Subtitle Edit

[Music Fan]

Hi,
I'm quite confused with a pattern to replace, I can't target 3 suspension points following a new line and followed by another character.
I want to add a space after these 3 points, thus only when they begin the line (I have another reg ex when they are preceeded by a space).

For instance, I'd like this ;
1
00:02:37,309 --> 00:02:39,516
Hello...

2
00:02:39,687 --> 00:02:42,641
...hi ...you.
to become ;
1
00:02:37,309 --> 00:02:39,516
Hello...

2
00:02:39,687 --> 00:02:42,641
... hi ...you.

I believed this could work but no ;
(\n)(\.{3})(\w)
replace by ;
$1$2 $3
:o

[markfilipak]

Hi,
(\n)(\.{3})(\w)

Slow down my friend. You're telling regexp to find a word boundary.

Try this instead:

(\n)(\.{3})([^\s])

Also try this:

(\n)(\.{3})([\S])

Also try this:

^(\.{3})([\S])

[Music Fan]

Thanks but neither of these 2 detects the pattern.

[markfilipak]

Thanks but neither of these 2 detects the pattern.

Then try these:
(\n)(\.\.\.)([\S])
(\n)(\.\.\.)([^ ])
^(\.\.\.)([^ ])

[Music Fan]

Great, this works ;
^(\.{3})([^ ])
equals ;
^(\.\.\.)([^ ])
replace by ;
$1 $2
:thanks:

Do you know why \n is not taken into account ?

[markfilipak]

Great, this works ;
Do you know why \n is not taken into account ?
Nope. Every regexp processor I've used has bugs. SE can't seem to make up its mind regarding how to handle line-endings between subs versus line-endings within subs.

[Music Fan]

Ok, good to know.

[markfilipak]

Do you know why \n is not taken into account ?
If
- your subs are text (e.g., SRT), and
- you have a text editor for UTF-8, and
- your text editor does regexp.
I'd recommend you use your text editor rather than SE.

[markfilipak]

"The Ghost and Mrs. Muir" [1947] DVD, UPC 024543071426 (2023)

I've written about this in the past two days.
As it turns out, the real problem is that the audio is out of sync. The movie's running time is 102 minutes. By the 30 minute mark, the audio lags by about 200ms and placing subtitle cues begins to get difficult. By the end, the audio lags by about 600ms.

The fault is not with SE. The fault is with Mill Creek Entertainment, the folks who made the DVD. Perhaps my experience, documented here, will help other SE users.

FOLLOW-UP: Below is how I fixed the movie MP4. I
- sped up the dialog audio (-map "[a]") by a factor of 1.00005 (found by trail and error) and
- mixed in the sped up dialog (-c:a:0 ac3) and
- added the two audio comment streams (-map 0:a:4 -map 0:a:5, -c:a:1 copy -c:a:2 copy) and
- deleted closed captions
using the following script:
set _SOURCE_="c:\The Ghost and Mrs. Muir\copy_b.mp4"
: copy_b is the concatenated VOBs made via copy /b then transcoded to AVC using my special sauce
set _CHAPTERS_="c:\The Ghost and Mrs. Muir\chapters.txt"
set _TARGET_="c:\The Ghost and Mrs. Muir [1947].mp4"
set _STRETCH_AUDIO_=-filter_complex "[0:a]atempo=1.00005[a]"
set _DELETE_CC_=-bsf:v "filter_units=remove_types=6"
ffmpeg^
-i %_SOURCE_%^
-i %_CHAPTERS_%^
%_STRETCH_AUDIO_%^
-map 0:v -map "[a]" -map 0:a:4 -map 0:a:5 -map_metadata 1^
%_DELETE_CC_%^
-c:v copy -c:a:0 ac3 -c:a:1 copy -c:a:2 copy^
-dn^
%_TARGET_%
I will now fix the SRT subs that I made in a previous step, which should be much easier now, and mix them in at the end.

[Music Fan]

If
- your subs are text (e.g., SRT), and
- you have a text editor for UTF-8, and
- your text editor does regexp.
I'd recommend you use your text editor rather than SE.
Ok, I believe I can do that with PowerShell (I don't know other methods).

[TR-9970X]

"The Ghost and Mrs. Muir" [1947] DVD, UPC 024543071426 (2023)

I've written about this in the past two days.

The fault is not with SE.

I will now fix the SRT subs that I made in a previous step, which should be much easier now, and mix them in at the end.

I just had to say that you've gone to a huge effort to "fix" this, and it's still DVD resolution (unless you're going to upscale in the future).

And progressive sync problems are a real pain.

The copy I got is 720p, with sub's, and it's all good, despite it being B&W, and a little grainy, it's good.

[markfilipak]

This should be useful for anyone experiencing audio out of sync.
I just had to say that you've gone to a huge effort to "fix" this, and it's still DVD resolution (unless you're going to upscale in the future).
Every problem holds a serpent in one hand and a laurel wreath in the other. Wisdom comes through struggle. (Note: I'm puzzled by "upscale". Why would anyone ever do that?)

And progressive sync problems are a real pain.
The audio is slightly early and needed to be stretched.

From ffprobe -show_streams
Before stretching: duration=6265.248021 seconds = 1:44:25.248021
After stretching: duration=6265.595708 seconds = 1:44:25.595708
Amount of stretching: 6265.595708/6265.248021 ==> atempo=1.00005
What waveforms shows doesn't make sense:
In the case of the DVD above, audio was early relative to video. Don't let that confuse the issue. Regardless how well audio syncs with video, sound and a picture of the sound should always be identical. If a sound is early, then a picture of the sound should also be early. When SE waveforms draws pictures of sounds (audio), it must be using improper methods, methods derived from video timing.

[iKron]

is there anyway we can use https://github.com/tesseract-ocr/tessdata_best

[markfilipak]

Below are suggestions for integrating keyboard and mouse operations so to speed editing in the waveform panel.
INTERVAL is an internal process.
REWORK is suggestions shown to users after INTERVAL completes after displaying REWORK flags.
ALOOP, EDCUES, and MOVINTERVAL are user functions and will be of most interest to readers.

INTERVAL process: An automatic 'waveform' process that is run when SE knows where utterance_1 ends and utterance_2 starts.
This is called by EDCUES (below), not directly by users.
Let out_pad = 1000ms, in_pad = 300ms -- they can be changed of course, as can the 100ms shot change encroachment value.
"gap" is the minimal subtitle gap set by users.

// Set the gap (i.e., the minimum allowed space) to 2 frames.
// Adjust out-pads to 1000ms after utterances.
// Adjust in-pads to 300ms before utterances.

IF (X - gap) > 1300ms
// ___out-cue_1 ___in-cue_2
// / /
// ---utterance_1---|<--------------X-------------->|---utterance_2---
// ------------> <----
// 1000ms 300ms
// ---subtitle_1------------------| |---------subtitle_2----

ELSE IF (X - gap) = 1300ms
// ---utterance_1----------|<----------X----------->|---utterance_2---
// -------------> <----
// 1000ms 300ms
// ---subtitle_1--------------------------|gap|---------subtitle_2----

// When out-cue_1 and in-cue_2 collide with the gap, eat into out-pad_1.

ELSE IF (X - gap) >= 600ms
// ---utterance_1-------------------|<------X------>|---utterance_2---
// <----
// 300ms
// ---subtitle_1--------------------------|gap|---------subtitle_2----

// When the utterances are even closer, eat into both, but center in-cue_2 between the utterances -- favors subtitle_2.

ELSE IF X > gap
// ---utterance_1-----------------------|<----X---->|---utterance_2---
// <----
// X/2
// ---subtitle_1--------------------------|gap|---------subtitle_2----

// When the utterances are closer than the gap, allow the gap to overlap utterance_1 -- preserves the start of subtitle_2.

ELSE
// ---utterance_1---------------------------------| |---utterance_2---
//
// ---subtitle_1--------------------------------|gap|---subtitle_2----

IF utterance_1 ends less than 100ms after a shot change, move out-cue_1 to the shot change and flag the interval "REWORK".
IF a shot change is within out-pad_1, move out-cue_1 to the shot change, then move in_cue_2 to either (out-cue_1 + gap) or (in-cue_2 - 300ms), whichever is later.
IF a shot change is within the gap, do nothing.
IF a shot change is within in-pad_1, move in-cue_2 to the shot change, then move out-cue_1 to either (out-cue_1 + 1000ms) or (in-cue_2 - gap), whichever is earlier.
IF utterance_2 begins less than 100ms before a shot change, move in-cue_2 to the shot change and flag the interval "REWORK".

REWORK (recommendations)
1, If utterance_1 and utterance_2 have the same speaker, merge the subtitles, then split them somewhere else.
2, Consider merging the subtitles as a 2-speaker dialog subtitle.
3, Do nothing, retain the current in- and out-cues.


ALOOP user function: A user 'waveform' function for arbitrary audio looping.
THE_KEY is a label for the key that has been assigned to invoke this function.
Press and hold THE_KEY, click point_A, click point_B.
(The audio continuously loops.)
Drag either point any number of times.
(The audio loop responds to the new loop.)
Release THE_KEY.
(The audio stops.)
Point_A and point_B can be _any_ two points.


EDCUES user function: A user 'waveform' function for setting cues.
THE_KEY is a label for the key that has been assigned to invoke this function.
Press and hold THE_KEY, click an existing out_cue, click an existing in_cue.
(The audio begins to loop continuously.)
Drag out_cue to the left while audio loops until you hear the end of utterance_1.
Drop out_cue at the end of utterance_1.
Drag in_cue to the right while audio loops until you hear the start of utterance_2.
Drop in_cue at the start of utterance_2.
(The audio continues looping.)
Repeat drag-n-drop as needed, adjusting their positions if needed, while holding THE_KEY.
Release THE_KEY.
(The audio stops).
SE now knows where utterance_1 ends and utterance_2 starts and automatically runs the INTERVAL process.
Repeating EDCUES on the same cues or any other pair of cues is allowed at any time without restriction.


MOVINTERVAL user function: A user 'waveform' function for moving both in- and out-cues simultaneously.
THE_KEY is a label for the key that has been assigned to invoke this function.
Press and hold THE_KEY.
Click on an interval between subtitles.
Drag-n-drop the interval.
(Both out-cue and in-cue have been moved.)
Release THE_KEY.


FOCUS: Abandon the current focus strategies. Toggle the focus between text and waveform via a single control: The "Tab" key for example.


SPACE: Instead of the user setting the number of pixels in a space, could you prompt the user to identify the center of a space during training and then simply measure it yourself?

[dngnt]

There is a long standing bug in Subtitle Edit I suggest for fixing: when a mkv file contains multiple audio tracks, SE always chooses the first track for displaying and editing, it's impossible to choose the 2nd track.

[GCRaistlin]

There is a long standing bug in Subtitle Edit I suggest for fixing: when a mkv file contains multiple audio tracks, SE always chooses the first track for displaying and editing, it's impossible to choose the 2nd track.

You need to select "mpv" or "VLC media player" as Video engine in Settings - Video player. Then you'll be able to choose audio track in Video menu.
VLC should be x64. Also, see http://forum.doom9.org/showthread.php?p=2017155#post2017155 .

[dngnt]

You need to select "mpv" or "VLC media player" as Video engine in Settings - Video player. Then you'll be able to choose audio track in Video menu.
VLC should be x64. Also, see http://forum.doom9.org/showthread.php?p=2017155#post2017155 .

Thanks a lot!
I've changed from DirectX to mpv and a "choose your audio track" appears and it works!

[jay123210599]

Can someone help me with these?

https://forum.doom9.org/showthread.php?t=186374
https://forum.doom9.org/showthread.php?t=186365

[markfilipak]

Can someone help me with these?

https://forum.doom9.org/showthread.php?t=186374
https://forum.doom9.org/showthread.php?t=186365
What is it you are trying to do? I rip Blu-rays all the time and I have fast and effective subtitling methods that avoid the kind of troubles you are having.

[jay123210599]

What is it you are trying to do? I rip Blu-rays all the time and I have fast and effective subtitling methods that avoid the kind of troubles you are having.

Like what?

[markfilipak]

Like what?
What is it that you are trying to do?

[jay123210599]

What is it that you are trying to do?

I'm trying to replace the subtitle at the top of this (https://imgur.com/5gN0B3m) with this (https://imgur.com/AwW5b8e).

[jay123210599]

How does any of that help with my task?

[Emulgator]

I would like to control ±1 frame movements within "Wave Form/Spectrogram" window with Left/Right Arrow keys.

For some versions now, independent from where focus sits:
Assigning Left/Right arrow keys to have "Video -> One frame back/forward"
unfortunately takes the usual "Move cursor left/right" away from Text Box edits, always performing Video frame decrement/increment.
And Shortcuts->Text Box don't seem to address "Move cursor left/right" at all.

Clearing these Left/Right Arrow key assignments from "Video" restores Text Box "Move cursor left/right",
but as I would like to control ±1 frame movements with Left/Right Arrow keys within "Wave Form/Spectrogram" window
(as I am used to that function from Vegas) it is not offered there, so I went to "Video".

BTW, the shortcuts for focus on"Video" don't seem to apply if focus is on video preview window.
Or am I missing something here ?

Just tested: For 3612, 3613, 407, 408, 4010, 4012 I had not changed defaults.
These are playing nice, arrow keys perform per focus.
Text box: "Move cursor left/right"
Wave Form/Spectrogram "One second back/One second forward".

Too coarse for me, but well, for now I will settle for the offered ±100ms.

[markfilipak]

Two issues:
1, What do you make of this?
https://forum.doom9.org/attachment.php?attachmentid=18970&stc=1&d=1749943258
Note that I then click "Skip".

2, What's happened to SE? It seems to have lost its memory of past training.
https://forum.doom9.org/attachment.php?attachmentid=18971&stc=1&d=1749943492
Oh, dear.

--Mark.

