UsEac3to v1.3.6: a different GUI for eac3to

[Hellboy.]

tebasuna51
When i use UsEac3To to convert a DTS-HD 5.1 to ac3 5.1 using ffmpeg-A/V Recode, the command line use -center_mixlev 0.707.
What is the benefit of using this? Because the ffmpeg documentation say that is used "when downmixing to stereo".
Thanks.

[tebasuna51]

That parameter, included in the BSI info of each AC3 frame, is only used when it is downmixing automatically to stereo, yes.

Like the default used by ffmpeg AC3 encoder is -center_mixlev 0.500 (I don't know for what, in old Aften encoder was 0.707) I recommend put it to 0.707 to avoid the know problem of low volume dialogs when downmix.

Of course playing it at 5.1 do nothing.

[hypnotize666]

Hi everyone, I would like to know the step to follow to have an aac type file at the output He is normalized so I imagine that it must be cvbr (aac lc sbr) I tried with this type of parameter but without success.Thanks
http:////imgur.com/Y5qGdNM
http://imgur.com/yelyyjJ
Does the info of the quality 5.1 then stereo are good if not can we have more info to understand how to configure it plz?thx

[tebasuna51]

The CLP you use is:

stdout.wav -normalize | qaac -v 99 --he --ignorelength --adts --no-delay -o %_.aac -

output a error:
--num-priming is only applicable for AAC LC.

In the qaac help show:
Applicable only for AAC LC.
--num-priming=0 is the same as --no-delay.
Without the --no-delay parameter the output work fine:
AAC-HE Encoder, CVBR 96kbps,...

Using qaac --formats you can see the allowed bitrates for HE CVBR 48KHz 5.1 sources:
...
HE 48000Hz 5.1 (C L R Ls Rs LFE) -- 80,96,112,128,160,192
...
If you put 99 take 96 Kb/s

Without the --no-delay I recommend output in mp4 container:

stdout.wav -normalize | qaac -v 96 --he --ignorelength -o %_.m4a -

[Emulgator]

Like the default used by ffmpeg AC3 encoder is -center_mixlev 0.500
(I don't know for what, in old Aften encoder was 0.707)
I recommend put it to 0.707 to avoid the know problem of low volume dialogs when downmix.
Just as I came across audio 2.0 tracks consisting of the original mono copied to L and to R, so phase correlation = 1:
If those in-phase tracks would happen to reach 0dB (what they rarely do) -center_mixlev 0.500 would be only safe way to stay below clipping.
-center_mixlev 0.707 would allow 1.41 FS for that corner case, so I guess ffmpeg just wanted to stay away from clipping.
If user could rely that each such fully correlated track would never exceed -6dB then -center_mixlev 0.500 would be safe.
In the end at my place it would boil down to inspect the waveform first, then decide how to continue.

[Hellboy.]

I know that to preserve the Atmos data you must extract it with -keepDialnorm (eac3to).
But if i have an Atmos that i don't know if -keepDialnorm was used, how i can check?
Thanks.

[tebasuna51]

...-center_mixlev 0.707 would allow 1.41 FS for that corner case, so I guess ffmpeg just wanted to stay away from clipping.
Well, I prefer some clip than don't listen the dialogs, with the 0.707 coeficient the volume of center channel is preserved, with low coeficients means attenuate the dialogs.
In the end at my place it would boil down to inspect the waveform first, then decide how to continue.
Of course the best option when that audio must be played with a stereo speakers is make a proper downmix, even with greater coeficient for center channel, and Normalize it like is explained here (https://forum.doom9.org/showthread.php?p=2003223#post2003223)

[Emulgator]

Yep, normalize before is a good solution.

[tebasuna51]

I know that to preserve the Atmos data you must extract it with -keepDialnorm (eac3to).
The data are preserved (same size and MediaInfo show both like JOC) but with Dialnorm changed (if show -31 dB maybe it is changed) seems not compliant.

Extracted with -keepDialnorm my Denon player show DD Atmos and decoded with Dolby Reference Player seems fine.

