Sunday 27 April 2014

Perfectly Executed Closed Captions

So this post from The Chive, "Perfectly Executed Closed Captions," deserves a mention. It's refreshing to see captioning whimsy emerge from a source other than mishaps, in at least a few of these. Describing non-verbal background cues is a strange little window of creativity for a live captioner. We have some leeway to describe music as we feel is appropriate, and to determine when silence is sufficiently portentous to warrant a mention.

So "screaming like a sissy" might fall foul of what we would consider appropriate gender discourse as broadcasters. But it might be the perfect articulation of the sound if, for example, one character had just demanded of another that they "stop screaming like a sissy".

"Walt Jr. vomiting" may not warrant a mention if we had just seen it happening in a preceding shot, but if the camera lingered on Walt, it might be important to know the sound persists. And whether to call him Walt Jr (I would elide the full-stop as our software tends to read that as the end of a sentence) or Flynn might depend on context.

"Phrase whore" makes me wonder a little. It could be a homophonous name which failed. Or it could be a verbal shortcut which went awry. Perhaps the software misheard "phrase four" which was supposed to be automatically replaced by a specified name. In any case, coarse language house styles should probably pick that up.

"(Dramatic instrumental music)," "(BLOWING SOUND)", "Cat sound", "(silence)", "laughing", "(jazz music playing)", "(sneezes, farts)," "(epic thrash metal)", "(melancholy guitar music)" and "(JOCKS LAUGHING)" each reflect really interesting creative choices by captioners. Whether to include these non-verbal cues and how to describe them is obviously integral to how closed-caption viewers interpret the program. You can see immediately, reading "dramatic instrumental music" contains none of the drama of actually hearing it, but it can still clarify the meaning of the visual information. Louis CK's words really would have less meaning if his syntax was transcribed without his many rhythmical and expressive pause utterances.  We see in "laughing" how precisely duplicating onscreen information can become tedious or absurd, a kind of parodic reflection, but clarifying that she isn't screaming may be crucial.

The tennis screencap stretches credibility. For one thing, the sounds of tennis can reasonably be inferred from onscreen information. For another, captioning delay of a couple of seconds can render such pedantic play-by-play description rather meaningless. More on that in future posts about sport captioning.

Thanks also to The Chive for the banner which heads this blog. :)



Disclaimer.

1 comment:

  1. Sound labels are a strange one. Thankfully the ones I see on broadcast TV generally make sense. Some that I see posted online I wonder about the source because they are describing things that are visual and I don't think are necessary to include, or as you say like the tennis example serve no informative purpose to deaf and hard of hearing viewers. I think the purpose of them is sometimes misunderstood.

    ReplyDelete