[LOCKED] EAC3TO Bug Reports and Feature Requests Only

[Curly]

Ripping Elemental UHD right now...

[skull]

Beautiful song selections!

Here is mine:

I ran some tests on latest v3.40 test version, with UHD BD that had Dolby Vision as well as variety of audio tracks, including TrueHD (no seamless branching or anything unusual). Results are excellent. No issues and created .thd, .ac3 and merged .thd+ac3 files, as expected. I ran -demux and track selected command lines, with and without .ini file present, all good!

Keep up the great work Curly!

P.S. A small suggestion that occurred to me - while in beta/test versions only, it may be good to have the build date/timestamp of eac3to_mod to be written at the beginning of the run (e.g. eac3to_mod v3.40 build 2023.11.11.101010), at least for the -log file? This would ensure you can tell which exact build folks are running, if they submit a bug report. Just a thought I had.

[Curly]

Thank you for your testing, skull. So far, so good.

It's a good thought for build numbers but ya know the effort in managing that is not inconsiderable. There's a double negative to confuse you, knerk! For DG-tools slipstreams we would just ask people for the file created date. It's probably enough here too. Some compulsives would put all test stuff under SVN, but that's overkill. We do a system backup every day. Only formal releases (not _test stuff) get put under SVN.

[skull]

Thank you for your testing, skull. So far, so good.

It's a good thought for build numbers but ya know the effort in managing that is not inconsiderable. There's a double negative to confuse you, knerk! For DG-tools slipstreams we would just ask people for the file created date. It's probably enough here too. Some compulsives would put all test stuff under SVN, but that's overkill. We do a system backup every day. Only formal releases (not _test stuff) get put under SVN.

No problem. Makes sense to me, not worth the trouble when you explain it.

[Curly]

I just did a test for the 4K Blu-ray of Elemental (2023), a disk with seamless branching. It has seven audio tracks, but only the TrueHD Atmos and the AC3 compatibility track included are synced (even though the eac3to log claims they can't be removed). All the others are out of sync.

I started working on this. The decoding/encoding messages are gone because we don't do that anymore. I'll check them all for sync. And I'm looking into the message about the THD can't be corrected when it is in fact corrected. Finally, gonna get rid of all the "Audio has a gap..." messages. That should be for debugging only.

I like your idea to use a code block for long logs, etc. So here is my rip log (with the audio gap messages suppressed):