Without -keepDialnorm (or converted to mkv from m2ts with MkvToolnix) my Denon player show DD+ and decoded with Dolby Reference Player we obtain only silence.

Like I say MediaInfo show JOC for both and Cavernize decode both with the same output (seems more permissive).

[hypnotize666]

Hi everyone, I would like to know the step to follow to have an aac type file at the output He is normalized so I imagine that it must be cvbr (aac lc sbr) I tried with this type of parameter but without success.Thanks
http:////imgur.com/Y5qGdNM
http://imgur.com/yelyyjJ
Does the info of the quality 5.1 then stereo are good if not can we have more info to understand how to configure it plz?thx
Hello again, here is my structure of my folder with the codecs I have chosen the qaac codec but why in the log does it indicate a nero direct show codec error?This is the log file:eac3to v3.36
command line: "C:\Installation\Encodeurs\eac3to336-UsEac3to133\eac3to.exe" "K:\00-Classics\Teleseries\MED (2015) {tvdb-304283}\s01\MED s01e01 French Hdtv x265 Hypnotize.mkv" 2: stdout.wav -normalize -progressnumbers -log="K:\00-Classics\Teleseries\MED (2015) {tvdb-304283}\s01\MED s01e01 French Hdtv x265 Hypnotize.mkv_Job_1.log"
------------------------------------------------------------------------------
MKV, 1 video track, 1 audio track, 0:22:22, 30p /1.001
1: h265/HEVC, 720p30 /1.001 (16:9), 10 bits
2: AAC, French, 5.1 channels, 24kHz, 40ms
"Francais 5.1 AAC@192Kbps HE"
[a02] Extracting audio track number 2...
[a02] Decoding with DirectShow (Nero Audio Decoder 2)...
[a02] Getting "Nero Audio Decoder 2" instance failed. <ERROR>
Aborted at file position 262144. <ERROR>
http://imgur.com/h4hmNKg
http://imgur.com/Z38eAp8

[hypnotize666]

I found it a bit fishy but I still tried with this command: stdout.wav -normalize | qaac -v 192 --he --ignorelength -o %_.m4a - but the strange thing is that by clicking on the menu at the end it does not mark a .m4a type file but rather an .aac type file as I wasn't sure I tried 2 ways but without success 2: stdout.wav -normalize | qaac -v 192 --ignorelength -o %_2fre.aac -

[tebasuna51]

... but why in the log does it indicate a nero direct show codec error?

Because eac3to need Nero 7 installed to decode AAC, NeroAacDec is not enough. Decode it before to wav with ffmpeg (A/V Recode) and after normalize and recode to AAC.

[tebasuna51]

I found it a bit fishy but I still tried with this command: stdout.wav -normalize | qaac -v 192 --he --ignorelength -o %_.m4a - but the strange thing is that by clicking on the menu at the end it does not mark a .m4a type file but rather an .aac type file as I wasn't sure I tried 2 ways but without success 2: stdout.wav -normalize | qaac -v 192 --ignorelength -o %_2fre.aac -

What menu?
Without success?

Both commands works fine for me (of course over a source than can be decoded by eac3to), the first create a file_.m4a (aac HE in mp4 container) and the second a file_2fre.aac LC

[mannequin80]

Well, I prefer some clip than don't listen the dialogs, with the 0.707 coeficient the volume of center channel is preserved, with low coeficients means attenuate the dialogs.


is there a universal way to enhance the dialogue (center channel) without applying additional filters (like dynaudnorm)?

[tebasuna51]

@mannequin80
Read the thread convert 5.1 to stereo (https://forum.doom9.org/showthread.php?t=185520)

Players can be instructed to enhance the FC in the downmix, for instance with mpc-hc:

[mannequin80]

@mannequin80
Read the thread convert 5.1 to stereo (https://forum.doom9.org/showthread.php?t=185520)


i'd love to keep it at 5.1 and make it playable across various devices with good sounding dialogue while maintaining the dynamic range as much possible and not using any hardware methods like DRC or Night Mode. after looking at multiple methods, i can't find a better solution than "volume=x.xdB (if needed), dynaudnorm=p=0.9:m=2:g=31:f=300". there's no easily detectable volume pumping even in complicated scenes BUT the drawback of this method is the loud parts (like music, etc) become even louder so it's difficult to preserve the balance. wondering about any additional tweaks to try.

