UsEac3to v1.3.6: a different GUI for eac3to

[oniiz86]

@tebasuna51 I was just curious as to why there is a difference concerning H264 & H265 streams when selecting the Output Format, the H264 stream has the option to output the stream into the MKV container or as an H264 Elementary Stream but with a HEVC/H265 stream it only offers to output as an H265 Elementary Stream, I may be overlooking something so simple. :)

https://i.imgur.com/gYvKLEc.jpeg

https://i.imgur.com/EDJ8ElW.png

[tebasuna51]

Because gdsmux (part of old Haali Media Splitter used by eac3to) does not support h265 (https://haali.net/mkv/codecs.pdf).

[oniiz86]

Because gdsmux (part of old Haali Media Splitter used by eac3to) does not support h265 (https://haali.net/mkv/codecs.pdf).
@tebasuna51 Thanks for that, I find it surprising though that UsEac3to will allow a H265/HEVC stream from an MKV file to be extracted as a H265 Elementary Stream.

https://i.imgur.com/ORikrkJ.png

If I import the H265 Elementary Stream into UsEac3to, the Output Format options are only H264 or MKV.

https://i.imgur.com/nKXKUUj.png

I know you said it's a gdsmux limitation with the haali media splitter but I was surprised that it can be extracted fine from the MKV container as H265.

[Columbo]

Why do you find it surprising? Muxing and demuxing are different things.

[oniiz86]

I understand they're different things but was curious why you can extract a HEVC stream from an MKV file but if a H265/HEVC stream is imported into UsEac3to it has H264 or MKV options instead for a H265/HEVC stream, I get it because @tebasuna51 mentioned it's a gdsmux limitation with the haali media splitter not offering H265 support but it appears you can still get around the issue if a MKV file is input.

[tebasuna51]

If I import the H265 Elementary Stream into UsEac3to, the Output Format options are only H264 or MKV.

Of course eac3to can't recode h265 to h264 and show:
This video conversion is not supported. <ERROR>

Also can't mux h265 to mkv and show a error:
h265/HEVC video track muxing to MKV is currently not supported. <ERROR>

It's a clear bug in UsEac3to when it offers those options (I'm only looking at the first three characters, 'h26' for h264).

I'll fix it in the future; it's obvious that eac3to can't do anything with an h265 track.

[DanDare1983]

Hey teba, whilst trying to convert a DTS-HD MA 2.0 track to Flac i encountered an error/crash which was something yo do with flac first pass. I'm using the latest eac3to version 3.57 with the latest UsEac3to. Also I have quite alot of atmo tracks/7.1 tracks and would it be ok to simply encode to ac3? Would i have to use the -down6 option or can I just leave it as just ac3?

[tebasuna51]

Hey teba, whilst trying to convert a DTS-HD MA 2.0 track to Flac i encountered an error/crash which was something yo do with flac first pass. I'm using the latest eac3to version 3.57 with the latest UsEac3to.
No problem here:
eac3to v3.57
DTS Master Audio, 2.0 channels, 16 bits, 48kHz, dialnorm: 0dB
(core: DTS, 2.0 channels, 1509kbps, 48kHz, dialnorm: 0dB)
Decoding with libDcaDec DTS Decoder...
Encoding FLAC with libFlac...
Creating file "D:\Test\AudioN\DtsHD\20\stream2MA.dts_.flac"...
The original audio track has a constant bit depth of 16 bits.
eac3to processing took 1 second.
Done.
Please upload the eac3to log or a sample of your DTSMA

[EDIT]Try also ffmpeg recode with UsEac3to option 'A/V Recode' selecting 'flac' in 'Decode, Lossless,...' options.

Also I have quite alot of atmo tracks/7.1 tracks and would it be ok to simply encode to ac3? Would i have to use the -down6 option or can I just leave it as just ac3?
Never use the old Aften encoder included with eac3to, use always ffmpeg like ac3 encoder. You have 3 options:

1) Decode with old eac3to libav decoder and downmix with eac3to -down6 using the ac3-ffm Output format, like this:
-down6 stdout.w64 | ffmpeg -i - -c:a ac3 -b:a ...

2) Recode with ffmpeg automatic downmix with UsEac3to option 'A/V Recode' selecting only the bitrate.