Code: Select all

D:\Don\Programming\Delphi\EAC3TO Mod>eac3to_mod g:\rips\elemental 2) -demux
Running in fast mode
Keeping dialnorm
M2TS, 1 video track, 7 audio tracks, 3 subtitle tracks, 1:41:29, 24p /1.001
1: Chapters, 27 chapters
2: h265/HEVC, 2160p24 /1.001 (16:9), 10 bits
3: TrueHD/AC3 (Atmos), English, 7.1 channels, 48kHz, dialnorm: -27dB
   (embedded: AC3, 5.1 channels, 640kbps, 48kHz, dialnorm: -27dB)
4: E-AC3, English, 7.1 channels, 1024kbps, 48kHz, dialnorm: -27dB
   (core: AC3, 5.1 channels, 576kbps, 48kHz, dialnorm: -27dB)
5: AC3, English, 5.1 channels, 640kbps, 48kHz, dialnorm: -27dB
6: AC3, English, 2.0 channels, 320kbps, 48kHz, dialnorm: -27dB
7: AC3, English, 2.0 channels, 320kbps, 48kHz, dialnorm: -27dB
8: E-AC3, French, 7.1 channels, 1024kbps, 48kHz, dialnorm: -27dB
   (core: AC3, 5.1 channels, 576kbps, 48kHz, dialnorm: -27dB)
9: E-AC3, Spanish, 7.1 channels, 1024kbps, 48kHz, dialnorm: -27dB
   (core: AC3, 5.1 channels, 576kbps, 48kHz, dialnorm: -27dB)
10: Subtitle (PGS), English
11: Subtitle (PGS), French
12: Subtitle (PGS), Spanish
Creating file "00705 - Chapters.txt"...
a08 Extracting audio track number 8...
a09 Extracting audio track number 9...
a03 Extracting audio track number 3...
a04 Extracting audio track number 4...
a07 Extracting audio track number 7...
a05 Extracting audio track number 5...
s11 Extracting subtitle track number 11...
s12 Extracting subtitle track number 12...
a06 Extracting audio track number 6...
a03 Extracting audio track number 3...
s10 Extracting subtitle track number 10...
a03 Extracting AC3 stream...
v02 Extracting video track number 2...
a03 Extracting TrueHD stream...
v02 Creating file "00705 - 2 - h265, 2160p24.h265"...
a03 Creating file "00705 - 3 - TrueHD+AC3 (Atmos), English, 7.1 channels, 48kHz.thd"...
a09 Creating file "00705 - 9 - E-AC3, Spanish, 7.1 channels, 1024kbps, 48kHz.eac3"...
a08 Creating file "00705 - 8 - E-AC3, French, 7.1 channels, 1024kbps, 48kHz.eac3"...
a04 Creating file "00705 - 4 - E-AC3, English, 7.1 channels, 1024kbps, 48kHz.eac3"...
a03 Creating file "00705 - 3 - TrueHD+AC3 (Atmos), English, 7.1 channels, 48kHz.ac3"...
a05 Creating file "00705 - 5 - AC3, English, 5.1 channels, 640kbps, 48kHz.ac3"...
a06 Creating file "00705 - 6 - AC3, English, 2.0 channels, 320kbps, 48kHz.ac3"...
a07 Creating file "00705 - 7 - AC3, English, 2.0 channels, 320kbps, 48kHz.ac3"...
a03 Skipping identical AC3 frames (seamless branching)...
a03 Skipping identical AC3 frames (seamless branching)...
a03 Optimizing TrueHD stream (seamless branching)...
a03 Optimizing TrueHD stream (seamless branching)...
a04 Skipping identical AC3 frames (seamless branching)...
a09 Skipping identical AC3 frames (seamless branching)...
a08 Skipping identical AC3 frames (seamless branching)...
a05 Skipping identical AC3 frames (seamless branching)...
a07 Skipping identical AC3 frames (seamless branching)...
a06 Skipping identical AC3 frames (seamless branching)...
s10 Creating file "00705 - 10 - Subtitle (PGS), English.sup"...
s11 Creating file "00705 - 11 - Subtitle (PGS), French.sup"...
s12 Creating file "00705 - 12 - Subtitle (PGS), Spanish.sup"...
a03 The audio gaps/overlaps can't be removed from the TrueHD bitstream.
a03 In order to remove them you'll have to transcode to another format.
a03 Starting 2nd pass...
a03 Realizing (E-)AC3 gaps...
a03 Creating file "00705 - 3 - TrueHD+AC3 (Atmos), English, 7.1 channels, 48kHz.ac3"...
a04 Starting 2nd pass...
a04 Realizing (E-)AC3 gaps...
a04 Creating file "00705 - 4 - E-AC3, English, 7.1 channels, 1024kbps, 48kHz.eac3"...
a05 Starting 2nd pass...
a05 Realizing (E-)AC3 gaps...
a05 Creating file "00705 - 5 - AC3, English, 5.1 channels, 640kbps, 48kHz.ac3"...
a06 Starting 2nd pass...
a06 Realizing (E-)AC3 gaps...
a06 Creating file "00705 - 6 - AC3, English, 2.0 channels, 320kbps, 48kHz.ac3"...
a07 Starting 2nd pass...
a07 Realizing (E-)AC3 gaps...
a07 Creating file "00705 - 7 - AC3, English, 2.0 channels, 320kbps, 48kHz.ac3"...
a08 Starting 2nd pass...
a08 Realizing (E-)AC3 gaps...
a08 Creating file "00705 - 8 - E-AC3, French, 7.1 channels, 1024kbps, 48kHz.eac3"...
a09 Starting 2nd pass...
a09 Realizing (E-)AC3 gaps...
a09 Creating file "00705 - 9 - E-AC3, Spanish, 7.1 channels, 1024kbps, 48kHz.eac3"...
Video track 2 contains 145943 frames.
Subtitle track 10 contains 1950 captions.
Subtitle track 11 contains 1258 captions.
Subtitle track 12 contains 1265 captions.
eac3to processing took 2 minutes, 9 seconds.
Merging THD and AC3 for "00705 - 3 - TrueHD+AC3 (Atmos), English, 7.1 channels, 48kHz"
Done.

