eac3to - audio conversion tool

[mindbomb]

For demuxing subtitles with zlib compression in an mkv, eac3to demuxes the file sucessfully, but it doesnt remove the compression.

[cafevn]

hi, if i only need extract DTS-HDMA5.1 audio i don't need arc soft decoder?

[mindbomb]

hi, if i only need extract DTS-HDMA5.1 audio i don't need arc soft decoder?

correct.

[cafevn]

correct.

thank you, i only need remux blu-ray, not encode anythings

[kypec]

thank you, i only need remux blu-ray, not encode anythings
MakeMKV (http://www.makemkv.com/) is probably more suitable for your purpose than eac3to if all you need is demux/remux Bluray streams. ;)

[cafevn]

MakeMKV (http://www.makemkv.com/) is probably more suitable for your purpose than eac3to if all you need is demux/remux Bluray streams. ;)

i have 2-disc blu-ray want to remux to 1 mkv file (titanic 3d)

i thing i need demux 2 disc and use mkvmerge to append to 1 mkv file

that right :D

[dansrfe]

Is there support planned for MP4 files?

[DigitalfreakNYC]

Every time I try to convert an uncompressed wav to thd, I get "audio not supported." Any suggestions?

[filler56789]

Every time I try to convert an uncompressed wav to thd, I get "audio not supported." Any suggestions?

eac3to cannot encode to TrueHD. From the help screen:

Decoded audio data can be stored as / encoded to:
(1) RAW, (L)PCM
(2) WAV (PCM only), W64, RF64, AGM
(3) WAVs (multiple mono WAV files, PCM only)
(4) AC3
(5) DTS
(6) AAC
(7) FLAC

[dansrfe]

Will there be support for stdin any time soon? I need to transcode audio that resides within an MP4 container so I need to pipe MP4Box's audio extract output to eac3to's input then pipe that back to MP4Box to mux the audio and encoded video to MP4.

[Snowknight26]

Use ffmpeg instead?

[thecross]

I really like this tool. Only thing I wish it could handle was multithreading. No matter what I do it seems to saturate only one "core", so on a "quad core" CPU eac3to will max out at 25%. On an "octal core" CPU it will max out at 12.5%. Perhaps there is a command line argument to set threading options or the number of threads to use, or a special super secret "build from source" multithreaded version I am unaware of, but I can't seem to find a way to make eac3to use my entire CPU.

[Q-the-STORM]

any way to speed up audio to a specific framerate?
this works fine:

eac3to source.wav output.wav -25.000 -changeTo24.000

but I would need something like this
eac3to source.wav output.wav -25.000 -changeTo25.650

seems like eac3to only supports the common framerates, but wouldn't it be possible/easy to open it up to more specific values?

[Shevek]

any way to speed up audio to a specific framerate?
this works fine:

eac3to source.wav output.wav -25.000 -changeTo24.000

but I would need something like this
eac3to source.wav output.wav -25.000 -changeTo25.650

seems like eac3to only supports the common framerates, but wouldn't it be possible/easy to open it up to more specific values?

Use Audacity (http://audacity.sourceforge.net/)

[tebasuna51]

Or sox:

sox source.wav output.wav speed 1.026

Where 1.026 = 25.650/25

[dansrfe]

Is there any particular reason why eac3to can't read from and mux to an MP4 container? Is it deliberate by design?

[Kurtnoise]

why not remux your MP4 into MKV instead ?

MKV support in eac3to is not perfect, yes, but it's much better than nothing.

[tebasuna51]

eac3to was born to manage E-AC3 audio (for that the name) in EVO container.
And many other audio conversions.
After was added support for m2ts, ts, VOB, ...
Later a partial support for mkv files.

MP4 support was never added.
Maybe because there are other tools to do the job, maybe because patent conflicts, standard full specs, different formats, ...

[MerolaC]

Hello.
madshi, Thank you a lot EAC3To. It's awesome!

I'm having a doubt. Which is the last version of Nero 7 that works for EAC3To. I have the version 7.10.0 (Activated and all to Premium Edition) and when I do the test in EAC3To, it still says that I don't have Nero 7 installed.
I'm on Windows 7 Ultimate x64
Is that the problem?

In any case, thank you a ton!

[sephirotic]

Please forgive me if this is a common question, i've spent a lot of time searching for answers for my problem with no avail.

