DVBSub2Text and French DVB satellite TV subtitles

[naszdom]

Does anyone Know about DVBSub2Text.

Its a marvellous at extracting DVB closed subtitles from Polish satellite TV TS streams and converting them to srt with tessdata.
With French subtitles it recognises subtitles are there but is unable to read them. Videoredo does not recognise them at all. I can extract them project X but then I have to OCR them myself (subtitle edit is useless with satellite DVB subtitles).

Is there any way DVBSub2Text can be made to read them or is there any other similar programme to DVBSub2Text.

I tried looking around to try to contact the maker of DVBSub2Text but there seems no trace.

[Bonnevue]

I wrote a program, a small one, that does the folowing:
it takes audio all streams
it takes video
it takes dvb-subtitles
it converts everything to mpeg2 in a mkv contaner.

You can watch it in vlc, mplayer, mediaplayer classic.

It's very simple to use unpack the zip-file in the same folder as your videocaptures. Type the name of the program ts2tv.exe. thats all, it scans automatcly for all ts files in that folder and converts them. when its done it saves the original files with a bak extension, so you can alsways go back.

If you're interested, let me know on this forum on this thread or you can pm me.

[DjeiPi]

Hello

=>Bonnevue

I wrote a program ...
If you're interested, let me know on this forum on this thread or you can pm me.

Yes, I'm very interested, it's just what I'm searching for, many thanks


=> 2naszdom
DVBSub2Text and French DVB satellite TV subtitles

Does anyone Know about DVBSub2Text.

I just discovered this program, now 0.9.4.0

I used it with France TNT and Astra satellite FTA.

When it works, OCR is more better than SubtitleEdit 3.4.13 for getting out fine SRT.
But :
1) It seems not to see hear impaired stream
2) Sometimes original langage audio stream is correctly detected in left part of the window, but nothing can be read
3) An sometimes, nothing is detected

In all these 3 cases, MPC-HC, VLC, SMplayer always fine display all subtitles streams.

I meet same kind of problems but slightly different with ProjectX :
However, when ProjectX works fine, both hear impaired and normal subtitles streams are extracted as SUP.SUB and SUB and IDX files.
Last can be embedded inside or aside MKV for reading in external readers, both can be OCR converted by SubtileEdit to SRT.

SUB can also be hard written with VLC using MPEG-TS container as output one
Same for SRT using Avidemux, any output container.

Presently, depending on Bonnevue's software, I have to initiate myself with FFMPEG :

http://www.pouchintv.fr/forum/viewtopic.php?f=31&t=1356
https://www.autoitscript.fr/forum/viewtopic.php?f=6&t=14205

[davejavu]

I don't know if this will work for French subtitles as I don't have any, but it works great for my English DVB recordings.

Try using MKVToolNix to repackage your TS file into an MKV container file, then use gMKVExtractGUI or MKVCleaver to extract the subtitle streams only. The DVB subtitles in my recorded programs are recognised by MKVTooNix as SRT subs and if I extract them I get a normal SRT text file.

[DjeiPi]

Hello

Thks, Davejavu, but I've already tried this with MKVTN.
Unfortunatly, when demuxing french (Astra sat/fta * and terrestrial) TS files, MKVTN sees only subtitles files when they come as a pair "file.sub" and "file.idx".

In most cases, ProjectX provides such pairs from TS for both standard subtitles and hear impaired ones, but not always, as I told it above.

* If you are sat equipped, you could try with ArteHD channel

[davejavu]

OK, sorry. I have sat but my dish is pointed at 28.2 so I can't get any French channels to test with unless I move it.

From your info above it looks like your channels broadcast graphic based subtitles but here in the UK they use text based ones instead.

You say 'when demuxing french TS files' that MKVTN only sees idx/sub pairs. Are you demuxing before loading into MKVTN? I just load the original TS file and MKVTN lists all the video/audio/subtitles streams it finds. MKVTN should be able to see any type of subtitle stream in the TS file, text or graphic, if they are there and have the proper stream PID's from the broadcaster. As I said, the UK channels use text subtitles and MKVTN has no problem with these.