[Curly]

Got some preliminary results. All the audio tracks are out of sync by a lot, except the THD track! The THD is the one that produces the warning saying it couldn't be corrected.

So then on a hunch I re-ripped everything with -no2ndpass. After that I tried with AC3 audio track 5 and amazingly it is in sync. I assume that will also fix the others but will test it.

Tomorrow I will tell you about what the 2nd pass is all about (madshi told me). So we just have to figure out what is wrong with the 2nd pass processing. I do know that eac3to and DGDemux do things completely differently.

[Curly]

Guys, please re-download build 3.40_test. With Elemental UHD, all the audio tracks are now in sync. You do not have to give the -no2ndpass option.

I haven't checked subs yet, will do that later today.

Thanks to madshi for support on this issue.

https://rationalqm.us/misc/eac3to_mod_3.40_test.rar

[Trunks]

Indeed, the audio tracks are now in sync. The subtitles are not, but I am sure you'll find a solution.

Good job!

[Curly]

You are welcome and thank you for your confidence in me. Now, how could I possibly disappoint you?

BTW, are the subs in sync with DGDemux?

[Trunks]

Yes, subtitles are in sync with DGDemux.

By the way, would it be possible to keep the debug messages like "[a08] Audio has a gap of 102ms at playtime 0:03:11. <WARNING>" with a setting in the .INI file?

[Curly]

Ha ha, it turns out the option isn't needed. Those messages were a result of the broken gaps processing that was fixed in the latest test version. I'll explain more later but it is perfectly correct and fine that those warnings are not triggered. I don't have to comment them out.

[Curly]

I confirm DGDemux sups are in sync all the way through. For eac3to_mod, they start in sync but are out by 5 seconds by the end.

Now I have to figure out how Rocky did gaps correction for sups, and then look at what eac3to_mod is doing. Nice work, Rocky!

[Thunderbolt]

is it still save enough to run the cmd line with the -fast command? or were those CRC checks superfluous?

Mod edit: this post is an example of what belongs in General Discussion thread, please use this as a lesson, one and all. Fair game to be deleted in future. Thanks!

[Curly]

Yes it is safe. They were not superfluous. They just caught some rare stream corruption errors. If you'd hit that your rip would be bad. The -fast option may also be used to control future performance improvements in other areas.

You are a bad boy for not posting this in the general discussion thread. Please do so in the future. This is for solid bug and feature stuff. Thank you.

[Curly]

OK, guys, after some squirreled detective work (not dogged, get it?), I found the cause of the subs sync issue. Now to figure out what to do about it...tomorrow.

[Trunks]

This needs confirmation, but it looks like there is a bug with eac3to_mod that doesn't exist with the original eac3to.

It seems you can't demux the TrueHD (Atmos) and the AC3 for compatibility when the name is the same (but the extension is different, of course). There is no problem when you use -demux.

For example: eac3to_mod.exe M:\Elemental 2) 3:English.thd 3:English.ac3 -progressnumbers

You will see that the demux process suddenly stops. The percentage when it stops is never the same.