So i succesfully decoded a DTS-ES file (6.1 channel) into a flac. I wasn´t satisfied with the file size so i decided to go lossy to save space. At first i considered going with AC3 but 640kbps wouldn´t fit DVD9 discs and i don´t want to reencode my video track again. That's when i started getting worried that AC3 448kbps quality wouldn´t satisfy me enough so i decided going on AAC multichannel. This was the first time i was encoding a 7th channel audio AAC instead of a simple 5.1, eac3to gave me the warning when encoding the audio:
"neroaacenc doesn´t support 6.1 encoding. Will double the 7th channel".
While i do know that AAC is much more efficient than AC3 i started worrying that this extra useless (doubled) channel would kill all the purpose of extra efficiency of the AAC. 8 channels vs 6 channels would represent roughly 30% of extra audio data.
Do people actually transcode 6.1 DTS-ES to AAC or should i stick with DD 448?

Compatibility is not a issue to me since the video track itself is a 10bit encode so i will use a software decoding pc in my living room for playing the file.

Another question is: eac3to does not support Adaptive bit rate of nero? -q settings are getting me very inconsistent sizes/bitrates.
(also, is ABR considerable much less efficient than VBR encoding? should i stay with VBR even if it gives 20kb/s less than a target abr setting?)

Thanks a lot.

[tebasuna51]

Which is the last version of Nero 7 that works for EAC3To.

I have Nero 7.11.10 and work.
BTW, now Nero decoder is needed only for AAC.

[MerolaC]

I have Nero 7.11.10 and work.
BTW, now Nero decoder is needed only for AAC.

Thanks for the answer. I will try reinstalling. But now that you say that. Is the Nero decoder obsolete? Or why is not needed anymore? The AC3 decoder I mean. I already have the one AAC working.


Edit:
Well, I re-read the OP and DRC is applied sometimes, I guess I know now.... Again, thank you tebasuna.

[tebasuna51]

...This was the first time i was encoding a 7th channel audio AAC instead of a simple 5.1, eac3to gave me the warning when encoding the audio:
"neroaacenc doesn´t support 6.1 encoding. Will double the 7th channel".

Preserve a 6.1 audio is only usefull when you have a 6.1 audio equipment (not very common).
If you have 5.1 speakers is better use -down6, the BC channel is mixed in surround channel and you still have the BC like phantom channel.

BTW, NeroAacEnc can encode 6.1. Try this:

eac3to input6.1.dts stdout.wav | NeroAacEnc -q 0.5 -ignorelength -if - -of output.m4a

The problem here is the order in back/rear channels. Maybe your AAC decoder can't restore the correct mapping.

Another question is: eac3to does not support Adaptive bit rate of nero? -q settings are getting me very inconsistent sizes/bitrates.

The NeroAacEnc options are:

Quality/bitrate control:
-q <number> : Enables "target quality" mode.
<number> is a floating-point number in 0...1 range.
-br <number> : Specifies "target bitrate" mode.
<number> is target bitrate in bits per second.
-cbr <number> : Specifies "target bitrate (streaming)" mode.
<number> is target bitrate in bits per second.
When neither of above quality/bitrate options is used,
the encoder defaults to equivalent of -q 0.5

Multipass encoding:
-2pass : Enables two-pass encoding mode.
Note that two-pass more requires a physical file as input,
rather than stdin.

To have a efficient -br encode you need -2pass and need a physical file. eac3to can't do this, decode your input to wav and use NeroAacEnc with the -2pass option.

[buffyangel108]

Does anyone know how I can pipe eac3to output to flac.exe?

I know eac3to has in-built flac encoding (using libFlac), but compression is seemingly set at level 8 which takes a lot longer than level 0 (and I don't need the extra compression).

I've tried:

eac3to file.dts stdout.wav | flac -0 - -o file.flac

...which works on short files but not on movie-length ones, throwing up a foreign metadata error towards the end of encoding. I'm guessing this could be where the stdout.wav temp file reaches the 4GB WAV header limit?

I've also tried doing it in two steps: dts>wav (with eac3to) then wav>flac (with flac), but again wav files larger than 4GB fail to encode in flac. :-(

Does anyone know a workaround? If not, could eac3to.exe or libFlac.dll conceivably be patched to output at level 0 flac compression instead of level 8?

[nhakobian]

...which works on short files but not on movie-length ones, throwing up a foreign metadata error towards the end of encoding. I'm guessing this could be where the stdout.wav temp file reaches the 4GB WAV header limit?


This is both true and false at the same time.

When you pipe something, there is no file created, so there is no filesystem bound filesize limit.