[markfilipak]

Version 4.0.12.
Dark theme OFF-to-ON preserves existing waveform colors.
Dark theme ON-to-OFF causes default waveform colors to overwrite existing waveform colors.
That was also true of version 4.0.8. Other versions, unknown.

[markfilipak]

When I started this post, I started really thinking, really hard.

Do you know how to get out of maze? A maze like Jack Nicholson is in, in "THE SHINING". Well, you put one hand on the wall beside you and walk, never taking your hand off the wall. You will eventually get to an exit or you will double back. If you double back, put your hand on the opposite wall and try again.

I think SE is doing something like that with the letters during OCR. I think that's why having just the outlines, and everything else transparent, works so well. SE is sticking a 'hand' out and 'walking' the inside of the outline.

I'm posting what I initially wrote because identifying the 'Image palette' elements is still a worthy effort...

The following is part of OCR.
color picker
/ tick box
/ /
[X] Use custom colors [ ] [ ] Transparent [ ] [ ] Transparent [ ] [ ] Transparent [ ] [ ] Transparent
background outline body (unknown)
(clearly true) or or
body outline
(which?) (which?)
Which of the four [ ] [ ] Transparent groups handles outline? Which handles body?

IDX metadata has a very similar entry:
custom colors: OFF, tridx: 0000, colors: 000000, 000000, 000000, 000000

Which 000000 handles outline? Which handles body? Nobody says. The US Library of Congress doesn't say. matroska.org doesn't say. All the places I searched during a day of searching... None of them stated what group covers which feature. Do they even know? Do you?

I guess it doesn't matter because SE seems to ignore the metadata that's in the IDX file, anyway.

This:
[X] Use custom colors [ ] [X] Transparent [ ] [ ] Transparent [ ] [X] Transparent [ ] [X] Transparent
seems to work the best. Color doesn't seem to matter.

[Emulgator]

If you want to guess coming from DVD, the sequence there is b, p, e1, e2 (background, pattern, emphasis 1, emphasis 2)
which can be implemented as
background, body, antialiasing, outline.
or
background, body, antialiasing, shadow.

Here one may assume background, body, outline, shadow ?

Subpic Pixeltype Values for reference (mpucoder):
background b 00
pattern p 01
emphasis1 e1 10
emphasis2 e2 11

[Thunderbolt8]

after the last update, all my list for removing hearing impaired stuff, multiple replace and fix common error settings are gone. whats up with that? is it possible to fix this?

[Emulgator]

If you installed to a new location, SE starts vanilla.
If you kept the old version: Settings.xml sits in <your SubtitleEdit path here>
ATM in my version 4.0.12 there are 1740 lines controlling Subtitle Edit behaviour.

For .xml editing you may want to install notepad++
Right click on this .xml.file -> Properties -> Open with -> <give the path to notepad++>
From now on any .xml will be opened by notepad++.
Why notepad++ ? notepad++ keeps a history of edits if you set it that way.
Nice to have for pruning as well: WinMerge.
You pull two or three versions of any file onto GUI and get them compared visually (highlighted in colors)
You may purge/edit/inherit as you like, have different SE version side by side and so on.
WinMerge keeps backups too, side by side at the source file place, file name extended by date/time

Go to <your SubtitleEdit path here>\Dictionaries.
There are .xml files controlling the behaviour per subtitle language.

In \Dictionaries folder .xml files sort per language:
de_<> as an example, _se (or no prefix) is what Subtitle Edit brings, _user contains your interactions.

de_DE_se.xml
de_DE_user.xml
de_names.xml
de_names_user.xml
de_NoBreakAfterList.xml
deu_OCRFixReplaceList.xml
deu_OCRFixReplaceList_User.xml

[markfilipak]

If you want to guess coming from DVD, the sequence there is b, p, e1, e2 (background, pattern, emphasis 1, emphasis 2)
which can be implemented as
background, body, antialiasing, outline.
or
background, body, antialiasing, shadow.

Here one may assume background, body, outline, shadow ?

Subpic Pixeltype Values for reference (mpucoder):
background b 00
pattern p 01
emphasis1 e1 10
emphasis2 e2 11
Well, Emulgator, I'm glad to meet someone in the know, again.

In 'FATAL ATTRACTION [1987].idx' -- MPV will play subtitles contained in *.idx & *.sub, I tried various 'custom colors' settings.

custom colors: ON, tridx: 0000, colors: 000000, 000000, 000000, 000000
custom colors: ON, tridx: 0000, colors: ffffff, ffffff, ffffff, ffffff
custom colors: ON, tridx: 1111, colors: ffffff, ffffff, ffffff, ffffff

None had any effect. Do you know why?

[Thunderbolt8]

If you installed to a new location, SE starts vanilla.
If you kept the old version: Settings.xml sits in <your SubtitleEdit path here>
ATM in my version 4.0.12 there are 1740 lines controlling Subtitle Edit behaviour.

For .xml editing you may want to install notepad++
Right click on this .xml.file -> Properties -> Open with -> <give the path to notepad++>
From now on any .xml will be opened by notepad++.
Why notepad++ ? notepad++ keeps a history of edits if you set it that way.
Nice to have for pruning as well: WinMerge.
You pull two or three versions of any file onto GUI and get them compared visually (highlighted in colors)
You may purge/edit/inherit as you like, have different SE version side by side and so on.
WinMerge keeps backups too, side by side at the source file place, file name extended by date/time

Go to <your SubtitleEdit path here>\Dictionaries.
There are .xml files controlling the behaviour per subtitle language.

In \Dictionaries folder .xml files sort per language:
de_<> as an example, _se (or no prefix) is what Subtitle Edit brings, _user contains your interactions.

de_DE_se.xml
de_DE_user.xml
de_names.xml
de_names_user.xml
de_NoBreakAfterList.xml
deu_OCRFixReplaceList.xml
deu_OCRFixReplaceList_User.xml
I didnt explicitly pick a new location during installation. I did what I always used to do with past updates.

has a new default path been added to his version?

theres also no settings file in my install directory. afaik that ones rather in users/appdata or something?

[Music Fan]

Hi,

the settings.xml file is in C:\Users\Name\AppData\Roaming\Subtitle Edit\.

I always copy it before to update SE.

[StainlessS]

Hi Nikse,
I got a little bug for you.

See here:- https://forum.doom9.org/showthread.php?p=2020443#post2020443

Hi MicroFiche, [EDIT: we had up to 34c here, not quite so bearable humidity]
https://www.mediafire.com/file/ctpxbpkuk7h9o9h/MicroFiche.7z/file

Above [~18KB] contains original extracted subs (outfile.srt) and my Result.srt with scrolling subs eradicated.
I have not spent that much time checking them all, but looks good to me.
I'm gonna have a nap now and as I only just got what I think is pretty much my best expected result,
thought you might like to take a gander.

I'll still spend some time cleaning up the script before I do post it (maybe tomorrow).

Also contains avisynth comparison script to show input and output subs.

Requires Avisynth and TextSub() from VsFilter.dll # https://github.com/pinterf/xy-VSFilter/releases
EDIT: Suggest loading script into VirtualDub2 for perusal.

EDIT:
About 24 mins in, the original outfile.srt has the missing "00:" leading hours timings missing for a time,
and in the comparison script, those subs dont show up. [only the result.srt subs show, as we fix the missing '00:']
as described here


EDIT: To fix the original Outfile.srt, load it into SubtitleEdit, and do a renumber from the edit menu (I think),
then save, that fixes the times missing '00:' and will then work ok in the comparison script.

EDIT: Below missing from above quote
[CODE]
569
00:23:59,071 --> 00:23:59,670
>> NO, I'M ASKING.
WHAT IS "SATURDAY NIGHT LIVE"?
I DON'T KNOW THAT.

570
23:59,672 --> 24:03,774 # BAD Format, missing leading "00:"
WHAT IS "SATURDAY NIGHT LIVE"?
I DON'T KNOW THAT.
>> IT'S "SATURDAY NIGHT LIVE."


# ...


1404
59:57,944 --> 01:00:09,154 # BAD Format, missing leading "00:"
¶¶
>>> IT'S WEEKEND UPDATE, 50th
ANNIVERSARY EDITION.

1405
01:00:09,156 --> 01:00:19,564
>>> IT'S WEEKEND UPDATE, 50th
ANNIVERSARY EDITION.
WITH COLIN JOST AND MIKELE --


# OK from here till End



when you do a renumber of that file Outfile.srt, where hours '00:' are missing from both start and end times,
then is fixed by SE ok, but if only the start time has missing '00:' then is not fixed by SE.
EDIT: As in the FINAL red marked line in above code block.

Example file (outfile.srt) contained in Microfiche.7z linked in quote.

EDIT: Also, renumber dont quite work, [EDIT: when only start time hours 00: missing]

1404
59:57,944 --> 01:00:09,154 # EDIT: NOT Fixed on start '00:'
¶¶
>>> IT'S WEEKEND UPDATE, 50th
ANNIVERSARY EDITION.

1404 # EDIT: repeat line 1404
01:00:09,156 --> 01:00:19,564
>>> IT'S WEEKEND UPDATE, 50th
ANNIVERSARY EDITION.
WITH COLIN JOST AND MIKELE --

[markfilipak]

If you know, please tell me. What was SE trying to do? Thanks--Mark.

Date: 07/05/2025 07:52:43
SE: 4.0.12.0 - Microsoft Windows NT 6.1.7601 Service Pack 1 - 64-bit
Message: DownloadAsync failed - https://raw.github.com/SubtitleEdit/plugins/master/Plugins4.xml

System.Net.Http.HttpRequestException
Message: An error occurred while sending the request.
StackTrace: at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()
at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)
at Nikse.SubtitleEdit.Core.Http.HttpClientDownloader.<DownloadAsync>d__9.MoveNext()
System.Net.WebException
Message: Unable to connect to the remote server
StackTrace: at System.Net.HttpWebRequest.EndGetResponse(IAsyncResult asyncResult)
at System.Net.Http.HttpClientHandler.GetResponseCallback(IAsyncResult ar)
System.Net.Sockets.SocketException
Message: An attempt was made to access a socket in a way forbidden by its access permissions 185.199.110.133:443
StackTrace: at System.Net.Sockets.Socket.InternalEndConnect(IAsyncResult asyncResult)
at System.Net.Sockets.Socket.EndConnect(IAsyncResult asyncResult)
at System.Net.ServicePoint.ConnectSocketInternal(Boolean connectFailure, Socket s4, Socket s6, Socket& socket, IPAddress& address, ConnectSocketState state, IAsyncResult asyncResult, Exception& exception)

Plugins4.xml? Is this important?

[jpsdr]

Hello.

Since several versions, downloading dictionnary, either tesserac or "standard" is not working anymore. I don't remember exactly when it stopped working. The process is stuck to "waiting...", and i can't stop neither close or stop, i have to kill the programm in task manager.

[TR-9970X]

Hello.

Since several versions, downloading dictionnary, either tesserac or "standard" is not working anymore. I don't remember exactly when it stopped working. The process is stuck to "waiting...", and i can't stop neither close or stop, i have to kill the programm in task manager.

See if this works:-

https://github.com/UB-Mannheim/tesseract/wiki

You may have to add it to the Subtitleedit folder in users

[Yosho]

I desperately need an auto sub syncer in Subtitle Edit... I'm finding subs that are 3 - 4 seconds too long in some cases. It takes so long to retime everything by hand using waveform that I'm getting burned out.

[Music Fan]

There's an option to limit timecode's duration.

[junah]

May I request a feature to batch delete last line on every subtitle.

[varekai]

Accidentally clicked on "Restore default settings" and now I no longer have right click on .sup file and open with Subtitle Edit.
For .srt the option is still there. Can't find any setting in options?
Hope I don't have to go through the complicated process of creating a registry key in Win 10x64, help is appreciated.

Best regards,
varekai

[Emulgator]

Options -> Settings -> File type associations -> tickbox .SUP -> "Update file associations"

[varekai]

Options -> Settings -> File type associations -> tickbox .SUP -> "Update file associations"
Did exactly that, a dozen times but no luck, even reboot PC.

[Music Fan]

And if you right click on .sup file, Open, select Subtitle Edit in the list, always use this program (or something like that) ?

[varekai]

And if you right click on .sup file, Open, select Subtitle Edit in the list, always use this program (or something like that) ?
Rightclick on .sup I get "How do you want to open the file? Continue using this app" chose OK and Subtitle Edit opens.
But it doesn't stick to Subtitle Edit, ""How do you..." pops up again.
Doubleclick .sup works fine, I'll use that.
Had a déjà vu moment that this has happened before with .sup and opening with the right-click context menu.

About Options->Settings->File type associations
Button "Restore default settings" should maybe only restore the actual window opened, now it is restoring all settings.
Perhaps "Restore default settings" button only in "General" window?

[Music Fan]

Rightclick on .sup I get "How do you want to open the file? Continue using this app" chose OK and Subtitle Edit opens.
But it doesn't stick to Subtitle Edit, ""How do you..." pops up again.
Depends maybe on the OS.
With Windows 10 : Rightclick on .sup, Open with, How do you want to open this file, below I can select SE or Search an app in the MS Store, or search another app on the pc, and below I can tick "Always use this app for this type of file" (I translate from french).

[varekai]

Depends maybe on the OS.
With Windows 10 : Rightclick on .sup, Open with, How do you want to open this file, below I can select SE or Search an app in the MS Store, or search another app on the pc, and below I can tick "Always use this app for this type of file" (I translate from french).
That's what I do, but it doesn't stick, always a new popup window with same message "How...". Go :confused: figure...

[Music Fan]

Ok, there is maybe a Window's setting for files association.

[Emulgator]

If it does not stick, then right-click on SE icon and "Run as administrator" might help.
Win10P64 here too, I had to make myself real Admin first.

WinUgly-and-mean: The real Admin account is hidden by default.
The user will only notice if he tries to rename his account to "Administrator"
Then he is prompted: "The name is already in use", or similar.

Workaround translation from German:

Press [Win] + [R], type "lusrmgr.msc".
Select "User" > "Administrator".
Remove tick "Account deactivated"

From now on you can use the real Admin account and may your inputs be respected.