3) Like 2) but do the recommended downmix 71-51o like this:

[Wile_E_Coyote]

That EAC3 5.1 from streaming don't have AC3 core, if your player don't support Atmos only the simple EAC3 (core) is played.

If you want recode it to AC3 to save space, and lose quality, you can use ffmpeg with the button 'A/V Recode' in UsEac3to.

I always recommend remove the Dialog Normalization and disable the Dynamic Range Compression, but everyone can do it or not.

Hi tebasuna51. Is it possible to "downmix" the added Atmos/JOC into a DD+/E-AC3 stream rather than convert to AC3 or does that not gain anything?

[tebasuna51]

Like you say that not gain anything.

[seagate]

tebasuna51, hello

Today I downloaded Hawaii Five-0 - Brian Tyler 2023 OST in flac files.

Only a few out of 41 flac files could not be converted to ac3.

I am using UsEac3to v.1.3.5, eac3to v.3.57. I am using Windows 10 22H2 (19045.2728) education.

I am attaching a picture so you can see the error.

"I DonТt Ride Shotgun.flac" not found. <ERROR>"

Specifically this file (flac) is not recognized, and most of the 41 files are successfully recognized by UsEac3to v.1.3.5.

I am also attaching only one of the several (5) problematic flac files for you to test and/or determine why there is a problem.

If you need it - I will also upload a flac file that was successfully recognized by UsEac3to v.1.3.5 and I converted it to AC3 without any problems.

https://i.postimg.cc/sgC5fD6v/haa.png

39 I Don’t Ride Shotgun.flac (https://seyarabata.com/695420bcb0e66)

Complete name : 39 I Don’t Ride Shotgun.flac
Format : FLAC
Format/Info : Free Lossless Audio Codec
File size : 11,9 MB
Duration : 55 s 693 ms
Overall bit rate mode : Variable
Overall bit rate : 1 793 kb/s
Album : Hawaii Five-0 (Original Series Soundtrack)
Album/Performer : Brian Tyler
Part : 1
Part/Total : 1
Track name : I Don’t Ride Shotgun
Track name/Position : 39
Track name/Total : 41
Performer : Brian Tyler
Genre : Score
Recorded date : 2023
ISRC : USLS52228539
Copyright : (C) 2023 Lakeshore Records (P) 2023 Lakeshore Records
Cover : Yes
Cover type : Cover (front)
Cover MIME : image/jpeg
UPC : 0780163628524
MEDIATYPE : album
ITUNESADVISORY : 0

Audio
Format : FLAC
Format/Info : Free Lossless Audio Codec
Duration : 55 s 693 ms
Bit rate mode : Variable
Bit rate : 1 549 kb/s
Channel(s) : 2 channels
Channel layout : L R
Sampling rate : 48.0 kHz
Bit depth : 24 bits
Compression mode : Lossless
Stream size : 10.3 MiB (86%)
Writing library : libFLAC 1.1.0 (2003-01-26)
MD5 of the unencoded conten : 00000000000000000000000000000000


Addendum:

Oh my god. I figured out why this problem was occurring.

The problem was due to the apostrophe in the name: "Don’t".

Can’t this be corrected somehow in the software so that it recognizes correctly?!

[Columbo]

if that is an issue with eac3to itself I can look at it. The ' character is legal in a filename.

[tebasuna51]

The apostrophe (') character ascii code 39 is legal without problem, but in your name "39 I DonÆt Ride Shotgun.flac" there are the ascii char 149 show in my codepage like other than ('), in your image is showed like T.

Edit the name and replace that character with the (') in your keyboard (ascii code 39).

[SeeMoreDigital]

There is a world of difference between don’t and don't...

[tebasuna51]

Today I downloaded ...

And remember rule:

6) No warez, cracks, serials or illegally obtained copyrighted content! Links to content of a questionable nature (e.g. anything you don't own and/or have downloaded), asking for, offering, or asking for help/helping to process such content in any way or form is not tolerated.

This problem hapen with names created in other countries or OS, not with names created by yourself.

[tebasuna51]

Hi
Is there a way to output RF64 (WAV) files with eac3to?
I can change the audio FPS without any problem, but I can’t find an option to force RF64 as the output format.
It looks like W64 is the only similar format available.
I’m currently automating the process with a batch script using eac3to + Dolby Encode Engine.