However, the WAV file format internally has a limit of 4GB filesizes due to how data is stored in its internal headers, which are passed between the two programs. Maybe the crash is occurring when this value overflows.

[the_weirdo]

@buffyangel108

You can try adding --ignore-chunk-sizes to FLAC command line and see if that can help. Also, you may try this version (http://lists.xiph.org/pipermail/flac-dev/2013-April/004057.html) of FLAC.

[superhil]

i found a bug, "-edit=0:00:0X,10000ms -silence" produce "invalid edit format" for X=0,1,..,10
but, start from X=11, e.g. "-edit=0:00:11,10000ms -silence" works well.

[tebasuna51]

i found a bug, "-edit=0:00:0X,10000ms -silence" produce "invalid edit format" for X=0,1,..,10
but, start from X=11, e.g. "-edit=0:00:11,10000ms -silence" works well.

It's a know bug. The amount to insert must be lower than the current position.
If you want insert 20 sec. the position must be at last 0:00:21

Maybe eac3to check first if can do a -loop option and show the <ERROR> because don't have enough audio to repeat.
But for -silence option should admit it.

[madshi]

If it's a known bug why is it not in the bug tracker? ;)

http://eac3to.bugs.madshi.net

[superhil]

If it's a known bug why is it not in the bug tracker? ;)

http://eac3to.bugs.madshi.net

then you should add it into bug tracker :D

because "-edit=0:00:00,10000ms -silence" is valid input

[Chumbo]

then you should add it into bug tracker :D

because "-edit=0:00:00,10000ms -silence" is valid input
You find it, you log it is how it works.

[superhil]

You find it, you log it is how it works.
I'm sorry I don't understand what you mean :confused:

That option should be a valid input, right? adding 10s of silence start from 00:00:00 until 00:00:10
But eac3to recognizes it as invalid input

I understand that "-edit=00:00:00,10000ms -loop" is invalid input, but "-silence" should work

[tebasuna51]

That option should be a valid input, right? adding 10s of silence start from 00:00:00 until 00:00:10

For that is enough. +10000ms

[superhil]

For that is enough. +10000ms
i understand it, for that purpose +10000ms will work well.

i just want to point out that "-edit=0:00:0X,10000ms -silence" for X=0,1,..,10 should be a valid input. ;)

[madshi]

If you want the bug fixed you need to add it to the bug tracker. It's as simple as that.

[Groucho2004]

@madshi
Is there a chance you could expand the "normalize" functionality so one can specify the peak level (instead of always normalizing to 0dB)? I sometimes like to do some post processing and having some headroom would be nice.
Maybe something like "-normalize -3dB".

[superhil]

If you want the bug fixed you need to add it to the bug tracker. It's as simple as that.

Maybe my explanation is not good, but as the developer i'm sure that you understand it very well :D

http://bugs.madshi.net/view.php?id=61

[superhil]

@madshi
Is there a chance you could expand the "normalize" functionality so one can specify the peak level (instead of always normalizing to 0dB)? I sometimes like to do some post processing and having some headroom would be nice.
Maybe something like "-normalize -3dB".

Is it similar to add gain after normalize?
Normalize to 0dB means that you add some positive or negative gain so that the peak become at 0dB.
If you want the peak at +2dB you can simply normalize and add +2dB gain

[tebasuna51]

If you want the peak at +2dB...

You can't have peaks at +2dB with the standar output 24 bit int.
The max value is 0 dB.

[superhil]

You can't have peaks at +2dB with the standar output 24 bit int.
The max value is 0 dB.
I'm sorry, I don't know about that :D
But, couple days ago i converted .ac3 audio to wav with eac3to. In logfile i saw that it applies -XdB gain (i don't remember exact value). Does it mean original audio has a positive peak value (+X dB)?

[Groucho2004]

Is it similar to add gain after normalize?More or less. I just want to skip the additional step of adding gain.

If you want the peak at +2dB
As already pointed out by tebasuna51 - :confused:

[Groucho2004]

I'm sorry, I don't know about that :D
But, couple days ago i converted .ac3 audio to wav with eac3to. In logfile i saw that it applies -XdB gain (i don't remember exact value). Does it mean original audio has a positive peak value (+X dB)?
The maximum level for PCM data is 0 dB, no matter what bit depth.

[DarkSpace]

I'm sorry, I don't know about that :D
But, couple days ago i converted .ac3 audio to wav with eac3to. In logfile i saw that it applies -XdB gain (i don't remember exact value). Does it mean original audio has a positive peak value (+X dB)?