You will have to transfer (copy) all your settings from <your previous user name> to <Administrator>
or input everything new, from the desktop to everything that is stored under <Users>

Besides of that: regedit
Computer\HKEY_CLASSES_ROOT\.sup
Computer\HKEY_CLASSES_ROOT\SubtitleEdit.sup\shell\open\command
"C:\_PROG\! Subtitle Tools\SubtitleEditBeta\SubtitleEdit.exe" "%1"
and so on... (match to your Path of course)

[varekai]

OK, thanks for input.
I'll try that, I'll clone my OS drive before testing, just in case things goes sideways...

[markfilipak]

The audio and the waveforms made from the audio do not match. They start out in sync, but get out of sync: waveform leading vocalizations, as the video proceeds, getting further and further out of sync. I've tried reloading the video but, since it's the same video, nothing happens. I've searched the UI and on-line for a solution but have not found joy.

Is there a way to force the waveforms to be re-rendered?
UPDATE: Yes, there is a way: Video -> Close video file. Then reopen the video file and generate new waveforms. I did that.

UPDATE: The SE-embedded MPV player shows correct lip-sync. And the shot-changes added to the waveforms are correct. So, the problem seems to be limited to waveform rendering. It's the waveforms, only, that are out of sync.

Edit: For my three-hour movie, the waveforms at the beginning match what I hear but at the end they are 1.2 seconds early. To state the problem clearly, I hear the audio 1.2 seconds past the time that the waveforms show.

Edit: Oh, and one last thing. I listed all the PTSes via FFprobe and checked all the PTS intervals. They are all 1001 PTS ticks in duration [note]. So, for sure, the video is not changing FPS and there are no laggard frames. The video is pristine.

[note] The video is a 24/1.001 FPS MP4 with a 24000 ticks/second timebase clock, hence, 1001 ticks/frame.

Edit: How did this happen? I may have found the cause... This gets a little complicated.

The video is an MP4. MP4s, like all MOOV-based packets, have an editlist atom that tallies frames to be dropped when the video is decoded. It does that when an edit is made but the video is not re-encoded -- essentially, it gets the decoder to do the final edit (similar to the way soft telecine gets the decoder to change 24/1.001 FPS to 30/1.001 FPS).

When I created the MP4, audio at the end was early. (Why audio was early is unknown.) I had to -itsscale:a the audio by 1.000127 (that's stretching the audio, or so I thought). FFmpeg documentation claims that -itsscale:a changes the audio PTSes (samples, actually), which should stretch the audio, which should leave tiny audio gaps. But -itsscale:a apparently does not stretch the audio. -itsscale:a instead apparently put 31 video frames into the MP4 editlist! -- the FFmpeg log shows them.
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 276480 @ 180
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 278016 @ 181
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 279552 @ 182
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 281088 @ 183
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 282624 @ 184
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 284160 @ 185
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 285696 @ 186
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 287232 @ 187
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 288768 @ 188
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 290304 @ 189
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 291840 @ 190
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 293376 @ 191
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 294912 @ 192
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 296448 @ 193
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 297984 @ 194
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 299520 @ 195
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 301056 @ 196
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 302592 @ 197
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 304128 @ 198
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 305664 @ 199
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 307200 @ 200
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 308736 @ 201
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 310272 @ 202
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 311808 @ 203
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 313344 @ 204
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 314880 @ 205
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 316416 @ 206
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 317952 @ 207
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 319488 @ 208
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 321024 @ 209
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263f280] drop a frame at curr_cts: 322560 @ 210

FFmpeg did that to sync the audio during decoding.

My mind is getting fuddled. I'm going to create a short test video, rather than using a three-hour movie.

I'm leaving this here, as is, so that someone more experienced with MOOV atoms can comment and perhaps help me. ...There is definitely something wrong with waveforms when rendering an MP4 video that contains an editlist.

[TR-9970X]

Music symbol never gets recognised properly :(

I have been using SE for years, and it ALWAYS has problems with ♪.

It doesn't make any difference how I extract the subs from the video file, but running .sup's thru SE, it's just not good.

And even with the "Music Symbol" check box, checked.

It's VERY time consuming to manually change the ♪ it's missed (or used a different character) in Notepad :(

This is the original subtitle:- https://www.mediafire.com/file/ez2yud8h2j2a03o/Batwoman.S01E14.Grinning.from.Ear.to.Ear.Repack.1080p.BluRay.DDP.5.1.H.265.-EDGE2020_track3_%255Beng%255D.sup/file

And this the same after running thru SE:- https://www.mediafire.com/file/y38tfiq8yzadi1b/Batwoman.S01E14.Grinning.from.Ear.to.Ear.Repack.1080p.BluRay.DDP.5.1.H.265.-EDGE2020_track3_%255Beng%255D.srt/file

You'll soon see what a mess it is, checking from line 109 and onwards, with the .srt conversion.

PS:- It would also be very handy if you could edit individual lines as they scrolled by, and you spotted an error (line's in Yellow, etc)

[Music Fan]

I have been using SE for years, and it ALWAYS has problems with ♪.
With "binary image compare" instead of Tesseract, it's ok.
But it introduces other problems : "I" instead of "l" in some words ("Iater") and spaces in the middle of some words ("Hi" becomes "H i") and in front of some comas and points.
These problems can perhaps be solved after doing OCR with "Fix common erros" in the Tools tab.

[TR-9970X]

With "binary image compare" instead of Tesseract, it's ok.
But it introduces other problems : "I" instead of "l" in some words ("Iater") and spaces in the middle of some words ("Hi" becomes "H i") and in front of some comas and points.
These problems can perhaps be solved after doing OCR with "Fix common erros" in the Tools tab.

Thanks for the suggestion :)

I have always used Tesseract, and have "Fix Common Errors" checked.

So, just to confirm, use "Binary Image" first, then Tesseract ?? or ??

[Music Fan]

Only "Binary Image", no Tesseract.
Then try to fix the errors.

[TR-9970X]

Only "Binary Image", no Tesseract.
Then try to fix the errors.

OK, will give it a shot.

:thanks:

[Music Fan]

I tested, the fixes work for unneeded spaces and some OCR errors ("I" instead of "l"), look at line 8, 9 and 11 ;

https://i.ibb.co/PGg9Wz9N/SE-Batwoman-S01-E14-Grinning-from-Ear-to-Ear-Repack-track3-eng-Binary-compare.jpg (https://ibb.co/5hLnjxn6)

But "H i" (line 13) stays as is, you can fix it with the Multiple Replace tool in Edit tab.

[markfilipak]

I'm in big trouble here.

Audio and video and timestamps and shot changes are all in sync. What's not in sync is the waveforms. Waveforms go from in sync at the beginning to early by over a second at the end.

The packets are MP4. The MP4 has a MOOV editlist that roughly matches how the waveforms go early. ...I'm out of my depth! Help! I'm drowning!
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] Processing st: 0, edit list 0 - media time: 2002, duration: 266122872
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] Offset DTS by 2002 to make first pts zero.
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] Setting codecpar->delay to 2 for stream st: 0
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] Unknown dref type 0x206c7275 size 12
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] Processing st: 1, edit list 0 - media time: 324580, duration: 532220352
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 276480 @ 180
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 278016 @ 181
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 279552 @ 182
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 281088 @ 183
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 282624 @ 184
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 284160 @ 185
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 285696 @ 186
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 287232 @ 187
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 288768 @ 188
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 290304 @ 189
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 291840 @ 190
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 293376 @ 191
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 294912 @ 192
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 296448 @ 193
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 297984 @ 194
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 299520 @ 195
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 301056 @ 196
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 302592 @ 197
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 304128 @ 198
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 305664 @ 199
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 307200 @ 200
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 308736 @ 201
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 310272 @ 202
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 311808 @ 203
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 313344 @ 204
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 314880 @ 205
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 316416 @ 206
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 317952 @ 207
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 319488 @ 208
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 321024 @ 209
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] drop a frame at curr_cts: 322560 @ 210
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000261f280] skip 484 audio samples from curr_cts: 324096

[markfilipak]

SE 4.0.12 embedded this FFmpeg:
ffmpeg version 2024-05-20-git-127ded5078-full_build-www.gyan.dev

That is the final FFmpeg that ran in Win7.

My first guess for this problem is that SE fails to process MOOV editlists.

I've tried to make a one-minute test video but have failed. FFmpeg seems to be having trouble with this:
works works
--------- ---------
aselect=between(n\,2997*2002\,3515*2002),asetpts=PTS-2997*2002
---------
broken

[markfilipak]

I apologize for monopolizing this space ...it's complicated.

The bottom lines are:
1 - I need help in the form of a coworker because the complications are boggling my mind and I need to interact with someone to keep sane, and
2 - I think SE is not honoring MP4 editlists.

Of course, video edits should be completed prior to coding. When MP4 edits are made, a special moov-atom child: edts, is created that includes the edits that should be done prior to final coding. I reckon this is done so that authors can quickly watch the results of their editing without having to transcode the video. That way, the author can do multiple quick edits, see the results, then include the edits in their final coding.

Because I didn't know about editlists, which are particular only to moov-based packaging, I created an MP4 that includes edts "drop a frame at curr_cts:" entries and passed it to SE as the video source of the video's subtitles. It appears that SE did not honor the edts in regard to the waveforms, but _may_ have honored edts in regard to video, audio, timestamps, and shot changes because they are all in sync, both in the embedded MPV and an external MPV.

The edits I did scaled the audio timestamps, which became progressively out of sync, earlier and earlier, as playback progressed. To achieve the scaling, I thought FFmpeg would stretch the audio by occasionally duplicating audio samples or adding single samples of silence. Doing so would not require recoding the video. But it appears that what FFmpeg actually did was create an edts that drops video frames. That makes no sense to me and I have a complaint with the FFmpeg folks, but that detail has nothing to do with the SE problem I've outlined. In fact, the way SE behaves in this circumstance makes recognizing what is actually happening much harder to figure out.

I'm doing my final coding now. The coding takes a few hours. The video is over 3 hours long.

I should add this so you know I'm not a compete idiot: As I stretched the audio to achieve sync, I did recode the audio, but not the video. I recoded the audio because I thought FFmpeg would duplicate audio samples, not drop video frames. Of course, I'm not happy with what FFmpeg did, but, again, that has nothing to do with the SE waveform problem (or so I think).

[TR-9970X]

I tested, the fixes work for unneeded spaces and some OCR errors ("I" instead of "l"), look at line 8, 9 and 11 ;

https://i.ibb.co/PGg9Wz9N/SE-Batwoman-S01-E14-Grinning-from-Ear-to-Ear-Repack-track3-eng-Binary-compare.jpg (https://ibb.co/5hLnjxn6)

But "H i" (line 13) stays as is, you can fix it with the Multiple Replace tool in Edit tab.

I tried it on several subs that I knew had the ♪ symbol, and it seems to work quite well, however, as you mentioned, it can cause other small issues, mainly a space after I (mainly) at the start of a new sentence/line.

It does seem to take longer than what I used to do, BUT at least I don't have to painstakingly do it manually in Notepad...

But other than that, I might have to get used to a new process.

Thanks.

[Music Fan]

Actually, once you saved the case ("H i", replace by "Hi") or others in the Multiple Replace tool, you don't have to do it again, that's quite fast.
For unneeded spaces, the other tool (Fix Common Errors) is also fast.

[markfilipak]

I've done a lot of searching. This appears to be a very old bug that produces a wide variety of symptoms, most of which I've seen. A year and a half ago, audio streams that don't terminate and jobs that buffer hundreds of gigabytes.

IMPORTANT: How does SE create the waveforms? Does SE do it, or does FFmpeg do it?

I should add that old versions of FFmpeg errored out and produced no output at all. Newer versions just produce audio stream errors.

I should also add that 'dref' means 'data reference index' https://developer.apple.com/documentation/quicktime-file-format/video_sample_description/data_format

There are many trouble reports such as
[mov,mp4,m4a,3gp,3g2,mj2 @ 000002e7d956ed40] Unknown dref type 0x206c7275 size 12
[mov,mp4,m4a,3gp,3g2,mj2 @ 0x1891c40] Unknown dref type 0x206c7275 size 12
[mov,mp4,m4a,3gp,3g2,mj2 @ 0x55fe07ff1180] Unknown dref type 0x206c7275 size 12
[mov,mp4,m4a,3gp,3g2,mj2 @ 0000010c8a3264c0] Unknown dref type 0x206c7275 size 12
[mov,mp4,m4a,3gp,3g2,mj2 @ 0x7f98a3008200] Unknown dref type 0x206c7275 size 12

but no one that I've seen blames dref. ...always the same "Unknown dref type 0x206c7275" in a wide variety of troubles. I've not encountered a list of known drefs.

[markfilipak]

The coding finished. I now have the video with no editlist.

The symptom has not changed. Audio, timestamps, and shot changes are all in sync. SE's waveforms start in sync, then get earlier and earlier until, at the end of 3 hours, waveforms are about 1.3 seconds early.

Has anyone else seen this symptom?

EDIT: I regressed back 2 1/2 years, back to version 3.6.13. This problem also occurs in 3.6.13 with this video.
I don't know how to continue.

[TR-9970X]

Actually, once you saved the case ("H i", replace by "Hi") or others in the Multiple Replace tool, you don't have to do it again, that's quite fast.
For unneeded spaces, the other tool (Fix Common Errors) is also fast.

I installed 4.0.14 on a different PC today. and I ran my first sub thru using Binary Image. and it was f*cking horrible !!!!

There were so many errors. just spaces EVERYWHERE, despite the preview window showing no issues, but in the edit window, what a mess :(

I thought I'd try Paddle OCR...I don't know where it was getting it's lines from, but it had nothing to with the sub I was working on.

So back to Tesseract 5.5.0 for me.

