then I'm happy to encode to 5.1 AAC.

[Mike]

hello friends I'm here with a doubt ..
I have a 384kbps audio to and can convert to 640kbps

It is allowed to convert audio unless kbps kbps for more ...

Or it can only be converted and is allowed to convert more kbps audio for less ...

thank you

[ndjamena]

Bitrate is irrelevant. Audio is decoded to PCM before being passed to an encoder, which has a far higher bitrate than any lossy codec.

[LigH]

But you will not raise quality by raising bitrate. What is lost in the original, can't be restored in the copy.

[Smithy]

Terminator 2 Judgment Day Skynet Edition 1991 Blu-ray 1080p EUR VC-1 DTS-HD MA

eac3to v3.28 (arcsoft 1.1.0.0)

M2TS, 2 video tracks, 9 audio tracks, 19 subtitle tracks, 2:17:19, 24p /1.001
6: DTS Master Audio, German, 7.1 channels, 24 bits, 48kHz
(core: DTS, 5.1 channels, 1509kbps, 48kHz)
[a06] Extracting audio track number 6...
[a06] Decoding with ArcSoft DTS Decoder...
[a06] Writing WAVs...
[a06] Skipping identical DTS frames (seamless branching)...
[a06] Original audio track: max 24 bits, average 16 bits, most common 16 bits.
[a06] Audio overlaps for 9ms at playtime 0:18:31. <WARNING>
[a06] Audio overlaps for 5ms at playtime 0:39:24. <WARNING>
[a06] Audio overlaps for 8ms at playtime 0:39:42. <WARNING>
[a06] Audio overlaps for 6ms at playtime 1:04:51. <WARNING>
[a06] Audio overlaps for 11ms at playtime 1:07:11. <WARNING>
[a06] Audio overlaps for 8ms at playtime 1:08:47. <WARNING>
[a06] Audio overlaps for 7ms at playtime 1:12:08. <WARNING>
[a06] Audio overlaps for 5ms at playtime 1:21:17. <WARNING>
[a06] Audio overlaps for 12ms at playtime 1:34:04. <WARNING>
[a06] Audio overlaps for 10ms at playtime 1:56:49. <WARNING>
[a06] Audio overlaps for 11ms at playtime 2:01:59. <WARNING>
[a06] Audio overlaps for 8ms at playtime 2:05:06. <WARNING>
[a06] Audio overlaps for 9ms at playtime 2:11:54. <WARNING>
[a06] Starting 2nd pass...
[a06] Extracting audio track number 6...
[a06] Decoding with ArcSoft DTS Decoder...
[a06] Writing WAVs...
[a06] Realizing RAW/PCM gaps...
[a06] Skipping identical DTS frames (seamless branching)...
[a06] Processed audio track: max 24 bits, average 16 bits, most common 16 bits.


eac3to v3.29 (dcadec)

M2TS, 2 video tracks, 9 audio tracks, 19 subtitle tracks, 2:17:19, 24p /1.001
6: DTS Master Audio, German, 7.1 channels, 24 bits, 48kHz
(core: DTS, 5.1 channels, 1509kbps, 48kHz)
[a06] dts, 48000, 7.1
[a06] Extracting audio track number 6...
[a06] Decoding with libDcaDec DTS Decoder...
[a06] Writing WAVs...
[a06] The libDcaDec DTS Decoder reported the error "Bitstream navigation error" while decoding. <ERROR>
Aborted at file position 16039026688. <ERROR>

[Smithy]

Originally Posted by Mike

hello friends I'm here with a doubt ..
I have a 384kbps audio to and can convert to 640kbps

It is allowed to convert audio unless kbps kbps for more ...

Or it can only be converted and is allowed to convert more kbps audio for less ...

thank you

Originally Posted by LigH

But you will not raise quality by raising bitrate. What is lost in the original, can't be restored in the copy.

thats right in theory,
but eac3to/libav ac3 encoder has wrong bandwidth in lower Bitrates for 5.1 like 384 (14 kHz) / 448 (16 kHz) vs Studio AC3 384 (18 kHz) / 448 (20 kHz)

To Save all Frequencies (no cutoff) from Source AC3 to Reencode AC3 with eac3to (libav), the min. Bitrate for 384/448 kbps is 576 kbps @ 5.1,
but 640 kbps are better choise because lossy Encode from lossy Source.
A Speedup for AC3 5.1 384 raise Frequencies near 20 kHz, so u need more Bitrate/Bandwidth for Reencode.
........
Better use Surcode for AC3 Encoder for lower Bitrates,
or AftenGui and EncWAVtoAC3 have bandwith Option from -2 to 60 (5.1 384 kbps 18kHz = 40 / 5.1 448 kbps 20kHz = 48)

[LigH]

To not lose quality, preferably don't recode (if the source format is supported by the target device). If you convert between different formats (e.g. multichannel AAC to AC3), you may of course use different bitrates because the codecs have different efficiency.