This is a private message received; please use the standard forum for technical questions like this.

In the help you have the answer, use .rf64 like extension.

In my next new GUI version 1.3.6 I offer that output and other new parameters.

[tebasuna51]

New version at first post with minor changes

Changelog
=========
- v1.3.6 2026-01-03 Add some options (-forced,...)

[oniiz86]

Thanks so very much for releasing v1.3.6 so swiftly Teba, a wonderful start to the New Year, it's so greatly appreciated, a very Happy New Year to you my good sir!! :thanks:

[sadpunk]

Thank you @tebasuna51

I’m changing the FPS of AAC audio files.
When decoding AAC, eac3to uses the Nero Audio Decoder. Would it make more sense to use FFmpeg instead of Nero, or is it sufficient to rely only on eac3to?

Currently, I decode AAC with FFmpeg using pcm_s32le and -rf64 always, then I change the FPS of the resulting RF64 audio file and finally convert it to AC-3 or E-AC3 using Dolby Encode Engine (DEE).

I’m using this workflow for AC-3 as well, since I don’t want to use libaften for AC-3 encoding and therefore prefer encoding AC-3 via DEE.

[tebasuna51]

eac3to decode aac with the old Nero7 decoder if it is installed.
I use always ffmpeg, also like ac3/eac3 encoder, then eac3to is not used at all.

If you want use DEE like encoder can do the first part with UsEac3to (select the fps change in the atempo parameter) like I show in the image:

[DanDare1983]

Hey teba, just a quick question. Is .w64 audio the same as .wav? I've seen some people converting .w64 audio to .wav and I wonder why? If I was to make a remux would it be ok to just leave the PCM audio as .w64?

[tebasuna51]

The audio data is the same, but the w64 header (also the rf64 format) have fields to support filesize greater than 4 GB (multichannel audio movie tracks needs that sizes).
It is safe convert w64 to wav if < 4 GB.
Some software can support also wavs > 4 GB with fake header but it is not recommended.

[IntelHEVC]

“The Dolby Vision data is being interpreted incorrectly. For example, DV 7.6 is physically present, but the GUI always shows it as DV 8.”

[tebasuna51]

The GUI only show the eac3to.exe output, seems it is a problem to be reported at https://www.rationalqm.us

[Columbo]

Or you can provide a source sample here and tell how you know it is 7.1 and DG Tools will look into it. Aternatively, provide a link to purchase that exact disk together with telling how you know it is 7.1.

[Columbo]

Please test this:

https://rationalqm.us/misc/eac3to.exe

Without a sample I can't prove the fix so your help will be appreciated.

[Kyaneos]

I hadn't used UsEac3to for a while, and after installing the latest version, I see that the `keep dialnorm` option is enabled by default. In previous versions, if I remember correctly, the `remove dialnorm` option was enabled by default.

I understood that dialnorm was obsolete and that it was recommended to remove it from Dolby tracks. Has the recommendation changed? Is it now recommended to keep it?

Thank you very much.

[tebasuna51]

It is optional, by default eac3to.exe remove dialnorm but in last versions there are a eac3to.ini with:

-fast
-keepDialnorm
-progressnumbers

and change the behaviour.
You can edit that eac3to.ini and remove the line -keepDialnorm like I have.

[tebasuna51]

There is a test version v.3.62 (https://www.rationalqm.us/eac3to/) of eac3to that allows AAC decoding using a limited version of ffmpeg.exe included in the package.

It works without apparent problems in UsEac3toGUI 1.3.6.

It does not yet support .m4a files, although ffmpeg should handle it without issue.

[oniiz86]

@tebasuna51 I don't believe it is working with UsEac3to v1.3.6 as unless I'm misunderstanding how it works, can you please show how aac decoding works with the latest eac3to v3.62 test build with ffmpeg invocation, the use of Nero no longer seems possible according to this post https://rationalqm.us/board/viewtopic.php?p=23820#p23820

I can get 7.1 EAC3 decoding to 7.1 WAV using the Output Format "wav" & "wavs" with the latest v3.62 test build & UsEac3to has no issues at all there but I'm stumped with how it handles AAC decoding.