But I do really like the Multiple Replace Tool, and the Fix Common Errors :)

[Music Fan]

So back to Tesseract 5.5.0 for me.
For subtitles with the ♪ symbol, if found the other method much faster.
There are spaces but they are all removed by Fix Common Errors.

[TR-9970X]

For subtitles with the ♪ symbol, if found the other method much faster.
There are spaces but they are all removed by Fix Common Errors.

Using Binary created so many spaces, I doubt that would be fixed that way.

Unless there is a different method, running a .sup thru, you have to accept or change ANY discrepancy it finds during the process, and thd amount of spaces would waste a LOT time.

Using Tesseract, adjust errors during the scan, then use Fix Common Errors, and maybe do further errors with option you told me about.

[Music Fan]

I tested both with your sup and the OCR process took around 20 seconds with binary image compare (Latin+latin next to Image database), I didn't have to accept anything during the scan (I also installed 4.0.14).
Then I went in Fix Common Errors and it needed 3 seconds to fix errors.
And finally in Multiple Replace to add "H i", replace by "Hi", "I n " by "In " (for line 18) and "U pg" by "Upg" (for line 19).

Less than 1 minute for the whole process and I got this ;
https://pastebin.com/WHEiu9Z0
Of course I didn't verify it entirely thus it may still contain errors.

[TR-9970X]

I tested both with your sup and the OCR process took around 20 seconds with binary image compare (Latin+latin next to Image database), I didn't have to accept anything during the scan (I also installed 4.0.14).
Then I went in Fix Common Errors and it needed 3 seconds to fix errors.
And finally in Multiple Replace to add "H i", replace by "Hi", "I n " by "In " (for line 18) and "U pg" by "Upg" (for line 19).

Less than 1 minute for the whole process and I got this ;
https://pastebin.com/WHEiu9Z0
Of course I didn't verify it entirely thus it may still contain errors.

OK, well, I'm pretty happy that you tested my "sample" :)

Well, I'm going to have to try this again with your settings.

Do you have all the options in Fix Commons Errors checked ??, or basically the defaults.

I guess the more additions to the Multiple Replace, the easier future scans will become.

Thanks so much for your ongoing help & suggestions :)

[Music Fan]

Do you have all the options in Fix Commons Errors checked ??, or basically the defaults.
I believe I had unchecked some of them.

I guess the more additions to the Multiple Replace, the easier future scans will become.
Indeed. And you can save your settings (C:\Users\name\AppData\Roaming\Subtitle Edit\Settings.xml) not to lose them in case of crash or future installation (after OS or PC change).

[TR-9970X]

I believe I had unchecked some of them.


Indeed. And you can save your settings (C:\Users\name\AppData\Roaming\Subtitle Edit\Settings.xml) not to lose them in case of crash or future installation (after OS or PC change).

Not having any reasonable success at all, I even downloaded the .sup that you ran, and I had to accept so many issues that I lost count.

So I'm clearly doing something VERY different to your workflow :(

As a side note, I am also trying to run a Star Wars .sup, and using Binary, it's just hopeless, I think it would take me hours to run it thru once, having to accept errors.....

Surely there has to better process :(

[Music Fan]

That's very strange, especially since we have the same SE version.
I simply open the sup file, the OCR window opens directly, I click on Start OCR and it's going fast. I don't have to accept anything.

https://i.ibb.co/0jrRfRx2/SE-Batwoman-S01-E14-sup.jpg (https://ibb.co/fYHzCzsF)

[TR-9970X]

That's very strange, especially since we have the same SE version.
I simply open the sup file, the OCR window opens directly, I click on Start OCR and it's going fast. I don't have to accept anything.

https://i.ibb.co/0jrRfRx2/SE-Batwoman-S01-E14-sup.jpg (https://ibb.co/fYHzCzsF)

I just noticed no OCR dictionary....

[Music Fan]

Ok, it's worth another try.

[TR-9970X]

Ok, it's worth another try.

I just ran two different subtitles through with no dictionary and the .srt is blank, except for time codes.

[Music Fan]

At the end of OCR process, I just click on OK, the OCR window closes and the subtitles appear in the main window.

[TR-9970X]

At the end of OCR process, I just click on OK, the OCR window closes and the subtitles appear in the main window.

So like you described, you click on the .sup file, it opens up into the ocr window, press start ocr, it goes through its process and then when you come to the end it's all in the window there that you can view, with no OCR dictionary.

Well, I did that on two separate.sup files, press start ocr faster than a blink of an eye, it's finished, and it's only time codes :(

I'm running Windows 11 25H2.

[Music Fan]

I have Windows 10 but it probably doesn't matter, your PC has been cursed.:scared:
Here is what I see at the end of OCR ;

https://i.ibb.co/tpgmQy7H/SE-Batwoman-S01-E14-sup-2.jpg (https://ibb.co/qYKy7btD)

And after clicking on OK ;

https://i.ibb.co/7dzb8pZ0/SE-Batwoman-S01-E14-sup-3.jpg (https://ibb.co/RkS7Z6sJ)

Then I go in Fix Common Errors then in Multiple Replace and it's over.

[TR-9970X]

I have Windows 10 but it probably doesn't matter, your PC has been cursed.:scared:
Here is what I see at the end of OCR ;

https://i.ibb.co/tpgmQy7H/SE-Batwoman-S01-E14-sup-2.jpg (https://ibb.co/qYKy7btD)

And after clicking on OK ;

https://i.ibb.co/7dzb8pZ0/SE-Batwoman-S01-E14-sup-3.jpg (https://ibb.co/RkS7Z6sJ)

Then I go in Fix Common Errors then in Multiple Replace and it's over.

Hi,

I'm using a different PC today, still W11 25H2, same SE.

If I use mine like you've shown above, it's just jibberish !!

I also have this pop up asking me to choose a character, every so often, like it's learning and adding to it's dictionary.

Has that Image database got anything to do with it ??

I also setup several "repairs" in Multiple Replace, but I can't get it to anything.

[Music Fan]

Hi,

I really don't know why we are not dealing with the same process (admitting our settings are the same), very strange.
Nikse555 could probably answer but he doesn't come often here.

[TR-9970X]

Hi,

I really don't know why we are not dealing with the same process (admitting our settings are the same), very strange.
Nikse555 could probably answer but he doesn't come often here.

Yes, it is VERY strange, for sure.

I guess sending him a PM to read these posts might be an idea.

Again, thanks for all your help.

[Music Fan]

I found a quicker method for you ;
-use Tesseract 5.5.0
-language : English
-engine mode : Original Tesseract only
-dictionary : english or none, it doesn't seem to matter
-during OCR, all music symbols are considered as J“
-once it's over, go in Multiple Replace and replace J“ by ♪

PowerShell can also be used to replace characters or words.

[TR-9970X]

I found a quicker method for you ;
-use Tesseract 5.5.0
-language : English
-engine mode : Original Tesseract only
-dictionary : english or none, it doesn't seem to matter
-during OCR, all music symbols are considered as J“
-once it's over, go in Multiple Replace and replace J“ by ♪

PowerShell can also be used to replace characters or words.

I think I may have tried something very similar..

If only it was only J", but I've seen so many other variants.

I will this a try tomorrow, and let you know.

Thanks again :)

[Music Fan]

With other engine modes, there are indeed several characters that replace ♪ depending on the lines.
But it seems that with "Original Tesseract only", they are all converted to J".

[Janusz]

Hi,
...
If I use mine like you've shown above, it's just jibberish !!

I also have this pop up asking me to choose a character, every so often, like it's learning and adding to it's dictionary.

Has that Image database got anything to do with it ??

I also setup several "repairs" in Multiple Replace, but I can't get it to anything.

1. To ensure we have the same settings in the program:
- Close SubtitleEdit if you have it running.
- Download version 4.0.14 of Subtitle Edit again from this address: https://github.com/SubtitleEdit/subtitleedit/releases/download/4.0.14/SE4014.zip
- Unzip the downloaded "zip" file to the "SE4014" directory on your desktop.
- Run the "SubtitleEdit.exe" program in this directory.
- Select "Layout" no. 12 on the toolbar.
- In the program options, find and uncheck "Tools/Prompt for unknow one letter words" (This option causes the program to prompt for each individual letter found during OCR.)
- Exit the program so that the changes are saved in a new "Settings.xml" file.

2. Restart the program and drag and drop your "Batwoman[eng].sup" file into the main program window.
- In the newly opened "Import/OCR" window, select:

[OCR method] = OCR via nOCR (this method introduces fewer errors into the recognized text, and besides, there's no need to add any new characters to the existing default database).

Increase [No of pixels is space] to 10-12 (to reduce the number of unnecessary extra spaces).

In [Language], leave Latin.

In the [Dictionary] section, set English to [Prompt for unknown words] and [Auto break paragraph if more then two lines] to [Disable] the options: [Prompt for unknown words] (the program will not expect acceptance for unknown words) and [Auto break paragraph if more then two lines].

3. Select [Start OCR].

In the [Subtitle text] window, lines with recognized text should begin to appear.

In the [Unknown words] window (on the right), you should see unrecognized words.

4. After the OCR process is complete, before closing the OCR window, you can add new words to the dictionaries using the buttons to the right of [Unknown words] to save them for future use.

5. You can now close the OCR window and continue working on the recognized text.

Edit:
I added a short video (https://www.mediafire.com/file/qckdwz44flnqp0q/K-002.mp4/file) showing what was described above.

[markfilipak]

I installed 4.0.14 on a different PC today. and I ran my first sub thru using Binary Image. and it was f*cking horrible !!!!

There were so many errors. just spaces EVERYWHERE, ... :(

I've not been following your thread, but I reckon you need to set the size of a space in the OCR. The size varies with the video source, so there's no one-size-fits-all setting. Try whatever comes up. If there are too many broken words, make the number of pixels in a space bigger. If there are too many joined words, make the number smaller. I'm not doing OCR at the moment, so I can't tell you where to find the size of a space or what it's called, but you will find it if you have a brain. :-)

Abort out of the OCR, set the size of a space, and restart OCR from zero. It's not too difficult.

Good Luck, Good Hunting.

By the way, I've found image OCR to be the best.

[TR-9970X]

OMG Janusz, this IS awesome :)

Your first post here in approx 2½ years, and that attached video was just so easy to follow, even if some of it didn't make too much sense to me, towards the end.

I have run several .sup's thru at your settings, and it does a brilliant job of the music symbols, but I've also learned a lot more about SE that I ever did.

I still have to do spell checks & corrections, and with Music Fan's help, I have also learned about the Fix Common Errors, and Multiple Replace, which is awesome, I may not need to use Notepad ever again. :)

I can't thankyou both enough for your help, I wish I had of known all this years & years ago...oh well...live & learn.

[TR-9970X]

I've not been following your thread, but I reckon you need to set the size of a space in the OCR. The size varies with the video source, so there's no one-size-fits-all setting. Try whatever comes up. If there are too many broken words, make the number of pixels in a space bigger. If there are too many joined words, make the number smaller. I'm not doing OCR at the moment, so I can't tell you where to find the size of a space or what it's called, but you will find it if you have a brain. :-)

Abort out of the OCR, set the size of a space, and restart OCR from zero. It's not too difficult.

Good Luck, Good Hunting.

By the way, I've found image OCR to be the best.

Hi, those instructions from Janusz have been a life saver :)

And thanks for your tips :)

If you choose OCR via nOCR, that setting (No of pixels is space) is in the top left corner.

Curious about this comment:-

By the way, I've found image OCR to be the best.

What is it referring to ? an SE setting or another app ?

[markfilipak]

What is it referring to ? an SE setting or another app ?

I should have written 'OCR method: Binary image compare'.

[markfilipak]

Hello, I couldn't find a way to merge Part-1 and Part-2. I have a plan and would appreciate your comments.

Part-1-subs and Part-2-subs already exist.

1 - I'll concatenate Part-1 and Part-2 videos to make Final video.
2 - Then I'll open Part-2-subs and use 'Synchronization'->'Adjust all times'
to move the Part-2-subs to the end of Final video.
3 - Then I'll save the SRT file and close SE.
4 - Then I'll make Final subs by concatenating Part-1-subs and Part-2-subs.
5 - Then I'll open Final subs and Final video and do final fixups.

Perhaps there's a simpler way, eh? Perhaps there's a way to import Part-1-subs at the end of step 2. I couldn't find it.

All bright ideas are appreciated. Thanks!

[Janusz]

...
All bright ideas are appreciated. Thanks!

You can find the subtitle merging tool here:
Menu/Tools/Append subtitle...
or:
Menu/Tools/Join subtitles...

[markfilipak]

@Janusz, Thank you.

Help me with the meaning of this:
"This will append an existing subtitle to the currently loaded subtitle which should already be in sync with video file."


This adds subtitles to the end of the current subtitles and it assumes the current subtitles are in sync with the video.

Is that correct, or is it backwards?

[Janusz]

@Janusz, Thank you.

...

This adds subtitles to the end of the current subtitles and it assumes the current subtitles are in sync with the video.

Is that correct, or is it backwards?

Yes. Both parts of the subtitles are properly synchronized, so:
Load the first part of the subtitles.
Select "Menu\Tools\Append subtitle" in the "Append subtitle" window, select [Yes], and in the next window, add the second part of the subtitles.
Don't select [Sync], just [OK]. And [OK] again.
The subtitles have been added to the end of the first part.
You'll be taken to the main program window. You can now load the combined video file and check the synchronization.

[markfilipak]

@Janusz, Thank you.

Your translator is very effective...!

I will try it when my current FFmpeg transcode is done. I will tell you the results. Thanks again.

[Music Fan]

In case you append subtitles, I wonder how SE can estimate the duration at the end of the first part to add the good duration before the second part.
In a movie, you can have several minutes after the last line of text and nothing in the srt indicates how many minutes of video there are after this line.

[TR-9970X]

In case you append subtitles, I wonder how SE can estimate the duration at the end of the first part to add the good duration before the second part.
In a movie, you can have several minutes after the last line of text and nothing in the srt indicates how many minutes of video there are after this line.

I think I've done this only once, and I was surprised that it was spot on.

