Well, you can decode, edit as text, and recreate the captions with McPoodle’s tools (McPoodle being the individual whose tireless work I referred to earlier). And then you can put the captions on VCDs or SVCDs.
But most DVD programs won’t let you author DVDs with Line 21 data – you have to have a high-end professional authoring system for that, like Spruce DVDMaestro (now defunct and taken over by Apple) or Sonic Scenarist. I use DVDMaestro 2.9, even though it’s aging and not too great these days, because it’s a heck of a lot easier to use than Scenarist.
Frankly, I’m currently thinking of just converting all my captured closed caption data over to subtitles. I can’t really think of a disadvantage, and it would work on all TVs, even old ones without closed captioning hardware … and it’s often much easier to turn subtitles on and off than it is to turn closed captions on and off (although new TVs often have a simple “CC” button on the remote … but all of my TVs have to go through a menu setup to turn them on or off).
The only downside is that, as far as I understand it, subtitles are stored as pictures rather than text data, so the actual text data will be lost. That might be bad if I ever needed to convert FROM the subtitles again…
Right but those VCD or SVCD “captions” would be subtitles, not true line-21 captions. However I believe that the S/VCD format uses text subtitles rather than image subtitles (I might be wrong here, its been a while since I played with that.)
Another one would be DVD Architect by Sonic Foundry which is what I use. It’s quite nice especially compared to Scenarist (with that interface from hell).
I use subtitles myself. I’ve thrown away the idea of using CC a long time ago. The main reason being with subtitles you can use non-text (images of phone ringing, music symbols, etc).
Well - that’s why you would keep the original text transcript somewhere on a trusty CD-ROM stored in a safe box located in a fallout shelter.
The reason they are stored as pictures is obvious - you could then do subtitles in an old dead runic language if you wanted to.
No, I’m talking about actual Line 21 closed captions, not subtitles. You can also add subtitles, of course, at least in SVCDs (I’m not sure if VCDs support subtitles or not). That’s why I said VCD/SVCD and not DVD – VCD and SVCD both have provisions for putting Line 21 caption data in a file on the CD. I believe it’s \EXT\CAPTxx.DAT, where xx is a number from 01-99 representing the particular stream.
All this is theoretical, though – I’ve never actually seen a VCD or SVCD with actual closed captioning, and I’ve never really tried to make one. And there seems to be some evidence that it may be impossible because of the way VCDs and SVCDs work. But then why build the capability into the (S)VCD standard?
Interesting! So DVD Architect will do real Line 21 data? If it still supported and being developed by Sonic, that would give it a big advantage over DVD Maestro.
About garbled CC data:
The CC information is encoded in a particularly brain-dead way on line 21. The data is sent as fixed-length dashes of black or white, you can see this on a TV if you have some way to turn on “underscan”, or on really old TV’s with a vertical hold control, roll the black bar down.
As an extremely feeble attempt to add some error-correcting, they send each character TWICE. That’s a particularly feeble method of error detection-- you don’t know WHICH of the two if any of the characters is the correct one. Nowadays there are error-correcting codes that can recover from single or multiple bit errors, but the CC protocol hasnt advanced to keep pace.
The bits are very prone to damage from any interference, ghosts, or some of the other signals that various links in the signal chain add to the vertical blanking interval. Often on line 20 you’ll see some color bars or sine-squared pulses. They’re very useful for adjusting the signal level and phase, but sometimes they can bleed onto line 21, or confuse the CC decoder in its line-counting.
Are you sure about this? It was my understanding that they send two characters per line. That would explain why when a frame of code on line-21 gets messed up there is always a 2 character drop out. I just asked Carl Jensema this question (he is one of the CC pioneers that I work with) and he says he doesn’t remember any error correction built into CC. Maybe I can convince him to talk on this board (he does subscribe to the SD).
No chance. Changing the CC standard will break every single CC decoder in use today. CC with digital TV was supposed to be an improvement (you are supposed to be able to do stuff like be able to select the font yourself, change how it is disiplayed, the colors used, etc) but I have not seen this yet in any DTV receiver.
For more info, you might check out the National Captioning Institute at http://www.ncicap.org/.