[Columbo]

@tebasuna51 I don't believe it is working with UsEac3to v1.3.6 as unless I'm misunderstanding how it works, can you please show how aac decoding works with the latest eac3to v3.62 test build with ffmpeg invocation, the use of Nero no longer seems possible according to this post https://rationalqm.us/board/viewtopic.php?p=23820#p23820
AAC decoding works fine in the 3.62 test build. What issue are you having specifically? If you insist on Nero and you have a working Nero install, you can remove ffmpeg.exe.

[tebasuna51]

Without problem with my GUI but the eac3to support is still limited, works this simple decode to wav (or wavs):

eac3to v3.62
command line: "C:\Portable\eac3to\eac3to.exe" "D:\Test\AudioD\Samples\aac\entrada.aac" "D:\Test\AudioD\Samples\aac\entrada.aac_-.wav" -progressnumbers -log="C:\Portable\eac3to\UsEac3to\UsEac3To.log"
------------------------------------------------------------------------------
Fast mode enabled
Dialnorm removal enabled
AAC, 5.1 channels, 24kHz, dialnorm: 0dB
Decoding AAC via ffmpeg...
eac3to processing took 1 second.
Done.

But not this recode to flac (or w64, ac3, pcm, rf64, etc.):
eac3to v3.62
command line: "C:\Portable\eac3to\eac3to.exe" "D:\Test\AudioD\Samples\aac\entrada.aac" "D:\Test\AudioD\Samples\aac\entrada.aac_-.flac" -progressnumbers -log="C:\Portable\eac3to\UsEac3to\UsEac3To.log"
------------------------------------------------------------------------------
Fast mode enabled
Dialnorm removal enabled
AAC, 5.1 channels, 24kHz, dialnorm: 0dB
Decoding with DirectShow (Nero Audio Decoder 2)...
Getting "Nero Audio Decoder 2" instance failed. <ERROR>
Aborted at file position 147018. <ERROR>

Also not all .aac are supported: mono, 7.1 or in m4a container

EDIT: also functions like slowdown, edit, ... are ignored

[Columbo]

Currently, ffmpeg is invoked only for decoding to wav/wavs. For all other scenarios the previous functionality with Nero/Sonic is invoked. You need a functioning Nero/Sonic install for that. In the future transcoding scenarios may be supported. The ffmpeg decoding was added to at least allow decoding to wav/wavs as few people have a functioning Nero/Sonic setup these days. We'll add a note about this to the 3.62 notes file.

If you can provide samples for the unsupported files we can look into supporting them.

[oniiz86]

@tebasuna51 Thanks so much for that, I'm sorry you're right, I'm a complete idiot for some silly reason I was confusing encoding with decoding, AAC decoding does work fine except for this AAC LC SBR test file that eac3to is reporting as 22kHz when it should be 44.1kHz found here https://www2.iis.fraunhofer.de/AAC/multichannel.html, eac3to has no trouble with the ffmpeg decoding the AAC LC variant at the link & reports correctly as 44.1kHz.

[oniiz86]

Currently, ffmpeg is invoked only for decoding to wav/wavs. For all other scenarios the previous functionality with Nero/Sonic is invoked. You need a functioning Nero/Sonic install for that. In the future transcoding scenarios may be supported. The ffmpeg decoding was added to at least allow decoding to wav/wavs as few people have a functioning Nero/Sonic setup these days. We'll add a note about this to the 3.62 notes file.

If you can provide samples for the unsupported files we can look into supporting them.

@Columbo Thanks for that, I have an issue with the ffmpeg decoding to WAV with this AAC LC SBR test file that eac3to reports as 22kHz https://transfer.it/t/PoelYd0RWDbZ

eac3to v3.62
command line: "C:\Users\MASTER\Desktop\UsEac3to136\eac3to.exe" "C:\Users\MASTER\Desktop\ChID-BLITS-EBU-Narration.mka" 1: "C:\Users\MASTER\Desktop\UsEac3to136\ChID-BLITS-EBU-Narration.mka_1-eng.wav" -progressnumbers -log="C:\Users\MASTER\Desktop\UsEac3to136\UsEac3To.log"
------------------------------------------------------------------------------
Running in normal mode
Dialnorm removal enabled
MKA, 1 audio track, 0:00:47
1: AAC, 5.1 channels, 22kHz, dialnorm: 0dB
[a01] Extracting audio track number 1...
[a01] Decoding AAC via ffmpeg...
[a01] ffmpeg failed decoding AAC. <ERROR>
Aborted at file position 942754. <ERROR>

[Columbo]

Thank you for the report, oniiz86.