The maximum level for PCM data is 0 dB, no matter what bit depth.

I think that with floating point samples, it is possible to have peaks above 0 dB. Also, since eac3to internally decodes ac3 as 64-bit floating point, this is quite likely what happened, and eac3to lowered the volume to prevent clipping.

[Groucho2004]

I think that with floating point samples, it is possible to have peaks above 0 dB.
No, it's not.

Edit:
Some light reading about the subject. (http://wiki.multimedia.cx/index.php?title=PCM)

[sephirotic]

@tebasuna51

BTW, NeroAacEnc can encode 6.1. Try this:

eac3to input6.1.dts stdout.wav | NeroAacEnc -q 0.5 -ignorelength -if - -of output.m4a

The problem here is the order in back/rear channels. Maybe your AAC decoder can't restore the correct mapping.

The NeroAacEnc options are:

Quality/bitrate control:
-q <number> : Enables "target quality" mode.
<number> is a floating-point number in 0...1 range.
-br <number> : Specifies "target bitrate" mode.
<number> is target bitrate in bits per second.
-cbr <number> : Specifies "target bitrate (streaming)" mode.
<number> is target bitrate in bits per second.
When neither of above quality/bitrate options is used,
the encoder defaults to equivalent of -q 0.5

Multipass encoding:
-2pass : Enables two-pass encoding mode.
Note that two-pass more requires a physical file as input,
rather than stdin.

To have a efficient -br encode you need -2pass and need a physical file. eac3to can't do this, decode your input to wav and use NeroAacEnc with the -2pass option.

Thank you very much for your answers. The eac3 unnoficial-faq neither the first page of this thread had a full list of options for neroaac parameters so this post was very useful. May i suggest adding a -h into the eac3to with all available commands in a future update? (is there another argument for showing help built in with the eac3to.exe x86 build? sorry if this was a dumb question)

Anw, thanks again.

[Snowknight26]

Why should another program's parameters be listed by eac3to?

[LigH]

IEEE floating point samples (single or double precision, MS codec 2CC = 0003h) would be able to store denormalized sample values beyond ±1.0; still, it should be avoided, the output from sound cards will usually be based on integer sample values, the D/A converter won't produce a higher voltage. But it may be useful for storing raw footage to be finalized.

There are several discussions if a floating point PCM audio with samples beyond 0 dBFS is "valid"; some say it was a purpose to use this format for cases where exceeding results are probable. I certainly share this opinion (http://forum.cockos.com/showpost.php?p=252083&postcount=8):

There's nothing "mad" about having 32-bit floating point files that go over the -1.0 - 1.0 range in the intermediate stages of audio production...But I do agree anything to be listened as the final product should be as standard as possible.

[madshi]

You can use the "-full" eac3to parameter to force eac3to to create a 64bit floating point file which can then store peaks above -1.0 - 1.0. When you're done with (external) processing, another eac3to run without "-full" will finally produce a 24bit file with no clipping. However, not many tools understand 64bit floating point wav files. Maybe the "-float32" switch (instead of "-full") has better compatability with a wider range of external tools, but it results in a small loss of precision, naturally.

[tebasuna51]

I think that with floating point samples, it is possible to have peaks above 0 dB. Also, since eac3to internally decodes ac3 as 64-bit floating point, this is quite likely what happened, and eac3to lowered the volume to prevent clipping.
Yes, is correct.
But take in mind than any output above 0 dB is always a error in the lossy process of encode/decode, because the original source never can have peaks above 0 dB.

This problem never occurs decoding lossless formats.

The eac3 unnoficial-faq neither the first page of this thread had a full list of options for neroaac parameters so this post was very useful. May i suggest adding a -h into the eac3to with all available commands in a future update? (is there another argument for showing help built in with the eac3to.exe x86 build? sorry if this was a dumb question)

eac3to don't support all parameters of internal encoders (Aften, NeroAacEnc and Flac) then can't be showed like eac3to parameters.

But you can use the 'pipe' way to use the full parameters of external encoders Aften.exe, NeroAacEnc.exe, Flac.exe and others:

eac3to input stdout.wav <eac3to parameters> | Ext_Encoder <Ext_encoder parameters>

You can know all parameters of external encoders with:
Aften -longhelp
NeroAacEnc -help
Flac --help
...

[wiggaz]

Hi,
I'd like to know how to cut an audio at a specific time and keep the first part of the cut.
I searched a lot, but I didn't found anything.
Thanks in advance.