But in the chance that the second part is out of sync, you could probably determine how out of sync, you could go back to the original and do an adjustment, and then appended again, until you get it right....

[markfilipak]

I have found two bugs in FFmpeg. They affect audio samples in MP4 editlists. The bugs cause waveforms to be erroneously drawn early (by 700 ms in one case, by 1300 ms in another, confirmed). There is a workaround.

When SE uses FFmpeg to get audio samples from which to draw waveforms, can it pass '-ignore_editlist' to FFmpeg?

[TR-9970X]

I have found two bugs in FFmpeg. They affect audio samples in MP4 editlists. The bugs cause waveforms to be erroneously drawn early (by 700 ms in one case, by 1300 ms in another, confirmed). There is a workaround.

When SE uses FFmpeg to get audio samples from which to draw waveforms, can it pass '-ignore_editlist' to FFmpeg?

Can you update the FFMPEG that SE uses ?

FFMPEG just got an update, to 8.01.

[markfilipak]

Can you update the FFMPEG that SE uses ?


ffmpeg version 2025-11-12-git-6cdd2cbe32-full_build-www.gyan.dev

I just isolated the bugs today. They are old, very old, but elusive. That's why -ignore_editlist exists: because they couldn't find the real bug, so they tried to paper over it. Well, the real bug ruins -ignore_editlist, too.

I think FFmpeg needs an -ignore_bugs option.

[markfilipak]

Yes. Both parts of the subtitles are properly synchronized, so:
Load the first part of the subtitles.
Select "Menu\Tools\Append subtitle" in the "Append subtitle" window, select [Yes], and in the next window, add the second part of the subtitles.
Don't select [Sync], just [OK]. And [OK] again.
The subtitles have been added to the end of the first part.
You'll be taken to the main program window. You can now load the combined video file and check the synchronization.

That produced a train wreck. SE misbehaved excessively.

Instead, I did what I previously outlined.
In SE
- Load file2.mp4 file2.srt.
- Synchronization->Adjust all times:
- Add the running time of file1.mp4.
- Save file2.srt.
In Windows
- copy file1.srt+file2.srt file.srt
In SE
- Load file.srt.
- Load file.mp4.
Done.
Well, I had to adjust the subtitles a bit.

Edit:
Thank you, Janusz. I was continuing to work on it hoping to add good news to my thanks. I do that now, but I don't have good news. :(

Until FFmpeg fixes the bugs I found, I'm stuck. And right now, the ffmpeg-user list is not functioning. :(

[markfilipak]

So that everyone understands the issue...

When I play part 1 via 'FFplay -ignore_editlist 1', the audio is approximately 700 ms early; for part 2, it's approximately 1300 ms early.

When I edit part 1 via SE, the waveforms that SE draws are approximately 700 ms early; for part 2, they are approximately 1300 ms early.

Those are not coincidences.

The FFmpeg report for part 1 shows
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000273cfc0] Processing st: 1, edit list 0 - media time: -1, duration: 31728

(31728 samples)/(48000 samples/sec) = 661 ms.

The FFmpeg report for part 2 shows
[mov,mp4,m4a,3gp,3g2,mj2 @ 000000000263cfc0] Processing st: 1, edit list 0 - media time: -1, duration: 63792

(63792 samples)/(48000 samples/sec) = 1329 ms.

So, when SE gets the audio, timestamps, and shot changes, '-ignore_editlist 0' is apparently in effect; ignore_editlist=0 is the default.

But when SE draws the waveforms, '-ignore_editlist 1' is apparently in effect.

Is there anything is SE's settings that controls this?

Now, understand that for these parts, there should be _no_ editlists, but FFmpeg has a bug. What I'm trying to do here on the SE forum is devise a workaround to get SE to '-ignore_editlist 1' for all operations.

Is there a way?

Thanks, and I think I'll have a drink now.

PS: Actually, '-ignore_editlist 0' for all operations would be okay, too.

If it's wrong, and it's bound to be wrong because of the FFmpeg bug, video that plays fine with MPV or VLC will have both audio and subtitles out of sync when played by most TV decoders. Whatever the case may be, when FFmpeg is fixed, I'll move all the subtitles to their correct times in a single operation using SE's excellent 'Synchronization' utility.

[junah]

Tnx for adding Google Lens under OCR options. It manages lots of languages very well, except it doesnt honor lowercase at the start of the line at all. Italic doesnt work aswell.

[Janusz]

So that everyone understands the issue...

When I play part 1 via 'FFplay -ignore_editlist 1', the audio is approximately 700 ms early; for part 2, it's approximately 1300 ms early.

When I edit part 1 via SE, the waveforms that SE draws are approximately 700 ms early; for part 2, they are approximately 1300 ms early.

Those are not coincidences.
...


I wouldn't worry about what FFplay shows with this or that parameter, as we're not using it in our case. All we need is the DELAY from our video.

Solution/workaround for your problem:

First, a demonstration... (https://www.mediafire.com/file/82q2b5x05ji823w/K-003_%25233.mp4/file)

1. Loading subtitles
2. Loading the original video (no. 1) (the video has a 29 ms audio offset relative to the video, but I'll skip that for now).
Important!!
Now, reset the program by using "Menu\Video\Set video offset..." [Reset]
As you can see, the video and subtitle synchronization is OK.

3. Loading a specially prepared video (no. 2) with audio, to which I've added 700 ms of silence at the beginning of the audio track (the video has a -700 ms audio offset relative to the video).
As you can see, the video and subtitle synchronization is OK.
SubtitleEdit ignored the negative audio offset relative to the video. The generated waveforms are exactly the same as in step 2.

Now for your case.
4. The next video (no. 3) has a 700 ms audio track trimmed at the beginning. To synchronize it with the video, I used a delay of 700 ms. The generated waveforms don't match the video at all. Shifting the subtitles is out of the question, as you can't synchronize the subtitles with both the video and the shifted audio track simultaneously.

Solution!!
Now you need to extract the audio track from your video. You can use ffmpeg, for example.

- In the program, close the video file ("Menu\Video\Close video file")
- Load the newly obtained audio track into the program by dragging it onto the waveform window,
- In the program, set the delay to 700 ms (Menu\Video\Set video offset...)
Syncing only with the audio will do the trick. Finally, you need to add a 700 ms offset for all subtitles so that everything plays correctly when combined with the video.

Edit:
Another solution:
Once you have the extracted audio track, generate an additional, empty audio file (containing only silence) no shorter than 700 ms with the same parameters as your original video file and combine them (silence + main file) without decoding. Add this combined file as a new audio track to your video with "delay=length of the added audio portion in ms" and use this new video to synchronize subtitles by selecting this new track to generate the waveforms in SubtitleEdit.

[markfilipak]

I wouldn't worry about what FFplay shows with this or that parameter, as we're not using it in our case. All we need is the DELAY from our video.

Solution/workaround for your problem:

First, a demonstration... (https://www.mediafire.com/file/82q2b5x05ji823w/K-003_%25233.mp4/file)

Thank you for your work, Janusz. You're very kind.

I downloaded your MP4. It's not what you think it is. The problems are caused by the structure of the MP4, not by SE. I'll write more about that in a bit.

First, your MP4 has an edit list. The source of the following lines is FFprobe's report (level 48), by the [mov,mp4,m4a,3gp,3g2,mj2 @ xxxxxxxxxxxxxxxx] demuxer.

Processing st: 1, edit list 0 - media time: 23956, duration: 4013585
drop a frame at curr_cts: 0 @ 0
drop a frame at curr_cts: 1024 @ 1
drop a frame at curr_cts: 2048 @ 2
drop a frame at curr_cts: 3072 @ 3
drop a frame at curr_cts: 4096 @ 4
drop a frame at curr_cts: 5120 @ 5
drop a frame at curr_cts: 6144 @ 6
drop a frame at curr_cts: 7168 @ 7
drop a frame at curr_cts: 8192 @ 8
drop a frame at curr_cts: 9216 @ 9
drop a frame at curr_cts: 10240 @ 10
drop a frame at curr_cts: 11264 @ 11
drop a frame at curr_cts: 12288 @ 12
drop a frame at curr_cts: 13312 @ 13
drop a frame at curr_cts: 14336 @ 14
drop a frame at curr_cts: 15360 @ 15
drop a frame at curr_cts: 16384 @ 16
drop a frame at curr_cts: 17408 @ 17
drop a frame at curr_cts: 18432 @ 18
drop a frame at curr_cts: 19456 @ 19
drop a frame at curr_cts: 20480 @ 20
drop a frame at curr_cts: 21504 @ 21
skip 1428 audio samples from curr_cts: 22528
That is an edit list. It's the first thing that may cause trouble.

The second thing that does cause trouble is a gap in the audio timestamps at the beginning of the MP4.

I've fixed my problem. Yeah! (I think.) The fix was very tedious involving several days of work.

I plan to write a detailed report. Before I do that, I'll finish the subtitles for part 1. Then I'll process part 2 in the same manner. Then I'll splice them together. The running time for parts 1+2 together is 5 hours.

I want to make sure that what I do is understandable and effective. Then I'll write the report, post it here, and then delete most of my previous postings on this subject. It will take me a few days.

By the way, the trouble was not caused by "Menu\Video\Set video offset..." -- I've never used it.

[markfilipak]

I'm still working on this.

Edit lists in MP4 and MKV TSes seems to be one of the best kept secrets in video. I've done a ton of searching and reading and everyone blames MP4 and/or MKV. From what I've read, no one seems to know about edit lists. Let me give you a simple, easy-to-understand example using something you probably already do know about: soft telecine. After that I'll tie this subject into SE.

Telecine takes runs of four frames and makes runs of five frames in order to play on legacy televisions. Stated another way, telecine converts eight fields played at 23.976 fps to ten fields played at 29.970 fps. It does it by duplicating two of the fields. Usually, [Aa][Bb][Cc][Dd] on disc becomes [Aa][Bb][Bc][Cd][Dd] sent to the TV. Note that the process is technically called 2-3 pulldown, but everyone calls it telecine because the machines that did it were called telecines.

Around the year 2000, DVD makers realized that putting ten fields on disc to get 29.970 fps was wasting 20% of the potential disc space! What they needed was a way to get the playback unit (the decoder) to accept [Aa][Bb][Cc][Dd], and to then make [Aa][Bb][Bc][Cd][Dd] to feed the TV. MPEG okayed the scheme and thus was born soft telecine.

Making edit lists is like making soft telecine in that it gets decoders to finish the editing, however, unlike soft telecine instructions, edit lists should never appear in final videos. That edit lists _do_ appear in what people _think_ are final videos is their mistake.

So, why do edit lists exist, eh?

Imagine you're an editor who's cutting and splicing video to make a final cut. _Without_ edit lists, you have to wait for hours of re-coding to finish after every edit before you can see the results (unless of course you're editing raw video). _With_ edit lists you see the results almost immediately. You only need to re-code one time, at the end, when you've worked out all the final cuts and splices. During the editing, you can even watch before-&-after, two-window comparisons by telling the before-window to 'ignore_editlist'.

What I am seeing in SE is actual audio, timestamps, shot changes, and subtitles that are all in sync, but audio waveforms that appear to be 'ignore_editlist'. I have already confirmed it because I've succeeded with DVD 1of2.

The project I'm working on is the American release of a five-hour, German movie that spans two DVDs, composed of VOBs that change back and forth between 29.970 and 23.976 fps, and with German-to-English dubbing (so lip-sync is very difficult to discern). I'm working on the five-hour version of "Das Boot".

[TR-9970X]

I'm still working on this.

The project I'm working on is the American release of a five-hour, German movie that spans two DVDs, composed of VOBs that change back and forth between 29.970 and 23.976 fps, and with German-to-English dubbing (so lip-sync is very difficult to discern). I'm working on the five-hour version of "Das Boot".

Hi, I'm curious about this 5 hour version ??

I checked IMDb, and the only thing that comes close is the "mini series", which is split over 3 episodes, but I've seen it available in 6 parts.

The movie is from 1981, and the series is 1985. is that what you're working on ??

But for me, they appear to be in German (obviously), and that's not ideal, so hopefully it has accurate subtitles.

It would be interesting to translate the German audio to English...can SE do that accurately ??

[markfilipak]

Hi, I'm curious about this 5 hour version ??

I checked IMDb, and the only thing that comes close is the "mini series", which is split over 3 episodes, but I've seen it available in 6 parts.

There have been five versions. I have [1981], [1996], and [2004]. I did the summaries below based on what I have, which is complete, plus what I researched. I have Das Boot [2004] on two DVDs. It's those that are driving me nuts. [1981] and [1996] are blurays and were a breeze.

Das Boot [1981][1984][1988][1996][2004]

Das Boot [1981] is the 2-1/2 hour, theatrical release version. It's actually an abbreviation of the miniseries and the resulting plot is mostly event-driven.

Das Boot [1984] is the nearly 5 hour, television miniseries broken into 3, 100-minute episodes prefaced by previous-episode summaries. To the theatrical version's tale, the 5-hour version adds countermeasures employed against attacking destroyers, evasive maneuvers used for escapes, British mistakes, and German luck. It presents more of the day-to-day story: coping with weeks-long periods of stormy weather, maintaining sleep cycles, treating incessant skin ailments, reacting to news from home (e.g., the firebombing of Hamburg & Cologne), suffering through boredom and loneliness, upholding morale (e.g., the captain's personal counseling), etc.

Das Boot [1988] is the same miniseries, but broken into 6, 50-minute episodes prefaced by previous-episode summaries.

Das Boot [1996] is the so-called DIRECTOR'S CUT that includes all of the event-driven plot found in the theatrical release combined with some of the character-driven scenes found in the miniseries. This version was digitally restored, recut with 1 hour of new scenes added, and sonically redesigned with new sound effects mixed into 8 channels of digital surround sound.

Das Boot [2004] is the miniseries except it's continuous, without episode breaks or previous-episode summaries.
The movie is from 1981, and the series is 1985. is that what you're working on ??