[tebasuna51]

Originally Posted by turab

I was thinking that maybe DTS-ES streams have some side information that's needed for matrix decoding that would get lost in the process.

To matrix decode SL-SR to SL'-BC-SR' you don't need a special side info.
The common parts between SL-SR are extracted to BC channel and elimitated from original SL-SR to output a new pair SL'-SR'.

Only DTS-ES 6.1 discrete have info to do a better channel separation:
http://www.avsforum.com/forum/90-rec...ml#post4224284

[Music Fan]

Originally Posted by tebasuna51

To matrix decode SL-SR to SL'-BC-SR' you don't need a special side info.
The common parts between SL-SR are extracted to BC channel and elimitated from original SL-SR to output a new pair SL'-SR'.

In this case, what's the difference between 5.1 and 6.1 matrix mix ? Because with a 6.1 (or 7.1) receiver, both can be converted to 6.1 (or 7.1).

[tebasuna51]

Originally Posted by Smithy

thats right in theory,
but eac3to/libav ac3 encoder has wrong bandwidth in lower Bitrates for 5.1 like 384 (14 kHz) / 448 (16 kHz) vs Studio AC3 384 (18 kHz) / 448 (20 kHz)

To Save all Frequencies (no cutoff) from Source AC3 to Reencode AC3 with eac3to (libav), the min. Bitrate for 384/448 kbps is 576 kbps @ 5.1,
but 640 kbps are better choise because lossy Encode from lossy Source.
A Speedup for AC3 5.1 384 raise Frequencies near 20 kHz, so u need more Bitrate/Bandwidth for Reencode.
........
Better use Surcode for AC3 Encoder for lower Bitrates,
or AftenGui and EncWAVtoAC3 have bandwith Option from -2 to 60 (5.1 384 kbps 18kHz = 40 / 5.1 448 kbps 20kHz = 48)

1) Of course if you need a re-encode because a speedup operation you can use, at your choice, higer bitrate output. But remember than speedup is a lossy operation and you lose quality always.

2) Of course if you own a commercial certified encoder maybe the output is better than use free AC3 encoders. Thats can't be discussed here.

3) But remember than quality is not only bandwith, at same bitrate you can choice between lose bandwith or lose precission.

4) I make a test encoding a Test.waw 5.1 48 KHz with different options:
Test-Aften-w40-384.ac3 (18 KHz)
Test-Aften-w48-448.ac3 (20 KHz)
Test-eac3to-384.ac3 (14 KHz)
Test-eac3to-448.ac3 (16 KHz)
Test-ffmpeg-384.ac3 (18 KHz)
Test-ffmpeg-448.ac3 (20 KHz)
Test-SoftEncode-384.ac3 (18 KHz, default like Studio AC3)
Test-SoftEncode-448.ac3 (20 KHz, default like Studio AC3)

I used spek to see bandwith, I upload the images here: https://www.sendspace.com/file/egd7tu

5) Remember than eac3to can use the external encoder Aften to override the default parameters used with eac3to:

eac3to Test.wav stdout.wav | Aften -b 448 -w 48 - Test-Aften-w48-448.ac3

6) Now ffmpeg is the free AC3 recommended encoder.

[tebasuna51]

Originally Posted by Music Fan

In this case, what's the difference between 5.1 and 6.1 matrix mix ? Because with a 6.1 (or 7.1) receiver, both can be converted to 6.1 (or 7.1).

Only commercial questions.

To obtain a full surround (plane) audio image for a listener with only two ears 5.1 speakers is more than enough. Systems with 6.1 or 7.1 speakers only want gain more money.

Do you have a ear in the nape? Your ears only listen a mix SL/BC and SR/BC or SL/BL and SR/BR.

[Smithy]

Originally Posted by tebasuna51

3) But remember than quality is not only bandwith, at same bitrate you can choice between lose bandwith or lose precission.

4) I make a test encoding a Test.waw 5.1 48 KHz with different options:
Test-Aften-w40-384.ac3 (18 KHz)
Test-Aften-w48-448.ac3 (20 KHz)
Test-eac3to-384.ac3 (14 KHz)
Test-eac3to-448.ac3 (16 KHz)
Test-ffmpeg-384.ac3 (18 KHz)
Test-ffmpeg-448.ac3 (20 KHz)
Test-SoftEncode-384.ac3 (18 KHz, default like Studio AC3)
Test-SoftEncode-448.ac3 (20 KHz, default like Studio AC3)

I used spek to see bandwith, I upload the images here: https://www.sendspace.com/file/egd7tu

1) Of course, example for Precission 5.1 AC3 640 Kbps use Bandwidth @ 48 (20 kHz), too.
Because it don't need more Bandwidth. or better use Bandwith near the Source Frequencies, maybe 18kHz = 40 or 16 kHz = 32 or whatever.
Eac3to/libav use Bandwidth @ 60 (24kHz) for 5.1 AC3 640 Kbps

2) Aften looks best to the Source but SoftEncode looks Crapy !