[DanDare1983]

Do you have to -keepdialnorm when extracting atmos? I remember using the old eac3to and it would automatically remove dialnorm as it was supposed to help in some way. I'm using -keepdialnorm from now on as I think it's the best thing to do especially as dgdemux and makemkv keeps dialnorm too.

[tebasuna51]

Do you have to -keepdialnorm when extracting atmos? I remember using the old eac3to and it would automatically remove dialnorm as it was supposed to help in some way. I'm using -keepdialnorm from now on as I think it's the best thing to do especially as dgdemux and makemkv keeps dialnorm too.

Just a few days before:

The data are preserved (same size and MediaInfo show both like JOC) but with Dialnorm changed (if show -31 dB maybe it is changed) seems not compliant.

Extracted with -keepDialnorm my Denon player show DD Atmos and decoded with Dolby Reference Player seems fine.

Without -keepDialnorm (or converted to mkv from m2ts with MkvToolnix) my Denon player show DD+ and decoded with Dolby Reference Player we obtain only silence.

Like I say MediaInfo show JOC for both and Cavernize decode both with the same output (seems more permissive).

[spinkeln]

Deleted

[MrVideo]

Do you have to -keepdialnorm when extracting atmos?
Using eac3to v3.52, I've discovered that using -keepdialnorm wasn't necessary as eac3to indicates, when it starts to operate on the source, that it is "Keeping dialnorm."

The job that was doing was extracting the thd and ac3 streams from a mkv file so that I can THD merge the two.

[MrVideo]

Is it possible with (Us)Eac3to to create an eac3 file with an empty core, like what can be done with a THD stream within a MKV file?

[tebasuna51]

Using eac3to v3.52, I've discovered that using -keepdialnorm wasn't necessary as eac3to indicates, when it starts to operate on the source, that it is "Keeping dialnorm."
It is not necesary if you have the -keepdialnorm in the eac3to.ini file.

Is it possible with (Us)Eac3to to create an eac3 file with an empty core, like what can be done with a THD stream within a MKV file?

UsEac3to can create eac3 files using ffmpeg (only to 5.1 and without ac3 core, only eac3 frames) but I can't understand the relation with THD.

To create a THD track BD compliant need a ac3 file, not eac3.
To create a eac3 track BD compliant you can't use ffmpeg (maybe Dolby Encoder Engine).

[MrVideo]

It is not necesary if you have the -keepdialnorm in the eac3to.ini file.
It is there, but I never put it there. Must be part if the distro.
UsEac3to can create eac3 files using ffmpeg (only to 5.1 and without ac3 core, only eac3 frames) but I can't understand the relation with THD.

THD was only mentioned as an example.

[tebasuna51]

Must be part if the distro.
Yes, it is. I don't like that option and I delete that option from the .ini file.

[MrVideo]

Yes, it is. I don't like that option and I delete that option from the .ini file.
Why don't you like the option?

[tebasuna51]

The Dialog Normalization different than -31 dB force to compliant decoders to attenuate the E-AC3 tracks all the sound.
Thats is a problem to recode, or listen, compared with others codecs.

For instance, with old DVD's with AC3 and DTS tracks all the people prefer the DTS track because sound high, normally 4 dB high with standard DN at -27 dB, than the same AC3 track.
The DTS encoder is not better, at same bitrate, than the AC3 only it is decoded without attenuation.
The same source encoded with GUIDE: How To Properly Encode Dolby Digital Audio (AC3) (https://forum.doom9.org/showthread.php?t=56020) recomendations sound with less volume than encoded with AAC, DTS,...

The same for the Dolby recomendation about Dynamic Range Compression, when we need recode a E-AC3 track we need force decoders (ffmpeg, avs decoders) to use drc=0.
When ffmpeg encode AC3 or EAC3 by default always use DN -31 dB and DRC=none to offer the same behaviour than others codecs.

