Closed Captioning question

[MarcB]

I was wondering - the people who do the closed captioning on a pre-recordedTV show - how do they do it? Do they have a copy of the script and then they send the tape to the various networks that air the show? The reason I ask is because of a goof on the episode of Full House I just saw on ABC Family.

Dialogue:

DANNY: Does anyone want to buy a donkey?
DANNY: Does anyone want a free donkey?

Closed Captioning:

Danny: Does anyone want to buy a donkey?
Danny: Does anyone want to buy a zebra?


"Does anyone want to buy a zebra?" makes no sense because the episode was about how Michelle bought a donkey.

 

[quadraphonic]

It's a system of levers and pulleys and the use, some would say abuse, of dimensional physics.

 

[Darth_vader]

This is the how I was explained some years ago, by a former Vitac captioner who lived near me:

Often when a pre-recorded show is in post-production, the captioners are sent a tape of the show, as well as a copy of the show's script from which to transcribe. This enables the captioner to insert the words in at the proper point in the show, as well as allow them to position the caption text where it needs to be (e.g. at the top of the screen if something is being displayed at the bottom.)

Sometimes last-minute corrections or changes to the show are made before the final product is syndicated but not before the captions are already written, or the captioner is given an earlier revision of the script (i.e. not the *final* final draft.) Even more disturbingly common still, are the studios that actually use an ABBREVIATED version of the script for captioning. (MGM seem notorious for doing this.)

And the transcribers themselves are human, and sometimes the occasional typo gets introduced as an operator error. Occasionally when the text is being proofread, an error gets overlooked and makes it into the final production.

Live captioning is just that, and generally the transcriber is sitting in a production booth, listening to the broadcast on headphones and typing on a stenotype machine (like what court reporters use to transcribe the court's proceedings.) Some newer stations use a "real-time" speech-to-text system, thus eliminating a transcriber from the production chain, and as a result what you read on the screen is the same thing the speaker is saying (albeit sometimes with less-than-stellar accuracy) and there's another transcriptionist ending up on the payroll at MickeyDee's.

That's the way I understand it works, anyways.

 

[DToTheJ]

I had previously inquired as to why contractions, when uttered, are translated into the pre-contracted term in closed captions. You'd think they'd want to save the space!

 

[anotherguyTN]

Don't live newscasts tie the captioning in to the Teleprompter?

 

[DToTheJ]

I have seen where that practice has taken place - a good example is the NY1 cable news channel in New York. Another example, albeit a little bit sloppy, is News 12 The Bronx. The closed captioning includes terms used in newscast script, so it's clearly connected with the script that the anchors/reporters read; the problem is, the closed captioning on News 12 runs a little bit ahead of the broadcast airing in real time.

 

[DToTheJ]

And here's a video of the closed captioning on ESPN during a discussion about quarterbacks, and when Tony Romo of the Dallas Cowboys is mentioned, the closed captioning reads - well, see for yourself:
http://www.youtube.com/watch?v=F19QAIqP8PQ

 

[Markxxx]

One thing I noticed at the gym is this. They have several TV sets all broadcasting the same channel and the closed captions on the different sets display the captions differently.

Some are accurate, some are not. What's surprising is one set will leave chunks out. So the way a set renders the caption must have something to do with it as well.

 

[Darth_vader]

"I have seen where that practice has taken place - a good example is the NY1 cable news channel in New York. Another example, albeit a little bit sloppy, is News 12 The Bronx. The closed captioning includes terms used in newscast script, so it's clearly connected with the script that the anchors/reporters read; the problem is, the closed captioning on News 12 runs a little bit ahead of the broadcast airing in real time."

The news on KPTV seems to read exactly as you described, with directions, timing cues and such like. They must play back the same text into the caption encoder that feeds the TelePrompTer or something!

 

[quadraphonic]

"I have seen where that practice has taken place - a good example is the NY1 cable news channel in New York. Another example, albeit a little bit sloppy, is News 12 The Bronx. The closed captioning includes terms used in newscast script, so it's clearly connected with the script that the anchors/reporters read; the problem is, the closed captioning on News 12 runs a little bit ahead of the broadcast airing in real time."

The news on KPTV seems to read exactly as you described, with directions, timing cues and such like. They must play back the same text into the caption encoder that feeds the TelePrompTer or something!

WTVD's always shows the to-be-accented words in quotes on their CC.

I still like the system of levers and pulleys....

 

[w9wi]

The news on KPTV seems to read exactly as you described, with directions, timing cues and such like. They must play back the same text into the caption encoder that feeds the TelePrompTer or something!

That *is* a common way to do it. Prompter has a serial port, which passes script data to the caption encoder.

The text on the prompter has to be in sync with what the anchor is reading, so it's a good source for caption text too.

I don't know what newsroom computer system KPTV uses. The one we use, you can mark text as "production cues". This text will display on the prompter but won't go to the caption encoder. Unfortunately, the writers don't always bother to mark it.....