I also worked at a local TV station years ago. On the premier of our captioning effort in the early 90s, some unprintable words were included on the caption broadcast. They were intended to be a cue for the sportscaster and were bleeped out of the tape, but “f___” was seen by all the hearing impaired who had lobbied for captioning. The station manager was pretty upset. Us production guys thought it was pretty funny.
The NCI is only one of many dozens of places that offers captioning services. For real captioning information (rules, regulations, standards, etc) visit the FCC web site. Here are some links:
Actually this is not true. Subtitles can very well have non-spoken sounds represented. In fact subtitles do a better job of doing this as they can use an image of a phone flashing to represent a telephone ringing, and so on. I have seen it done this way on some DVD’s. But really, with subtitles, anything is possible because it is really an overlaid image format rather than text-only. While there are some standards for representing non spoken sounds with line 21 closed captioning (varies by studio doing the captioning, but at least they have some) there are no standards whatsoever for subtitles. For example, some DVD’s are good with subtitles and actually make the subtitles followable by adding a dash at the beginning of the line when the speaker changes (like CC does) but other’s do not and it becomes impossible to tell who is talking. Stuff like that. Oh well.
I could be mistaken, but I believe chorpler was referring to recording his/her own captions onto original DVDs, and was not saying that DVD recorders wouldn’t capture the captioning encoded into off-air broadcasts. AFAIK, that should work. Can anyone confirm that? (I don’t have a DVD recorder yet, and my TV is pre-CC era.)
But if you send the HDTV signal into the RF input of the TV, won’t you get the TV’s closed captioning? Yes, I know the picture’s inferior to using the component cabling, but you’ll see the captions, right? You could switch between the TV’s inputs to see the captions or the better picture.
I’d like to keep this thread focused on the Staff Report – that is, on the question of how closed captioning works. If you want to ask about tricks that you can play or things that you can do based on closed captioning, please start a new thread, preferably in the General Questions forum… you’ll have a better chance of getting responses.
Umm - you can’t send a HDTV (ATSC) signal into RF input - only NTSC signals. Anyhoo, I get your point but that kinda defeats the whole point of having a HDTV and a progressive scan DVD player now, doesn’t it? :-\
I was referring to the VCR-like DVD recorders that have come onto the market in the past year or so and are still too expensive for most people to buy. I was considering buying one so I could save time setting up and burning DVDs on my computer, but so far everybody who owns one that I have talked to has reported that, when you record a TV program onto a DVD, the closed captions are not recorded.
It does seem odd, though … VCRs have no problem recording them, but then VCRs don’t digitize the signal and save it as MPEG-2 compressed video – since the captions are transmitted in an off-screen portion of the video signal, perhaps DVD recorders end up clipping off that portion of the signal before recording. I know that my computer’s MPEG-2 recorder crops the raw video signal quite a bit before saving it. But on the other hand, both of my TiVos and my ReplayTV have no problem recording and playing back captions, and they also digitize the video and save it in MPEG-2 format …
And, of course, I could be completely wrong, if the people I talked to were wrong. Or there could be newer recorders coming out that do indeed record the captions. I would welcome any report of a DVD recorder that is verified to save captions as well as video.
Or perhaps, like mherbold, most users of the recorders are connecting them through a video (composite or component) input, in which case the TV’s CC circuitry isn’t involved. But you could be right that MPEG digitizing doesn’t preserve them. I hadn’t thought of that.
I did appreciate the answer, and as someone with one foot in the hearing world and the other in the Deaf-world, I use CC all the time. Thank you for a fairly concise description of the CC process.
OOPS: My againg brain conflated a bit: The CC data is not repeated for error correction purposes. And there is a parity bit to aid in error detection, but not correction. So errors are going to show up as whatever the decoder decides to do, probably drop that character, or display a space, or whatever.
It’s still a mighty marginal design.
The TV’s closed caption decoding circuitry is only used for displaying the captions on the screen. The video signal coming into the DVD recorder should still carry the closed captioning data on line 21, regardless of the connection type. If you hook up a VCR via the coaxial cable input, or the S-Video input, or the composite line in, it will still record the closed captioning signal. It seems a shame that DVD recorders won’t.