If all the audio was encoded with Dolby defaults DN/DRC all can be ok, but if you listen Dolby compliants tracks and TV comercials or CD audio (Loudness war (https://en.wikipedia.org/wiki/Loudness_war)) or the same track encoded with other codec, you need modify the volume button every time.

[MrVideo]

Thanks for the info. I certainly do not want DRC. I'll remove the keepdialnorm option from the ini.

[SeeMoreDigital]

With regard to Dolby audio streams, is Dialnorm applied to all the available channels or just the centre channel?

[tebasuna51]

With regard to Dolby audio streams, is Dialnorm applied to all the available channels or just the centre channel?

To all. Just a part of GUIDE: How To Properly Encode Dolby Digital Audio (AC3) (https://forum.doom9.org/showthread.php?t=56020):
The decoder will perform an attenuation of (31 + dialnorm) dB to the program material when played back. So, in this case, the decoder will attenuate by (31 + -18) = 13 dB. This will bring the average sound level of the material to (-17.6 - 13) = -30.6 dBFS. The program is now played back at approximately -31 dBFS, the reference level.

-31 dBFS is a lower average volume level than what is typical from other sources. It will be noticeable that you will have to turn the volume up on your system when playing a DVD versus playing broadcast, tape, or other non-Dolby Digital program material.

[SeeMoreDigital]

To all. Just a part of GUIDE: How To Properly Encode Dolby Digital Audio (AC3) (https://forum.doom9.org/showthread.php?t=56020):
Thanks ;)

[tormento]

Does it work with last eac3to?

I have tried to read a mkv video and the eac3to about appears, instead of the track list.

Yes, I have set the exe paths.

[tebasuna51]

No problem here:

eac3to v3.52
command line: "C:\Portable\eac3to\eac3to.exe" "D:\Temp\t\00_HD.mkv" -progressnumbers -log="C:\Portable\eac3to\UsEac3to\UsEac3To.log"
------------------------------------------------------------------------------
Running in fast mode
Removing dialnorm
MKV, 1 video track, 1 audio track, 0:00:21, 25p
1: h264/AVC, English, 720p25
2: DTS Hi-Res, 7.1 channels, 2814kbps, 48kHz, dialnorm: 0dB
(core: DTS-ES, 5.1 channels, 1509kbps, 48kHz, dialnorm: 0dB)

[tormento]

No problem here
I did a clean installation of last version and no way to make it work, both drag and dropping and opening. Tried to rename file to a single character but nothing helped.

It opens a cmd window telling "analyze" and then the useac3to shows eac3to help page.

Would you please post here your eac3to.ini file?

Is there any way to trace the commands sent to eac3to? The log doesn't contain them.

[tebasuna51]

UsEac3to not need any installation other than fill the Settings options (if can't obtain them automatically):
Path to eac3to
Path to MkvMerge
Path to TsMuxer
Encoders Folder
Output Folder

You don't need eac3to.ini at all because the -progressnumbers -log= is added by UsEac3to (or can be added manually), but I have:
-fast
-progressnumbers

And the command line is show in the log like you can see in my previous post.

[tormento]

You don't need eac3to.ini at all because the -progressnumbers -log= is added by UsEac3to (or can be added manually), but I have
Ok, found the problem, I had '-nolog' in eac3to.ini.

[Axeldook]

Hi All,
when trying to encode to true hd. i lose channels.
5.1 come out 5.0
7.1 come out 5.0

how do i fix this ?

Thank you

[tebasuna51]

5.1 come out 5.0
Work fine here:
"C:\Portable\0\ffmpeg.exe" -i "C:\tmp\6p321.wav" -vn -strict -2 -acodec truehd "C:\tmp\6p321.wav_.thd"
ffmpeg version N-116720-g5c1c0325cd-20240817 Copyright (c) 2000-2024 the FFmpeg developers
Input #0, wav, from 'C:\tmp\6p321.wav':
Stream #0:0: Audio: pcm_s24le ([1][0][0][0] / 0x0001), 48000 Hz, 5.1(side), s32 (24 bit), 6912 kb/s
Output #0, truehd, to 'C:\tmp\6p321.wav_.thd':
Stream #0:0: Audio: truehd, 48000 Hz, 5.1(side), s32p (24 bit), 128 kb/s

