Issues creating .scc files from a DVD source that includes XDS

[Rycat]

Hi, I'm a newbie here, I wasn't quite sure where to put this as it is a rather odd problem. Lately I have been capturing TV recordings made on VHS, and I have been trying to work out a way to preserve the closed caption data present on the tape. My main setup was not capable of capturing the line 21 data, (A Sony Digital-8 camcorder with built-in TBC connected to the computer via FireWire) So I got a Magnavox DVD recorder working which captures the EIA-608 stream from the tape and writes it to the DVD. I can view these captions on VLC, and the formatting and positioning are correct. My end goal is to have the captions from the DVD saved to a .scc file, as the positioning information will be preserved.

Here is a screencap of VLC's rendering:

[Attachment 91212 - Click to enlarge]

Anyways, I thought the hard part was over, fired up McPoodle's (very old) SCC_TOOLS and did the following:

1. Follow McPoodle's directons for ripping SCC captions from a DVD
- Run VobSub Ripper Wizard
- run vobsub2scc.exe (McPoodle) on the .cc.raw file that VobSup Ripper Wizard created.
- I now have a .scc file (yay)

2. Verified that the .scc file I have wasn't just garbage
- Ran ccasdi.exe (McPoodle) on the .scc file to create a human-readable .ccd file
- opened the .ccd file in notepad++ and inspected the contents

This is where things went terribly wrong. Any tape that has XDS information turns out totally garbled, with the XDS data mixed into the captions. For example this tape was of the "Today" show on NBC, and the the network ID is repeated all through the captions, as well as other unrelated data. It looks like this:

InteNBraCctive H@DTV][v:NBC

ONNB TCHIS TNBUECSDAY MH@ORNINGC
C30thH@NBCNBCH@NBCNBCH@ TNBHACTBC

it should look like this:

ON THIS TUESDAY MORNING, OCTOBER
30th THAT HOUSE HAS BEEN AN
ANNUAL HOLE HAL WEAN TRADITION

(mistakes are present in the original live caption)

This was sub-optimal, and after I did some testing, only tapes that were new enough to have XDS information had this problem. I was able to get a good .scc file from tapes that were too old to have XDS information. Since VLC was unbothered by the XDS information, and the .srt files from VobSub Ripper Wizard were correct even on the tapes with XDS information, It seems to be an issue with McPoodle's software. I tried ccExtractor, but from what I can tell it cannot export .scc files, and no matter how much I faffed about with FFmpeg the only thing I could get it to output was error messages and frustration. (It very well may be my fault, I am not the sharpest tool in the shed when it comes to command line applications)

So I have essentially hit a dead end and I am looking for a way to:

- Strip the XDS data from the caption data before it is converted to .scc with SCC_TOOLS

---OR---

- Other software that can read the captioning data from the DVD and export it as a .scc file or another format that supports EIA-608 features such as positioning.

Anyways, sorry for such an odd problem. If in the end it isn't possible I can always use the .srt files without positioning information, but I would prefer to keep the original positioning.

[dubster]

Do you absolutely need scc files, specifically? What's wrong with just keeping the original subtitles in their Vobsub format?

Also do you have a sample file to play around with?

Edit: And CCExtractor should support SCC output - it's right there in their source code: https://github.com/CCExtractor/ccextractor/blob/master/src/lib_ccx/ccx_encoders_scc.c

[Rycat]

Hi, thanks for the response. Unexpectedly (but a welcome suprise) I was able to get it worked out before work this morning, I had a very old version of FFmpeg on my PC that did not have the ability to parse EIA-608 CC data, and I was able to use FFmpeg to export the .SCC file. apologies for the late response, I am just now getting time to write here. I did try CCextractor but I was getting some timing issues on .vob files.

[dubster]

Good to know that modern versions of FFmpeg can do the job. Thanks for following up.