[DjeiPi]

Hello

Things are not simple ! Let me roughly explain the french situation :
We have 20+ HD said free"channels".
They are broadcasted as :
- Terrestrial (TNT) recordable and usable in PC as .mts from top box adapters, PC cards. In recent TV sets with PVR, they are often encrypted, forget them for PC reworking.
- Satellite (Astra "TNTsat" duplicated on Eutelsat "Fransat", both are encrypted and need an access card with 4 or unlimimited years life). Records are not usable in PC.
A few channels are FTA duplicated on Astra or Hotbird. A bilangage german + french" ArteHD" channel slightly different from french "ArteHD" is broadcasted as FTA on Astra.
I speak about Arte because its interesting multilangage movies

Subtitles in both transmissions are bitmap (graphic) encoded

- Internet. Generally no subtitles for live broadcast
For replays, we can sometimes get "hardcoded" subtitles versions to choose or .webvtt aside files (to be converted to .srt with SubtitleEdit)

For me, the best tool for investigation is MediaInfo, directly usable in MPC-HC (File > Properties), which, as VLC and SMplayers (an others...) can read directly all recordable files said above.

Now we have to consider a lot of things :

- Several different encoding modes for broadcast that need several different adapted tools for processing raw records
- Available readers : with PC readers, no problem.
With external ones (top boxes, TV set), it's the "inkbottle", as well with raw records as with processed ones
All seem read MKV, some also can read TS, MTS, M2TS depending on some conditions, I believe that associated index files are important.
Concerning audio, some cannot read AC3, some cannot read AAC.
I own an old WD-HD which cannot read AC3 in MKV, but read it in TS...
Concerning subtitles, we have to consider graphic or alphanum files, embedded or in the same folder and also index problems.

My objective would be for archival file the capability to be played on my HD TVset, choosing audio langage, and (not hardcoded) subtitles.
Presently, my best provisional solution for movies is cleaning (begin, ads, end) with Avidemux at keyframes (SmartCutter if I intend to spend time later with subtitles, for use of graphics when DVBsub2text or ProjectX wants to work), saving as TS (with perhaps AAC<=>AC3 conversion for audio streams) and download SRT subtitles from internet (to be adjusted with SubtitleEdit) to be kept aside with the same name in the same folder.
If problems, I remux with MKVTN all what seem to be usefull.

PS : I just did found Bonnevue's program ts2tv here :

http://mediaf-u-an.123website.nl/

I cannot wait for testing it !

PS2 : For french HDTV and .mts recorded files, you have to use Bonnevue's program TSHD2MKV, dnl it here :
https://drive.google.com/drive/folders/0B6QqiXgVfptPdDJ4Z0J6aHdaUWs.
Presently, I get out only a partial working, but my main purpose was obtaining .sub and .idx, this using a tip runs fine on a firsttest movie, even for hear impaired subtitles.
I tried to contact the author

[DjeiPi]

Hello

Here are my first tests with TSH2DMKV.
Probably, I'd better to open a new thread, I'll do it when everything will be tuned fine

I check it on several channels of french TNT.
They generally broadcast two DVB graphic : normal subtitles and ear impaired ones in french langage

I unzipped "TSHD2MKV.zip" in any folder, this one will contain :

ffmpeg.exe
ffprobe.exe
mkvxextract.exe
tshd2mkv.exe

My topbox adapter delivers *.mts files.

If I put directly a .mts file in this folder, and launch tshd2mkv, it doesn't work
By luck(1), I renamed it *.mts.ts rather than *.ts and launch again tshd2mkv.exe

After a while, I got an error message :

Analyzing files and extracting subtitles WORKING---Error in main, line 148: could not open '*.mts.idx': No
such file or directory
---Error: Program stopped due to an error
---Program done, press RETURN--- PENDING


and found in the folder some text files, and :
*.idx
*.sub
*.mts.mkv

Fine, these are one of two tracks VOBSUB subtitles as files (that could also be easily edited as graphicor converted to SRT if interesting ) and as MKV packaged.,
Both can be muxed with *.mts in MKV toolnix.
At this step, when saving, you cut begin, ads, end (found using Avidemux at key frames).

Even if tsh2mkv extracts only one track, it works where ProjectX and/or DVBsub2text output nothing.

More to come, some things about DVB bitmapped subtitles are not clear to me !



(1) If you rename as *.ts, you'll find in the folder only the same file renamed *.ts.bak.
Useless !