But for me, they appear to be in German (obviously), and that's not ideal, so hopefully it has accurate subtitles.

It would be interesting to translate the German audio to English...can SE do that accurately ??
I don't know.

The subtitles are a funny story. The English dubbing is quite good regarding lip-sync, but the subtitles don't match the dubbing. You see, the subtitles are translations from German, and they're quite raunchy. The dub is often so different from the translated subtitles that you'd think they're from different movies. I like the subtitles because they illustrate just how prudish Americans are compared to Europeans.

I'm giving up on this project for a while.

I'm beginning to lean heavily in the direction of bugs in SE rather than MP4 edit lists as the source of problems. For example, I went to great effort to avoid MP4 and MKV in favor of M2TS and others (but there's not many others). When presented with AC3-in-M2TS, SE waveform behavior went wildly erratic: jumping around, stopping, skipping. I'd like to concatenate the VOBs via FFmpeg but FFmpeg just has too many bugs and the VOBs require too much clean-up. I tried, but have not succeeded. I've spent weeks on it to the exclusion of my actual life.

This has all been heartbreaking because I dearly love the 5-hour version, but it's on DVD and inaccessible because of FFmpeg and SE bugs.

[TR-9970X]

There have been five versions. I have [1981], [1996], and [2004]. I did the summaries below based on what I have, which is complete, plus what I researched. I have Das Boot [2004] on two DVDs. It's those that are driving me nuts. [1981] and [1996] are blurays and were a breeze.

Das Boot [1981][1984][1988][1996][2004]

Das Boot [1981] is the 2-1/2 hour, theatrical release version. It's actually an abbreviation of the miniseries and the resulting plot is mostly event-driven.

Das Boot [1984] is the nearly 5 hour, television miniseries broken into 3, 100-minute episodes prefaced by previous-episode summaries. To the theatrical version's tale, the 5-hour version adds countermeasures employed against attacking destroyers, evasive maneuvers used for escapes, British mistakes, and German luck. It presents more of the day-to-day story: coping with weeks-long periods of stormy weather, maintaining sleep cycles, treating incessant skin ailments, reacting to news from home (e.g., the firebombing of Hamburg & Cologne), suffering through boredom and loneliness, upholding morale (e.g., the captain's personal counseling), etc.

Das Boot [1988] is the same miniseries, but broken into 6, 50-minute episodes prefaced by previous-episode summaries.

Das Boot [1996] is the so-called DIRECTOR'S CUT that includes all of the event-driven plot found in the theatrical release combined with some of the character-driven scenes found in the miniseries. This version was digitally restored, recut with 1 hour of new scenes added, and sonically redesigned with new sound effects mixed into 8 channels of digital surround sound.

Das Boot [2004] is the miniseries except it's continuous, without episode breaks or previous-episode summaries.

I don't know.

The subtitles are a funny story. The English dubbing is quite good regarding lip-sync, but the subtitles don't match the dubbing. You see, the subtitles are translations from German, and they're quite raunchy. The dub is often so different from the translated subtitles that you'd think they're from different movies. I like the subtitles because they illustrate just how prudish Americans are compared to Europeans.

I'm giving up on this project for a while.

I'm beginning to lean heavily in the direction of bugs in SE rather than MP4 edit lists as the source of problems. For example, I went to great effort to avoid MP4 and MKV in favor of M2TS and others (but there's not many others). When presented with AC3-in-M2TS, SE waveform behavior went wildly erratic: jumping around, stopping, skipping. I'd like to concatenate the VOBs via FFmpeg but FFmpeg just has too many bugs and the VOBs require too much clean-up. I tried, but have not succeeded. I've spent weeks on it to the exclusion of my actual life.

I didn't realize there were so many variants...

I think I have a couple of different revisions of the movie, but until your previous post, I wasn't aware of a L-O-N-G mini series.

So I have downloaded the 6 part version, which looks pretty good video wise, the audio is German, but I have good (I hope) English subs.

So I'm going to edit each part, then join them altogether, should work out OK.

It's 1080p x265, with 6 ch E-AC3 audio.

It would be great to find an app that could translate the German to English, tho.

Cheers.

[markfilipak]

It would be great to find an app that could translate the German to English, tho.
Since you already have English subtitles, I assume you mean translating German speech to English speech. I think you would not like the result. Lip and mouth movements would look quite silly.

Dubbing into English is the art of substituting English words that match German lip and mouth movements while still maintaining the thread of conversations. Dubbing is not translation.

Edit: Have you seen the original Godzilla where they did exactly that: Translate Japanese dialog directly into English? Audiences laughed. Godzilla was a serious movie about the dangers of nuclear testing that became a sort of cult comedy.

[TR-9970X]

Since you already have English subtitles, I assume you mean translating German speech to English speech. I think you would not like the result. Lip and mouth movements would look quite silly.

Dubbing into English is the art of substituting English words that match German lip and mouth movements while still maintaining the thread of conversations. Dubbing is not translation.

Edit: Have you seen the original Godzilla where they did exactly that: Translate Japanese dialog directly into English? Audiences laughed. Godzilla was a serious movie about the dangers of nuclear testing that became a sort of cult comedy.

As mentioned in the PM's, I don't have to worry about translating, just sync & length.

But I have noticed that the subtitles don't match word for word with the dubbed English audio, I might try running the English audio thru SE using Whisper to create hopefully more accurate subs.

[markfilipak]

As mentioned in the PM's, I don't have to worry about translating, just sync & length.
Sync is not a problem.
But I have noticed that the subtitles don't match word for word with the dubbed English audio, I might try running the English audio thru SE using Whisper to create hopefully more accurate subs.
The subs are translations from the German dialog. The dubs are created dialog (art, as I tried to explain) so that lip and mouth movements don't promote laughter but the conversations still maintain the story. That's always true of dubs since the Godzilla fiasco.

I didn't 'fix' the subtitles because I think it's interesting how raunchy the actual German was. Americans tend to be prudish and the raunchy German shows how Europeans are different. ...Just makes for an extra study in culture along with the story. The English dubbing is so good, so clear, you don't really need subs.

[TR-9970X]

Sync is not a problem.

The subs are translations from the German dialog. The dubs are created dialog (art, as I tried to explain) so that lip and mouth movements don't promote laughter but the conversations still maintain the story. That's always true of dubs since the Godzilla fiasco.

I didn't 'fix' the subtitles because I think it's interesting how raunchy the actual German was. Americans tend to be prudish and the raunchy German shows how Europeans are different. ...Just makes for an extra study in culture along with the story. The English dubbing is so good, so clear, you don't really need subs.

I agree that the dubbing would be good, especially with Das Boot, and therein lies the problem with using the subs, in this case, as they are not accurate.

But it would still be interesting to create subs that DO match the dubbed English, wouldn't it ??

Have you used Whisper in SE to create subs for a video that hasn't got any ??

It takes quite a long time, depending on the used settings & libraries, but it turns out pretty damn good, I've used it a couple of times.

[markfilipak]

Have you used Whisper in SE to create subs for a video that hasn't got any ??
Nope.
It takes quite a long time, depending on the used settings & libraries, but it turns out pretty damn good, I've used it a couple of times.
Thanks for the tip. I'll try Whisper when the need arises.

[jay123210599]

Can anyone help me with this?

Now I see what my problem is. The shortcuts I set up for moving videos frame by frame work for this menu (https://imgbox.com/WcMhGnID), but not for this menu (https://imgbox.com/VNpn3u3V). But why is that, and what should I do to fix it?

[Janusz]

Can anyone help me with this?

Now I see what my problem is. The shortcuts I set up for moving videos frame by frame work for this menu (https://imgbox.com/WcMhGnID), but not for this menu (https://imgbox.com/VNpn3u3V). But why is that, and what should I do to fix it?


Let me put it this way... Something hasn't been fully implemented in the keyboard shortcuts in SubtitleEdit and its window for importing *.sub subtitles for editing. A workaround is to use [Video engine]=MPC-MC as the "Video player" and use the video player's shortcuts. For example, CTRL+Right Arrow moves the video forward by one frame, and CTRL+Left Arrow moves the video back by one frame. The spacebar starts video playback. These are the default settings for this video player, and you don't need to change anything. For this to work, in both cases, you need to click on the video window to make it active.

You can delete your settings in SubtitleEdit's "Settings/Shortcuts/Video" menu.

[jay123210599]

Let me put it this way... Something hasn't been fully implemented in the keyboard shortcuts in SubtitleEdit and its window for importing *.sub subtitles for editing. A workaround is to use [Video engine]=MPC-MC as the "Video player" and use the video player's shortcuts. For example, CTRL+Right Arrow moves the video forward by one frame, and CTRL+Left Arrow moves the video back by one frame. The spacebar starts video playback. These are the default settings for this video player, and you don't need to change anything. For this to work, in both cases, you need to click on the video window to make it active.

You can delete your settings in SubtitleEdit's "Settings/Shortcuts/Video" menu.

It worked, but whenever I move a frame forward or backward, this (https://imgbox.com/QcWaPElr) doesn't move along with the video. It always remained the same. How do I fix that?

[hidef_rec]

Is there a way to save a .sup (PGS) as a .sup after removing a few lines of the subtitle?

[Janusz]

@ hidef_rec
Yes. In the main program window, select File/Export/Bluray sup...

[Music Fan]

Or drag & drop the sup on SE main's window and the Import/OCR window will open.
There you can remove lines.
Then right click on the text, export, Blu-ray sup.

[Nikse555]

Is there a way to save a .sup (PGS) as a .sup after removing a few lines of the subtitle?

Or use File - Import - Blu-ray (.sup) subtitle file for edit...

Here you can delete/edit/adjust lines

[TR-9970X]

I have been using Whisper to transcode some videos that haven't got subtitles, and I'm using just the CPP with the large V2 model, it takes about 7 hours to do a 50-minute clip, and the first one I did was really good and accurate, the next one I did, far from it.

Is there a better, faster way or can I use whisper outside of SE ??

Cheers.

[Emulgator]

Yes and yes.
VoodooFX has built Purfview's Faster-Whisper-XXL which I use most of the time as the quick first approach for English,
then refine with additional passes for CPP and models v1, maybe v2, v3 missing more than hitting.
plus there are 4 more working Whisper implementations within SE, some running beautifully quick on GPU
(plus the last 2 I could not get to run), plus multiple Whisper standalones.

P.S. LARGE Models, that is, of course.

[TR-9970X]

Yes and yes.
VoodooFX has built Purfview's Faster-Whisper-XXL which I use most of the time as the quick first approach for English,
then refine with additional passes for CPP and models v1, maybe v2, v3 missing more than hitting.
plus there are 4 more working Whisper implementations within SE, some running beautilfully quick on GPU
(plus the last 2 I could not get to run), plus multiple Whisper standalones.

You're busy answering 2 threads :thanks:

OK, I just tried the Open AI model/engine with a medium model, didn't do a really good job.

Tried Const-me, which uses the GPU, so it's fast, but I only tried a small model, again, not too good.

Will try your suggestion, installing as we speak.

And I'm also trying a standalone, quite a lot to install for Whisper AI :(

So what model do you use for Purfview ??. and then run it thru again using CPP??


EDIT :- Just did a Purview run with medium model, started out in sync, then completely lost the plot :(

[VoodooFX]

EDIT :- Just did a Purview run with medium model, started out in sync, then completely lost the plot :(

What do you mean by "lost the plot"? You want to use large-v2 model for accuracy.
What language is that audio, can you upload audio to wetransfer?

[TR-9970X]

What do you mean by "lost the plot"? You want to use large-v2 model for accuracy.
What language is that audio, can you upload audio to wetransfer?

Meaning that it went way out of sync, that's all.

So recommending large v2, why not v3, and what's large turbo v3 ??

It's all English.

What is the point of just uploading the audio, when you've no video to check it against ??

[VoodooFX]

Meaning that it went way out of sync, that's all.

So recommending large v2, why not v3, and what's large turbo v3 ??

It's all English.

What is the point of just uploading the audio, when you've no video to check it against ??

large-v2 is best for English.
Turbo model is faster version of large-v3, but it's a bit less accurate.

Video is not needed at all to check audio/subtitles. Remux the file with MKVToolNix deselecting everything except audio.

Can you share srt too (where it's out of sync)?

[TR-9970X]

large-v2 is best for English.
Turbo model is faster version of large-v3, but it's a bit less accurate.

Video is not needed at all to check audio/subtitles. Remux the file with MKVToolNix deselecting everything except audio.

Can you share srt too (where it's out of sync)?

OK, I will download & try large v2, and if there's still issues, I will upload both.

Cheers & thanks.

EDIT :-

OK, I think there was an issue with the previous files I was using, when I demuxed it with MKVToolNix, it has some sync issues, already :(

So I used another old movie with no subs, and ran that thru with large v2. and it did it rather quickly, however, there are still a lot of sync issues, as you will probably find for yourself.

Here's the link to the audio & sub files, as requested.

https://www.mediafire.com/file/zo66yrozf4yq1xx/audio-subs.7z/file

[VoodooFX]

This tarzan audio is out of sync too. Your subtitles fits OK on waveform.
When you decode your audio it's shorter by 6 seconds.

[TR-9970X]

This tarzan audio is out of sync too. Your subtitles fits OK on waveform.
When you decode your audio it's shorter by 6 seconds.

Hi, well that's not good news, not good at all :(

Do you have any suggestions on what I can do ??

Is there a particular way to demux/decode so there is no sync issues, or is that just how it is ??

Can SE or other app re-sync everything ??

Thanks for your help :)

EDIT :-

I just ran a very different video thru SE, and using the large v2 model, it transcribed the 50 minute video very quickly.

However, on play back, there were a LOT of lines missed, that should have been recognised, but at least it was pretty much synced, all the way thru.

Any suggestions ??

[VoodooFX]

Do you have any suggestions on what I can do ??


Try "--ff_sync" argument, if you run from SE then be sure that SE passes the original file.