7.1 come out 5.0
You are right here:
"C:\Portable\0\ffmpeg.exe" -i "C:\tmp\8w341.wav" -vn -strict -2 -acodec truehd "C:\tmp\8w341.wav_.thd"
Input #0, wav, from 'C:\tmp\8w341.wav':
Stream #0:0: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 48000 Hz, 7.1, s16, 6144 kb/s
Output #0, truehd, to 'C:\tmp\8w341.wav_.thd':
Stream #0:0: Audio: truehd, 48000 Hz, 5.0(side), s16p, 128 kb/s

The ffmpeg thd encoder only can encode to 5.1, then need downmix previously 7.1 -> 5.1 and the last versions have a bug (already reported but ignored (https://trac.ffmpeg.org/ticket/10971)) when do so.

To recode to thd 7.1 you need any commercial encoder.

[Axeldook]

thank you !

[DanDare1983]

I've come across a DTS-HD MA 5.1 track which has a 1000ms DELAY. The Log is telling me that a remaining delay of -2ms could not be fixed. I understand that -2ms is nothing to worry about but could someone explain how eac3to has got the delay down to -2ms? I've also seen an ATMOS track with a delay of 3003ms too. Also if i was to just convert the track to ac3 would it sort the problem? I'll also add that the video is 2: h264/AVC, 1080i50 (16:9).

[tebasuna51]

To fix delay without recompress eac3to add silent frames (or remove frames).
Only multiples of 10.666 ms (512 samples in each frame at 48 KHz) can be fixed exactly.

[DanDare1983]

To fix delay without recompress eac3to add silent frames (or remove frames).
Only multiples of 10.666 ms (512 samples in each frame at 48 KHz) can be fixed exactly.

Thanks for the explanation. So there's nothing to worry about and eac3to is doing its job?

[tebasuna51]

You preserve the quality (without decode/recode) and the remaining delay is always less than 6 ms., far than the duration of a video frame.

[DanDare1983]

You preserve the quality (without decode/recode) and the remaining delay is always less than 6 ms., far than the duration of a video frame.

I'm currently trying to convert DTS-HD 7.1 to AC3 5.1 640kbps. Can I just choose the ac3 option or do I have to enter -down6 too?

[tebasuna51]

You can use the automatic ffmpeg downmix, the -down6 eac3to method or the recommended downmix 71-51o using F-FFMPEG functions

[SeeMoreDigital]

I'm currently trying to convert DTS-HD 7.1 to AC3 5.1 640kbps. Can I just choose the ac3 option or do I have to enter -down6 too?
Out of interest... Is there any reason why you don't want to use the 5.1 DTS 'core'?

[tebasuna51]

The core is a lossy encoded, for what use it if you have the losless source?

[SeeMoreDigital]

The core is a lossy encoded, for what use it if you have the losless source?DanDare1983 is encoding from lossless 7.1 DTS-HD MA to lossy 5.1 Dolby Digital. So my question is, why not extract 'and play' the lossy 5.1 DTS stream instead of encoding and playing a lossy 5.1 Dolby Digital stream?

[tebasuna51]

Of course, but some players (like some TV's, for instance the mine) can't support even the lossy core DTS, without problem with AC3 (my TV can send by SPDIF the AC3, also EAC3 5.1.2 to my Denon 5.1.2)

There are also the size to store, the AC3 640 Kb/s have equivalent quality than DTS core 1510 Kb/s (size x 2.3)

[DanDare1983]

Hi just a quick question. I have a playlist where the DTS-HD MA track has a delay of -1001ms. eac3to gets it down to -2ms which i understand is perfectly fine. Can I ask if the DTS core also has the same delay? If so, does that also get fixed?

[tebasuna51]

In fact when eac3to do the delay correction only add silence frames at the begining (delay +) , and that silence frames are only core frames because don't need MA extensions to codify silence.
To delay - delete first full frames (core+MA extensions)
Then the core is fixed also.