We were using fdkaac but as you have shown it turns out to be problematic. fdkaac as a decoder is poorly supported in ffmpeg. It was primarily designed as an encoder. For HE-AAC/SBR streams specifically, it doesn't properly set the output channel layout on decoded frames, causing the filter graph to fail. The native ffmpeg AAC decoder has no such limitation. For decoding, ffmpeg's native AAC decoder is fine. It handles LC, HE-AAC, HE-AAC v2, and SBR correctly. For decoding, the output is PCM either way and the difference is imperceptible.

Therefore, we've reverted the use of fdkaac. Please re-download the test build and update eac3to.exe. Report your results.

It's still reporting 22kHz. Classic HE-AAC/SBR issue. The ADTS header encodes the base layer sample rate (22050Hz). The SBR extension doubles it to 44100Hz during decoding. eac3to reads the rate from the ADTS header so it reports 22kHz. MediaInfo knows about SBR and reports the actual decoded rate. For the ffmpeg path this doesn't affect the output as ffmpeg decodes to the full 44.1kHz and the WAV(s) will be correct. It's purely a display issue in eac3to's reported format info. Fixing it properly would require parsing the SBR signaling in the AudioSpecificConfig or ADTS extension elements, which is non-trivial.

The ffmpeg trick is a hack to give at least correct decoding. We'll have to live with its limitations and document them well.

To avoid cluttering this thread we'd prefer further issues be addressed at DG forum.

[oniiz86]

@Columbo Thanks for that, I confirmed all is good now with the AAC LC SBR test file over at the DG forums & will address further issues there.

eac3to v3.62
command line: "C:\Users\MASTER\Desktop\UsEac3to136\eac3to.exe" "C:\Users\MASTER\Desktop\ChID-BLITS-EBU-Narration.mka" 1: "C:\Users\MASTER\Desktop\UsEac3to136\ChID-BLITS-EBU-Narration.mka_1-eng.wav" -progressnumbers -log="C:\Users\MASTER\Desktop\UsEac3to136\UsEac3To.log"
------------------------------------------------------------------------------
Running in normal mode
Dialnorm removal enabled
MKA, 1 audio track, 0:00:47
1: AAC, 5.1 channels, 22kHz, dialnorm: 0dB
[a01] Extracting audio track number 1...
[a01] Decoding AAC via ffmpeg...
eac3to processing took 1 second.
Done.

[tebasuna51]

We were using fdkaac but as you have shown it turns out to be problematic. fdkaac as a decoder is poorly supported in ffmpeg. It was primarily designed as an encoder...
The native ffmpeg AAC decoder has no such limitation. For decoding, ffmpeg's native AAC decoder is fine...

Therefore, we've reverted the use of fdkaac. Please re-download the test build and update eac3to.exe. Report your results.

1) Of course the native ffmpeg decoder is better and without problems.
Only the default encoder is worse than qaac or fdkaac.

2) There are a legal problem: you can't include in your pack a ffmpeg binary with:
--enable-libfdk-aac

3) Other question is how eac3to recognize or decode the aac streams.
In the attached logs there are some issues.

To avoid cluttering this thread we'd prefer further issues be addressed at DG forum.

Yes, UsEac3to users don't need, by the moment, that new eac3to feature because the A/V Recode functions work better, like the logs attached show.

BTW feel free to ask here some questions because I can't use the DG forum like you know.

[tormento]

Well, there is always a trick: point to https://github.com/MartinEesmaa/FFmpeg-Builds ;)

[Columbo]

Exactly. It's only a patent issue. Many of the patents have run out and the rest will be soon. Fraunhofer has never acted against any freeware/open source projects and there are many nonfree builds available for download across the internet.

eac3to's AAC parsing for format detection/reporting is primitive and we have no plans to rewrite it.

Anyone having issues with our site should contact our admin.

[tormento]

Anyone having issues with our site should contact our admin.
If only he would reply to e-mails.

If I recall correctly, I sent twice some years ago and I am still waiting for a reply.