Here is tarzan.srt: https://pastebin.com/D8RBuhUX [Produced with Faster-Whisper-XXL Pro, so it will be a bit more accurate]


I just ran a very different video thru SE, and using the large v2 model, it transcribed the 50 minute video very quickly.

However, on play back, there were a LOT of lines missed, that should have been recognised, but at least it was pretty much synced, all the way thru.

Any suggestions ??
Share the audio, the command used and timings where something is missing.

[TR-9970X]

Try "--ff_sync" argument, if you run from SE then be sure that SE passes the original file.

Here is tarzan.srt: https://pastebin.com/D8RBuhUX [Produced with Faster-Whisper-XXL Pro, so it will be a bit more accurate]


Share the audio, the command used and timings where something is missing.

Hi VoodooFX, I tried nearly everything in SE yesterday, and the results differ so much, but ALL still have some issues of sync, or missed lines or even added lines that are not in the audio :(

One annoying thing, if I try and load the Waveform, it throws up a whole list of errors, and it can't be used :(

So I've just about run out of options to get these movies I would like subtitles to.

I will try a few more things today, including your suggestion.

Will keep you updated.

Also, thanks for the "tarzan" subs :)

Cheers.

EDIT :- Have tried several more settings & models, it's just not working for me...well not as good as it should or could.

Maybe I'm expecting too much :(

[VoodooFX]

Hi VoodooFX, I tried nearly everything in SE yesterday, and the results differ so much, but ALL still have some issues of sync, or missed lines or even added lines that are not in the audio :(

So I've just about run out of options to get these movies I would like subtitles to.


Where do you get these out of sync files, from youtube?

Here is exact command that I've used:

faster-whisper-xxl.exe tarzan.mka -l en -m large-v2 --vad_method pyannote_v3 -o source --standard --max_gap 1 --ff_sync -hst 2 -ct float16 --realign


Remove "--realign" if you don't use Pro version, and remove "-ct float16" if you don't have GPU with CUDA. Remove "--ff_sync" if a file doesn't have sync problems.

Whisper model is non-deterministic by default, you can get different results every run, especially on this audio where monkeys are screaming and Aborigines are mumbling, model will increase temperature and will hallucinate on those parts.



One annoying thing, if I try and load the Waveform, it throws up a whole list of errors, and it can't be used :(


I didn't have any problems getting waveform from that mka file in SE, but it's useless because it's out of sync with the original audio.
And if you use "--ff_sync" then srt will be out of sync to the waveform, but it will be in sync with the actual video file.


EDIT :- Have tried several more settings & models, it's just not working for me...well not as good as it should or could.


Just share an audio with a problem and I can check it.

[TR-9970X]

Where do you get these out of sync files, from youtube?

No, they are "proper" downloaded files, rar sets, mainly.

Here is exact command that I've used:

faster-whisper-xxl.exe tarzan.mka -l en -m large-v2 --vad_method pyannote_v3 -o source --standard --max_gap 1 --ff_sync -hst 2 -ct float16 --realign


Can I add that code somewhere in SE ??, or do you use a cmd line process ??

Remove "--realign" if you don't use Pro version, and remove "-ct float16" if you don't have GPU with CUDA. Remove "--ff_sync" if a file doesn't have sync problems.

Don't have Pro (that's contribution version, isn't it??)

I have been using the PC with a 4080 Super.

Whisper model is non-deterministic by default, you can get different results every run, especially on this audio where monkeys are screaming and Aborigines are mumbling, model will increase temperature and will hallucinate on those parts.

Yes, I certainly seem to get different results :(


I didn't have any problems getting waveform from that mka file in SE, but it's useless because it's out of sync with the original audio.

I'm still curious why the wave form errors out.

Just share an audio with a problem and I can check it.

I will have some more attempts, but I've just about had enough for the time being :(

I REALLY appreciate your ongoing help.

I've been using SE for years, but only basically, it wasn't until I started asking certain question a couple of weeks ago that I have found just how much more SE can do, and it's SO much better than what I used to do.

[VoodooFX]

Can I add that code somewhere in SE ??, or do you use a cmd line process ??

In SE you can add the commands in the Advanced field there (but you need to understand what you are doing, because SE is passing some commands on its own too).
I always use it directly in console/terminal. You can check these links for fancy usage:
"One Click Transcribe" tool (https://github.com/Purfview/whisper-standalone-win/discussions/337)
Context menu tool (https://github.com/Purfview/whisper-standalone-win/discussions/539)


Don't have Pro (that's contribution version, isn't it??)

Yeah, "Pro (https://github.com/Purfview/whisper-standalone-win/discussions/456)" is not free version.

I'm still curious why the wave form errors out.

Maybe something with ffmpeg version. I use 7.1.

[TR-9970X]

In SE you can add the commands in the Advanced field there (but you need to understand what you are doing, because SE is passing some commands on its own too).
I always use it directly with console. You can check these links for fancy usage:
"One Click Transcribe" tool (https://github.com/Purfview/whisper-standalone-win/discussions/337)
Context menu tool (https://github.com/Purfview/whisper-standalone-win/discussions/539)

Well, I don't really understand what I'm doing :(
But I know where the Advanced commands go.
I'll just copy most of your line, and see what happens.


Yeah, "Pro" is not free version.

Is it a one-off payment, or subscription ??


Maybe something with ffmpeg version. I use 7.1.

Using 8.0.1

Thanks again, I will see if I can educate myself to do it differently.

[VoodooFX]

I know where the Advanced commands go.
I'll just copy most of your line, and see what happens.

Just don't add these to SE: tarzan.mka -l en -m large-v2 -o source.
And untick all postprocessing boxes there.

BTW, you want to check the whisper log file in SE, to be sure that it's using the original file as input and not a temp wav. In some cases SE can use temp wav, that's no bueno for "--ff_..." args, as those need original file input.

Is it a one-off payment, or subscription ??

It's one-off, but I wouldn't mind a subscription too. :D

[TR-9970X]

Just don't add these to SE: tarzan.mka -l en -m large-v2 -o source.
And untick all postprocessing boxes there.

BTW, you want to check the whisper log file in SE, to be sure that it's using the original file as input and not a temp wav. In some cases SE can use temp wav, that's no bueno for "--ff_..." args, as those need original file input.



It's one-off, but I wouldn't mind a subscription too. :D

Curiosity got the better of me, so I've run a couple of transcripts using your suggested command line, unchecked the post processing options, and both subs turned out pretty damn good, the sync seems very very close,
there are a couple of words that aren't where they should be, but one thing I noticed, it didn't pick up on any sort of background quiet audio (speech), only the main audio.

Are there any other commands that might help with that ??

I know you recommend large v2, but should I give v3 a try (not the turbo).

[VoodooFX]

...one thing I noticed, it didn't pick up on any sort of background quiet audio (speech), only the main audio.

Are there any other commands that might help with that ??

Maybe, dunno without the audio.

I know you recommend large v2, but should I give v3 a try (not the turbo).

Try it if you want. I wouldn't.

[TR-9970X]

Maybe, dunno without the audio.



Try it if you want. I wouldn't.

I will have another session tomorrow, and see how I go, probably won't get the v3 (but you never know).

I'll just have to send you the audio track, and the best .srt that I can produce.

Currently the audio is .flac, but that shouldn't be an issue, I can convert it to .ac3 if you think that might make a difference.

Cheers.

[VoodooFX]

I'll just have to send you the audio track, and the best .srt that I can produce.

Currently the audio is .flac, but that shouldn't be an issue, I can convert it to .ac3 if you think that might make a difference.

I don't need srt, just write the times where something is missing.
No need to touch audio.

[TR-9970X]

I don't need srt, just write the times where something is missing.
No need to touch audio.

Hello again,

OK, done more testing, now it's over to you to see if anything else can be done.

I have added the audio, a couple of .srt's (they're small for comparison), and the error msg that Waveform pops up...

And a read me.txt, so please read thru that, saves me adding it here.

https://www.mediafire.com/file/yahfak2vt9t0g6d/New+folder.7z/file

[VoodooFX]

https://www.mediafire.com/file/yahfak2vt9t0g6d/New+folder.7z/file

Those missing phrases/words were not identified as voice by vad, so there was no attempt to transcribe those areas.

Here is srt produced with Pro version: https://pastecode.io/s/nds33xaw


faster-whisper-xxl.exe boot.flac -l en -m large-v2 --vad_method pyannote_v3 -o source --standard --max_gap 1 -hst 2 -ct float16 --ff_vocal_extract mb-roformer --realign




BTW, that CPP srt is weird, 00:00:00-00:00:30 is not in audio, timestamps are off by up to dozens seconds.

[TR-9970X]

Those missing phrases/words were not identified as voice by vad, so there was no attempt to transcribe those areas.

That's interesting, I'm sure that other models did pickup some of that "missing" audio/text.

I will compare them tomorrow.

Here is srt produced with Pro version: https://pastecode.io/s/nds33xaw


faster-whisper-xxl.exe boot.flac -l en -m large-v2 --vad_method pyannote_v3 -o source --standard --max_gap 1 -hst 2 -ct float16 --ff_vocal_extract mb-roformer --realign


I should be able to use some of this command line in non Pro, there's only a little bit more than the previous.

BTW, that CPP srt is weird, 00:00:00-00:00:30 is not in audio, timestamps are off by up to dozens seconds.

Well, I'm glad that I sent them too :)

From the quick look I've had of the subs you sent, it looks like it grabbed a LOT of extra dialogue :)

BTW, did you bother to check if my files opened Waveform ??

[VoodooFX]

That's interesting, I'm sure that other models did pickup some of that "missing" audio/text.

Not "did pickup", as they don't try to pickup anything. They never checked that it's actually a speech what they are trying to transcribe. That's why you get there those hallucinations when there is no speech at all.


I should be able to use some of this command line in non Pro

You wont be able to use that as those are only Pro features.

From the quick look I've had of the subs you sent, it looks like it grabbed a LOT of extra dialogue

Yes, because cpp is weird, instead of transcribing stuff, cpp outputs "(speaking in foreign language)" bs.

BTW, did you bother to check if my files opened Waveform ??

Yes, opened after I named your files properly.

[TR-9970X]

Not "did pickup", as they don't try to pickup anything. They never checked that it's actually a speech what they are trying to transcribe. That's why you get there those hallucinations when there is no speech at all.




You wont be able to use that as those are only Pro features.

I used this :- --vad_method pyannote_v3 -o source --standard --max_gap 1 --ff_sync -hst 2 -ct float16

and all I can see that's different is this :---ff_vocal_extract mb-roformer

Yes, because cpp is weird, instead of transcribing stuff, cpp outputs "(speaking in foreign language)" bs.



Yes, opened after I named your files properly.

OK, so what did you rename them to, so I can check if it's my different version of FFMPEG that is the problem. I'm guessin' just a really simple name, then.

So one question about Pro version, is it a standalone command line, or can it be used within SE ??

Thanks

[VoodooFX]

OK, so what did you rename them to, so I can check if it's my different version of FFMPEG that is the problem. I'm guessin' just a really simple name, then.

To the same name as audio. BTW, I use older SE version - 4.0.12.


So one question about Pro version, is it a standalone command line, or can it be used within SE ??

Yes to both.

[markfilipak]

Above this message there are a lot of pleas for help regarding audio sync and loss of sync and clipped audio. I'm sorry that I don't have time to respond but I hope what I write here will help.

ISOM containers like MP4 and MOV can contain edit lists. I believe that's also true of MKV but I don't use MKV. If you edit audio without recoding it, that will likely result in an edit list being generated and attached to the AV. If you then audition the AV without recoding the audio, and if there is an edit list, then the player/utility may or may not appear to have synchronized audio depending upon whether it honors edit lists and how well it honors edit lists. You can detect and view edit lists via FFprobe, but it's complicated.

SE's waveform rendering -- specifically, the rendered pictures of the timing -- seems to have some issues with edit lists. What I do is recode the audio in an MP4 so that there's no edit lists. The audio must be cleaned of edit lists. For example, don't just 'FFmpeg -ss...' or 'FFmpeg -to...' or 'FFmpeg --itsoffset...' or 'FFmpeg -itsscale:a...' without recoding the audio -- for the final AV and any temporary AV that you're going to use in SE.

If I'm going to be working on the audio extensively, I recode it to frac so that it's a separate .FRAC stream and then edit to my heart's content. If I need to see it in an AV, I make a temporary AV just to do the viewing. Then, in the final MP4, I recode the frac to whatever audio I want. By keeping the frac audio separate, it stays pristine and the final MP4 has no edit lists. All that takes planning and it is tedious to make temporary AVs just to see some result I need to see, but it's worth it in the end.

All that I've written above is in addition to some SE waveform rendering issues that appear to be unrelated to edit lists. I'm still trying to figure it out.

I've written extensively about edit lists during the past month and my explorations are here in the SE forum. I promised to clean up my documentation and I reiterate that promise. But that has to wait until the time when I've figured it all out.

I hope my experiences help you. In the mean time I'm positioning subtitles based on the audio that MPV plays, not on the positions of the waveforms that SE renders.

[jay123210599]

Let me put it this way... Something hasn't been fully implemented in the keyboard shortcuts in SubtitleEdit and its window for importing *.sub subtitles for editing. A workaround is to use [Video engine]=MPC-MC as the "Video player" and use the video player's shortcuts. For example, CTRL+Right Arrow moves the video forward by one frame, and CTRL+Left Arrow moves the video back by one frame. The spacebar starts video playback. These are the default settings for this video player, and you don't need to change anything. For this to work, in both cases, you need to click on the video window to make it active.

You can delete your settings in SubtitleEdit's "Settings/Shortcuts/Video" menu.

It worked, but whenever I move a frame forward or backward, this (https://imgbox.com/QcWaPElr) doesn't move along with the video. It always remained the same. How do I fix that?

@Janusz Did you get this? How do I solve this problem?

[TR-9970X]

@ VoodooFX,

Me again, I have a very short clip (35 seconds) that is in German or French, and it's singing (sort of), and I just can't get all the words.