Code: Select all

eac3to_mod.exe M:\Elemental 2) 3:English.thd 3:English.ac3 -progressnumbers
Running in normal mode
Keeping dialnorm
analyze: 1%
analyze: 2%
analyze: 3%
analyze: 4%
analyze: 5%
analyze: 6%
analyze: 7%
analyze: 8%
analyze: 9%
analyze: 10%
analyze: 11%
analyze: 12%
analyze: 13%
analyze: 14%
analyze: 15%
analyze: 16%
analyze: 17%
analyze: 18%
analyze: 19%
analyze: 20%
analyze: 21%
analyze: 22%
analyze: 23%
analyze: 24%
analyze: 25%
analyze: 26%
analyze: 27%
analyze: 28%
analyze: 29%
analyze: 30%
analyze: 31%
analyze: 32%
analyze: 33%
analyze: 34%
analyze: 35%
analyze: 36%
analyze: 37%
analyze: 38%
analyze: 39%
analyze: 40%
analyze: 41%
analyze: 42%
analyze: 43%
analyze: 44%
analyze: 45%
analyze: 46%
analyze: 47%
analyze: 48%
analyze: 49%
analyze: 50%
analyze: 51%
analyze: 52%
analyze: 53%
analyze: 54%
analyze: 55%
analyze: 56%
analyze: 57%
analyze: 58%
analyze: 59%
analyze: 60%
analyze: 61%
analyze: 62%
analyze: 63%
analyze: 64%
analyze: 65%
analyze: 66%
analyze: 67%
analyze: 68%
analyze: 69%
analyze: 70%
analyze: 71%
analyze: 72%
analyze: 73%
analyze: 74%
analyze: 75%
analyze: 76%
analyze: 77%
analyze: 78%
analyze: 79%
analyze: 80%
analyze: 81%
analyze: 82%
analyze: 83%
analyze: 84%
analyze: 85%
analyze: 86%
analyze: 87%
analyze: 88%
analyze: 89%
analyze: 90%
analyze: 91%
analyze: 92%
analyze: 93%
analyze: 94%
analyze: 95%
analyze: 96%
analyze: 97%
analyze: 98%
analyze: 99%
analyze: 100%
M2TS, 1 video track, 7 audio tracks, 3 subtitle tracks, 1:41:29, 24p /1.001
1: Chapters, 27 chapters
2: h265/HEVC, 2160p24 /1.001 (16:9), 10 bits
3: TrueHD/AC3 (Atmos), English, 7.1 channels, 48kHz, dialnorm: -27dB
   (embedded: AC3, 5.1 channels, 640kbps, 48kHz, dialnorm: -27dB)
4: E-AC3, English, 7.1 channels, 1024kbps, 48kHz, dialnorm: -27dB
   (core: AC3, 5.1 channels, 576kbps, 48kHz, dialnorm: -27dB)
5: AC3, English, 5.1 channels, 640kbps, 48kHz, dialnorm: -27dB
6: AC3, English, 2.0 channels, 320kbps, 48kHz, dialnorm: -27dB
7: AC3, English, 2.0 channels, 320kbps, 48kHz, dialnorm: -27dB
8: E-AC3, French, 7.1 channels, 1024kbps, 48kHz, dialnorm: -27dB
   (core: AC3, 5.1 channels, 576kbps, 48kHz, dialnorm: -27dB)
9: E-AC3, Spanish, 7.1 channels, 1024kbps, 48kHz, dialnorm: -27dB
   (core: AC3, 5.1 channels, 576kbps, 48kHz, dialnorm: -27dB)
10: Subtitle (PGS), English
11: Subtitle (PGS), French
12: Subtitle (PGS), Spanish
a03 Extracting audio track number 3...
a03 Extracting audio track number 3...
a03 Extracting AC3 stream...
a03 Extracting TrueHD stream...
a03 Creating file "English.thd"...
a03 Creating file "English.ac3"...
process: 1%
a03 Skipping identical AC3 frames (seamless branching)...
a03 Skipping identical AC3 frames (seamless branching)...
a03 Optimizing TrueHD stream (seamless branching)...
process: 2%
a03 Optimizing TrueHD stream (seamless branching)...
process: 3%
process: 4%
process: 5%
process: 6%
process: 7%
process: 8%
process: 9%
process: 10%
process: 11%
process: 12%
process: 13%
process: 14%
process: 15%
process: 16%
process: 17%
process: 18%
process: 19%
process: 20%
process: 21%
process: 22%
process: 23%
process: 24%
process: 25%
process: 26%
process: 27%
process: 28%
process: 29%
process: 30%
process: 31%
process: 32%
process: 33%
process: 34%
process: 35%
process: 36%
process: 37%
process: 38%
process: 39%
process: 40%
process: 41%
process: 42%
process: 43%
process: 44%
process: 45%

Actually, I was able to demux with this command, but it took several attempts. And when it worked once, it failed immediately after the next try.

There is no problem when the names are different (like 3:Atmos.thd 3:Compatibility.ac3). It should be noted that two files are not merging into a thd+ac3 file here. I am assuming it's because of the different names.

[Curly]

Try with spaces after 3:

eac3to_mod.exe M:\Elemental 2) 3: English.thd 3: English.ac3 -progressnumbers

Working fine here. And yes, we merge only when the filenames are the same.

[skull]

Try with spaces after 3:

eac3to_mod.exe M:\Elemental 2) 3: English.thd 3: English.ac3 -progressnumbers

Working fine here. And yes, we merge only when the filenames are the same.

Yes, the syntax for eac3to always required the space. Curly, I wonder if there is some internal logic reason for this or could a future version be more flexible and allow to work okay, if someone forgets/doesn't put a space after the trackno: ?

[Curly]

We should do something, I agree. Probably issue an error message. Just crashing is low class.

Making progress on sups. It is a complicated problem. I have it working perfectly with one stream, but multiple streams collide. Gonna disambiguate using the PID.

[skull]

We should do something, I agree. Probably issue an error message. Just crashing is low class.

Making progress on sups. It is a complicated problem. I have it working perfectly with one stream, but multiple streams collide. Gonna disambiguate using the PID.

Take your time. The turtle always bests the hare.

[frankaboy]

Hello Curly,
as you know I am on Linux.
So I will wait patiently for the Linux port of eac3to_mod.

May you can give a statement to following subject, please.

As far as I know, eac3to is able to demux base and dependent streams from 3D Disk backups.
That works fine, if the backup tool (Anydvd) creates a one-to-one copy of SSIF folder.

In case of a MakeMKV backup, eac3to does not find the "dependent stream".
MakeMKV creates only ssif.smap files in the SSIF folder, in order to safe hard disk space.

From my point of view it would be great,
if eac3to_mod is able to demux the dependent stream of MakeMKV backups, as well.
However, it is just a suggestion by me.

If that make sense from your point of view too, and it is not to much work,
please implement demuxing of dependent stream for MakeMKV backups, as well.

Many thanks for your work and engagement.

[Curly]

greets Frankaboy i missed u
you shud get an avatar
so everybody can luv u
please give source, command u issued, and results that did not satisfy
that is what we call solid

[frankaboy]

Hi Curly,

Backup 3D disk with Anydvd:
Command line:
C:\eac3to\eac3to_mod.exe G:\HOBBIT_BOTFA_EE_3D_PART1\BDMV\PLAYLIST\00100.mpls

Eac3to read the corresponding ssif file and is able to demux of base and dependent streams.
1) 00100.mpls, 00049.m2ts, 1:30:54
- Chapters, 27 chapters
- h264/AVC (left eye), 1080p24 /1.001 (16:9)
- h264/MVC (right eye), 1080p24 /1.001 (16:9)

Backup same 3D disk with MakeMKV:
Same command line:
C:\eac3to\eac3to_mod.exe G:\HOBBIT_BOTFA_EE_3D_PART1\BDMV\PLAYLIST\00100.mpls

Eac3to do not find the base and dependent streams.
1) 00100.mpls, 00049.m2ts, 1:30:54
- Chapters, 27 chapters

Demuxing of base und dependent streams should be possible for MakeMKV backups too.

[Curly]

Thanks, I got it now. Adding to the list.

[frankaboy]

Perfect, many thanks Curly