Any suggestions ??, could I maybe send it to you, to have a play ??

Regards

[VoodooFX]

could I maybe send it to you, to have a play ??


Sure.
To stop offtopic, continue in the right thread: https://forum.doom9.org/showthread.php?t=184859

[TR-9970X]

Sure.
To stop offtopic, continue in the right thread: https://forum.doom9.org/showthread.php?t=184859

Sure, no worries, I was not aware of "your" thread.

Thanks, will send it off shortly.

[markfilipak]

I have some solid information.

I found a situation that 'shows' speech (waveform) on one side of a scene change but 'says' speech (MPV) on the other side of the scene change.

The hh:mm:ss.mmm of the scene change is 100% correct.

If I put the subtitle where the speech is shown (waveform), it's wrong (external MPV). If I put the subtitle where the speech is said (internal MPV), it's right (external MPV).

Background:
- Waveform was extracted and drawn by internal FFmpeg.
- What waveform shows gets earlier and earlier.
- I had captured the audio via MKVToolNix.
- I had to use -itsscale:a in (external) FFmpeg to sync the audio.
- There is no edit list in the resulting audio stream.

_No_ edit list, yet waveform is being drawn by internal FFmpeg as though there _is_ an edit list that internal FFmpeg is _ignoring_. That would mean that there actually is an edit list which internal MPV, external MPV, and external FFmpeg are all honoring but internal FFmpeg (waveform) is not.

But FFreport shows no edit list. I am totally baffled. The 'error' in waveform is too specific and too accurate and too persistent (video-to-video-to-video) to be accidental. Yet FFreport shows _no_ edit list. ...Is there an edit list or isn't there an edit list? Hell, I don't know anymore.

I need someone to help, to work with me. Nik is busy re-inventing the wheel.

[Nikse555]

Subtitle Edit 5 beta-1 is now out :)
https://github.com/SubtitleEdit/subtitleedit/releases
If you're on mac, be sure to follow the installation instructions: https://github.com/SubtitleEdit/subtitleedit?tab=readme-ov-file#macos

SE 4 was getting increasingly hard to maintain — especially on Linux. SE 5 has rewritten UI in Avalonia UI and works much better on Linux now. Works on mac as well (cross platform with single codebase).
Note that SE 4 will still be maintained for a while, and you can have both SE 4 and 5 installed (use portable versions).

Please give the beta a spin and let me know if you run into trouble.
Note: SE 5 requires Win10 or newer.

Yeah, re-inventing the wheel surely takes a lot of time ;)

[jay123210599]

Subtitle Edit 5 beta-1 is now out :)
https://github.com/SubtitleEdit/subtitleedit/releases
If you're on mac, be sure to follow the installation instructions: https://github.com/SubtitleEdit/subtitleedit?tab=readme-ov-file#macos

SE 4 was getting increasingly hard to maintain — especially on Linux. SE 5 has rewritten UI in Avalonia UI and works much better on Linux now. Works on mac as well (cross platform with single codebase).
Note that SE 4 will still be maintained for a while, and you can have both SE 4 and 5 installed (use portable versions).

Please give the beta a spin and let me know if you run into trouble.
Note: SE 5 requires Win10 or newer.

Yeah, re-inventing the wheel surely takes a lot of time ;)

Can it solve this problem?

Let me put it this way... Something hasn't been fully implemented in the keyboard shortcuts in SubtitleEdit and its window for importing *.sub subtitles for editing. A workaround is to use [Video engine]=MPC-MC as the "Video player" and use the video player's shortcuts. For example, CTRL+Right Arrow moves the video forward by one frame, and CTRL+Left Arrow moves the video back by one frame. The spacebar starts video playback. These are the default settings for this video player, and you don't need to change anything. For this to work, in both cases, you need to click on the video window to make it active.

You can delete your settings in SubtitleEdit's "Settings/Shortcuts/Video" menu.

It worked, but whenever I move a frame forward or backward, this (https://imgbox.com/QcWaPElr) doesn't move along with the video. It always remained the same. How do I fix that?

[Nikse555]

Can it solve this problem?

I'm not really sure what you mean, but the window with edit of image based subtitles does support the video movement shortcuts of the main window (in SE 5 beta 3+).

[jay123210599]

I'm not really sure what you mean, but the window with edit of image based subtitles does support the video movement shortcuts of the main window (in SE 5 beta 3+).

The beta does move the video and time when doing frame-by-frame when editing PGS subtitles, but it doesn't move the subtitles alongside them when I selected an image. It also doesn't show the end times for the subtitles and the PGS subtitle edit menu doesn't show the video in this format: 00:00:00:00

[tormento]

What is the most accurate OCR now, able to recognize italic, subtitle position and so?

I am stuck to binary, as it gives me the most predictable results but any suggestion is welcome.

[Nikse555]

What is the most accurate OCR now, able to recognize italic, subtitle position and so?

I think you're still stuck with BinaryOCR, nOCR, or Tesseract with old engine.

The AI models are certainly getting better, but none I have seen can detect italic yet - if you have a nvidia gfx card then running OCR via Ollama and glm-ocr is pretty good/fast:
https://ollama.com/library/glm-ocr

[TR-9970X]

I'm currently using Beta 5, and overall, I like all the changes, however, in "Multiple Replace", I think the previous 4.0.15 version is easier to work with.
With Beta 5, once you create a new category, it's a little bit more of a procedure to add a new rule, as you have to locate the little button with the 3 dots, and then click to show the menu, but depending on the screen layout, that button can be very hard to find, and I seem to be having to change the size & positions of other windows I have open at the time.
It would be so much easier & quicker just to be able to right button click anywhere under the main category, to add a new rule.
Also, if you have multiple categories, I've found that more often than not they ALL expand, before you get a chance to expand the one you want, so then you have to back track, and close them all.
So I've exported the other categories, and then deleted them from that list, I'll simply import them, when I need other categories.

So with the Github listing of available beta's, are there any changes, or are they simply just generated "clone's" ?

[Nikse555]

SE 5 Beta 9 is out now: https://github.com/SubtitleEdit/subtitleedit/releases

Multiple replace now has expand (+) and collapse (-) buttons, and remembers which categories were expanded.
Added Ctrl+N shortcut for new rule + Ctrl+D for duplicate rule.
Also added some coloring and info about what rules are involved in a subtitle line fix.

Change log is here: https://github.com/SubtitleEdit/subtitleedit/blob/main/ChangeLog.txt

[TR-9970X]

SE 5 Beta 9 is out now: https://github.com/SubtitleEdit/subtitleedit/releases

Multiple replace now has expand (+) and collapse (-) buttons, and remembers which categories was expanded.
Added Ctrl+N shortcut for new rule + Ctrl+D for duplicate rule.
Also added some coloring and info about what rules are involved in a subtitle line fix.

Change log is here: https://github.com/SubtitleEdit/subtitleedit/blob/main/ChangeLog.txt

Sounds good, will check it out, thanks :)

EDIT:-

OK, so I have spent the best part of today getting to know Beta 9, and I really appreciate the changes you've made since Beta 5.

Here are a couple of observations & suggestions. :)

#1. With the Category minimized, it's very easy to access the 3 dot button to add a new Category or Rule, however, once you've added a new rule, the whole Category expands, and the 3 dot button sort of goes into hiding.
So it requires the Category to be minimized again, add what you want, rinse, repeat :(

It would be great if the Category stayed minimized, until you need to expand it.

#2. The individual 3 dot buttons get inconveniently hidden under the scroll bar when the Category is expanded, making it difficult to access the 3 dot buttons menu.

The scroll bar seems to have a partial auto hide action, so when you get too close to it with the mouse pointer, it pops out, hence the above comment.

#3. When you have many, many Rules in the Category, and you discover that you need to edit or find a Rule, it's all but impossible to scroll up & down, until you find it. (Hard on the eyes).

Would it be possible to have some kind of search for Rule names/text ?

EDIT #2:-

Later in the day, I found myself having to add some lines to an existing subtitle, using the waveform, and I had been doing this quite a bit with SE 4.0.14, and found it quite easy to navigate the timeline, but with SE 5 B9, it seems to be quite different (to me), I found I had to jump it back by selecting a subtitle from before the section I wanted to edit, from the listing above the timeline window.

Maybe it needs some "obvious" navigation buttons.....:confused:

[TR-9970X]

I just had a strange and annoying thing happen.

I was editing subs with SE 5 B9, and I had a "huge" set of rules in the Multiple Replace, and for whatever reason the PC locked up, when I got back into SE, that whole set of rules (and others) were gone :mad::mad::angry::confused:

[TR-9970X]

Another thing...with SE 5 B9

How long does the "Use Always" option last when doing Spell Check ??

I have been doing a LOT of TV series subs, and the same words keep popping up, and I choose "Use Always", but then those words will show up, again & again...like it's not remembering the choice.

[TR-9970X]

Have been keeping a watchful eye on when B10 was coming out, and now it has :)

Have been using it for about 30 minutes, and it's a lot "faster" to navigate around the Multiple Replace screen, and adding new Rules & Category's....and the "Find" function is great :)

The only thing that would be a little easier is, when/if you have the screen full size, and you're adding rules, I found that the - & + buttons down the bottom left corner would be more convenient if they at the top left corner, closer the other "button's". Just a lot less mousing around, bottom to top.

Nice change since B9 :)

[TR-9970X]

I have a strange issue with the PGS subtitles on the Downton Abbey series, that I got recently.

When OCR'ing them with SE 5 B13, using either nOcr, Tesseract, and I even downloaded and tried PaddleOCR, the results for each are not good, the are many "typo's", grammar errors and extra characters throughout, it would take a long time to manually go through each sub, to correct :(

I could send you one, if you're interested.

UPDATE:- I thought I'd try Google Lens Standalone, and beside it being a little slower, it worked VERY well on the Downton Abbey sub's, but on some other it doesn't play well with ♪..it finds them well enough, but likes to put them somewhere else.

It's reasonably easy to shuffle them into the correct spot, tho.

[checky]

I tried to install version v5.0.0-beta22, but I got a pop-up saying:
Subtitle Edit requires the .NET 10 Runtime

So, what version should I download:
.NET Runtime 10.0.7
or
.NET Desktop Runtime 10.0.7

Here is the link where both runtimes could be downloaded:
https://dotnet.microsoft.com/en-us/download/dotnet/10.0

[TR-9970X]

I tried to install version v5.0.0-beta22, but I got a pop-up saying:


So, what version should I download:
.NET Runtime 10.0.7
or
.NET Desktop Runtime 10.0.7

Here is the link where both runtimes could be downloaded:
https://dotnet.microsoft.com/en-us/download/dotnet/10.0

It states that the Desktop has both. Read!

[GCRaistlin]

Nikse555, please take a look at https://github.com/SubtitleEdit/subtitleedit/issues/9525 and https://github.com/SubtitleEdit/subtitleedit/issues/9324 . These issues got auto-closed because of inactivity.

[markfilipak]

Waveforms are both width- and height-scaled.
Suggestion: Scale just the width; leave the height at 100%.

[DMD]

Good morning

Subtitle: Blu-ray SUP exported from SE 5.1.0.11 rejected by BDSup2Sub

Anyone else seeing this? A Blu-ray SUP exported from Subtitle Edit 5.1.0.11
won't load in BDSup2Sub++ 1.0.3 — it fails immediately with:

Subpicture too large: 4004x2004 at offset 0x00000068

The exact same subtitle exported from SE 4.0.16 (same settings) loads fine.
Looks like the new SUP export writes an image object larger than the video
canvas declared in the file. Export resolution was 3840x2160 (auto-selected
from the loaded 4K video).

I've filed a GitHub issue with the sample files: https://github.com/SubtitleEdit/subtitleedit/issues/12299
Just checking if others can reproduce, especially at non-4K resolutions.

Thank you very much

[Nikse555]

Good morning

Subtitle: Blu-ray SUP exported from SE 5.1.0.11 rejected by BDSup2Sub



Thx for the info :)

Hopefully fixed in SE 5.1.0 beta 13: https://github.com/SubtitleEdit/subtitleedit/releases

[DMD]

Thx for the info :)

Hopefully fixed in SE 5.1.0 beta 13: https://github.com/SubtitleEdit/subtitleedit/releases

Thank you very much for your promptness and attention.

[miket2]

In version 5.2.0 Beta 20, the text in WebVTT is duplicated when the text and subtitle numbers are the same.
In version 4.0.16, with the “Merge lines with same text on load” setting, the duplicate text was correctly removed.

For example:
Original:

WEBVTT
X-TIMESTAMP-MAP=MPEGTS:900000,LOCAL:00:00:00.000

STYLE
::cue(.styledotebebeb) { color:#ebebeb }

14
00:12:58.333 --> 00:13:00.125 align:middle line:85%,start position:50%,middle
<c.styledotebebeb>Do not translate this.</c>

WEBVTT
X-TIMESTAMP-MAP=MPEGTS:900000,LOCAL:00:00:00.000

STYLE
::cue(.styledotebebeb) { color:#ebebeb }

14
00:12:58.333 --> 00:13:00.125 align:middle line:85%,start position:50%,middle
<c.styledotebebeb>Do not translate this.</c>

With version 5.2.0 Beta 20:

...
00:12:58.333 --> 00:13:00.125 align:middle line:85%,start position:50%,middle
<c.styledotebebeb>Do not translate this.</c>
<c.styledotebebeb>Do not translate this.</c>
...

With version 4.0.16:
...

00:12:58.333 --> 00:13:00.125
<c.styledotebebeb>Do not translate this.</c>
...

Could you please fix this?

Thank you very much

[Nikse555]

@miket2: thx, this should be fixed in 5.2.0 Beta 21: https://github.com/SubtitleEdit/subtitleedit/releases
(duplicates are now ignored on load)

[miket2]

@miket2: thx, this should be fixed in 5.2.0 Beta 21: https://github.com/SubtitleEdit/subtitleedit/releases
(duplicates are now ignored on load)

Thank you very much. It works now as expected. :thanks: