Friday 27 June 2014

Captioning live sport, old sport.

Hello again, gentle blogglers. A terrifying two months ago (where in the Seven Hells has the year gone?) I detailed how live voice captioners cover the news, which is an enormous part of what we do when we do what we do like we’re doing it for TV. Since then I’ve been prattling on about various issues and considerations involved in live captioning, but I have yet to tackle the other key area of live broadcast TV captioning – the sport. I’m sure there’ll be future posts covering the interesting nuances of captioning some individual sports, but I thought I’d give an overview first.
Some overviews are more useful than others.

Exclusive live sport is increasingly integral in sustaining the business model of broadcast television. Sport remains stubbornly at its best when viewed live, socially (in person), and on those megalithic miniature cinemas in our living rooms or pubs, rather than the matchbook-sized televisions in our pockets, or the procrastination-and-cat-picture machines on our desks. While the world moves online, broadcast sport is flourishing almost as much as ever, so by volume, the combination of news and sport constitutes the bulk of the work of the live captioner.

But some very distinct disciplines are involved in transcribing the vocal crescendos of a roomful of sweaty and excitable besuited millionaires watching fit young people do skilful things with some kind of orb, in a way which enhances viewer comprehension. The news is compiled from pre-scripted fragments and a fluid running order, and presents an aurally complicated but visually bland narrative, centred on the newsreader’s voice. Live sport, on the other hand, is unscripted, visually engaging, and with a narrative at least as much driven by the shot selection (here I’m talking directors, not batsmen) as by the commentators. If captions for a couple of sentences in a row of news content were missing, the comprehensibility of that story has definitely taken a hit; in sport, the visual narrative spine can survive some missing auditory information. We all know this, because not every sports bar enables captions on its screens (though the better ones do), and because decorum obviously dictates very different levels of quiet for a beer-and-movie night than for a beer-and-football night (different again for a beer-and-Four Corners night, because ABC thug life).
Talk during Kerry's opening monologue, and Mary-Sue McStockphoto will end you.

This means our first duty as sports captioners is to ensure we leave the visual narrative unimpeded. That dictates the cardinal rule – stay out of the way. Sometimes literally – as I’ve said, line 1, on the top of the screen, making sure to clear any score graphics, is a good rule of thumb for sports caption placement. There’s a bit more to it than that though.
Bottom line, stay out of the way, or people can't see the TV.

Captions can be ‘in the way’ in another sense – as a distraction. An eye movement study cited by Media Access Australia found that closed captions viewers spend 88 percent of their viewing time reading scrolling captions, 67 percent of their time if it’s block captions. But block captions are problematical for live programming anyhow, of course, since all the words need to be received before the caption can begin to be displayed, which exacerbates delays. So scrolling captions it is. Now, it’s safe to assume that seasoned caption viewers avoid that 88% figure while watching sports, by picking and choosing when they read the commentary. But it illustrates that as a matter of course, scrolling captions are a significant distraction from the visuals, and like reading text messages while driving, it may take more of your attention than you realise.

So that’s the next aspect of “stay out of the way” – avoid being a distraction. That particularly comes into play in sports where all the action comes in brief, sporadic bursts – like tennis, golf, cricket and American football. In the first three of those, the commentators conveniently play ball as well, usually falling into an atmospheric hush a little before each play, often leaving enough time to let the captions finish and then quickly clear them before anything starts to happen. It’s fun, much tea is sipped while respeaking these sports.

In other sports, however, action and commentary don’t alternate so conveniently, and then “stay out of the way” imposes two further strictures: don’t caption anything you don’t need to, and don’t caption anything you can’t. Take a look at this 15 second clip of soccer commentary:
TRANSCRIPT: “They’re having a pre-game dance, I know the drugs are good here, but that’s arrogant. He shows him a Red Mitsy. Your party’s over. Here’s Cavani, shakes – what the fuck happened there? Fuck no, Suarez, he is a rodent! Finds small spaces, bites cunts’ faces. What a fucking pest.”

Warning: the preceding video and transcript contain language which some may find Australian. Reader discretion was advised.

There’s another post in the works about how we tackle obscenity, but in the meantime, let’s look a little more closely at how we would caption that. First of all, you avoid captioning anything the viewer can see for themselves. So “they’re having a pre-game dance” adds little in the way of clarification or analysis and only duplicates what obviously appears onscreen. But “I know the drugs are good here, but that’s arrogant” adds colour and opinion, and should be captioned. For clarity you might then change it to “that dance was arrogant”, remembering that the captioning delay means it won’t sync up. Similarly, “he shows him a Red Mitsy” is a colourful turn of phrase which adds something to the commentary, but if all he had said was “red card”, the visuals would adequately have it covered. If he’d dropped the umpire’s name though, that may be some clarification which goes beyond the visually obvious, and could be included. He doesn’t mention scores or other statistics, but if he had, the same rules would apply – if it’s already onscreen (which it usually is), don’t caption what you don’t need to.
Dog

Then there’s what we can’t caption – the moment-by-moment play of the ball. In news captioning, a four second delay is reasonably harmless, since the bulk of the information comes in the form of an uninterrupted aural stream, and even if you read the final words four seconds into the end credits, you can comfortably say you’ve watched the news and I can say I’ve captioned it. But in sport, a four-second delay might mean five passes, making the descriptive captions quite meaningless. It would mean, in the above clip, by the time that “Here’s Cavani” appeared onscreen, we might be seeing the blocked shot at goal, or Chiellini clutching his shoulder, or maybe Suarez high-fiving Hannibal Lecter. Probably not Cavani though. But again, opinion and commentary aren’t so meticulously time-sensitive. Hearing “Fuck no, Suarez, he is a rodent” a few seconds late still pretty much conveys its meaning, so that should be captioned.

The question “what the fuck happened there?” is more ambivalent. Often when a commentator asks a question, the delay makes it redundant. So if in the next few seconds, we saw a slow-motion replay which told us exactly what the fuck happened there, as is likely, that question appearing afterwards might not be much use. It’s more pointedly illustrated by simpler questions like “he kicks long – will it go in?” It would be a long kick indeed which hung in the air long enough for the caption to appear before the sphere reaches the checkpoint (I don’t really know sports). Of course, the Bogan Aussie commentator’s question was probably rhetorical, expressing disbelief or asking what had happened in Suarez’s head, so we may well caption it. The good news is, we do tend to still catch the highlights of the play of the ball. When someone takes an extended possession, or causes a stoppage for any reason, that’s conveniently when commentators tend to talk about them more tangentially (and less time-sensitively), and that’s likely to be the “clarification and analysis” stuff which we do caption. That’s when we’ll hear what a good season he’s having, or how many marks he’s taken this game, or who he transferred from last season. So in that context, captions viewers will still see the names of players who go for a run, or get shown in replays, or start a fight, or challenge a call, just not the five names in a row you might get during a quick succession of passes. Unfortunately, we don’t caption what we can’t caption clearly, and “stay out of the way” condemns superflux.
...or for those reading verbatim captions, “here's the interception, but he's got a lot of work ahead of him to get past these four defenders. I wonder if he can make something of this?”

So the other thing to think about when captioning sport is vocab. I mentioned way back in my very first post how we can write our own sets of custom autocorrect rules, called house styles. Well, sports are where they really come into their own – most captioners will maintain their own separate house style for each sport they caption, so that when you have the WWE one enabled, it knows to capitalise the Yes Movement, but when you don’t, you won’t unintentionally make the Scottish independence campaign look like a weird cult. When you have your soccer house style enabled, it knows to replace “read card” with “red card”. And when you have the cricket house style enabled, it knows to replace “England win” with “England wrest defeat from the jaws of victory”, because that has to be what you meant to say in that context.

You also need to train your Dragon to recognise general vocab for the sport. This will include the names of great former players, current champions, commentators, positions, and technical terms.
Otherwise you end up with this.

Once that’s done, you’re almost good to go – just need to google the game-specific vocab (from a reputable source!), names of the players, stadium, umpires and teams (including nicknames) playing today. Train those into Dragon, look out for errant homophones (this excellent piece in The Monthly illustrates the difficulties presented by the AFL’s proliferation of different spellings of “Jarrod”), create verbal shortcuts for difficult names, and keep a team list open while you’re on air, so anything which is coming out wrong can be copied in.
May need a house style.

At the end of all this, there is one uniquely rewarding thing about captioning sport – exposure. So many public places show sport with captions, and the audience are often passionately attentive viewers. That’s a nice thing to remember, if, like me, your understanding of sport is at times a little limited.



Disclaimer.

Monday 23 June 2014

Mishaps, but prefers Ms Haps.

More substantial posts are well on the way, honest. In the meantime, funny captioning mishaps!

It's been an entertaining week for it, some good slow fun captioning some informative offline documentaries. I was alarmed at one about cheese-making which saw the ingredients include "Kurds". There is surely no need to give moral vegans even more ammunition. A wildlife reporter described dolphins as "big and muscly", but Dragon, turning its mind ever towards duplicity (and roasting livestock, amirite Khaleesi?) decided it should be "bigamously". And finally, a documentary on Captain Cook, the dramatic tale of the Endeavour's hull being torn to shreds by some reef or other. The situation was so dire, he had to "throw the bankers overboard". Well, that's one way to keep them honest.

It's funny the way phrases used everyday will just fail once in awhile. I would have thought the personalised Dragon profiles of people who respeak the news for a living would be pretty used to the phrase "and now, onto our main news". But alas no: "and now, onto our menus". And who knows how many times I've respoken "Bin Laden" during the world news, only to have Dragon suddenly decide "bin lard and" made more sense.

And then there are some that just bewilder me. As a reminder for any new readers, Dragon's voice recognition makes contextual guesses about what the speaker most likely means. So, presented for your consternation: the phrases "write a horse" and "how can babies be borne?" Each of these fails to surmise what is literally the single verb most likely to be used in connection with that noun. The latter phrase remains an excellent question, though.

And finally, more merriment from the golf. I'll post more fully later on what makes the golf such a treasure trove of errors. For now though, a doozy from a colleague. And I must say, if I was going to be surprised on the 17th green, I would rather it be by a "birdie putt" than a "body part".

Why? Why does this clip-art exist?


Disclaimer.

Saturday 14 June 2014

Mishap Mish-Mash

This week has been the blurst of times in captioning land. Going away for the weekend in a moment, but first a few quick mishaps from your humble Rogue.

Lots of intense military commentary this week, one strategist talked about "our objective", and Dragon decided he meant "Arab jetty". Which, I don't know, maybe? Also in international diplomacy, we say Putin "engaging" with his diplomatic counterparts. Or as Dragon had it, "Putin in gay and" with his diplomatic counterparts.

The British passport crisis is in the spotlight, but I think Dragon might have the solution. The Home Secretary Theresa May has been surreptitiously replaced by the Spanish sausage "chorizo may". Also in UK politics, the Newark by-election provided many hours of unscripted commentary, and thus some giggles. One profile of the area described it as "90 minutes from London by rail, longer by boat". But Dragon got its anti-by-election snark on by making it "longer by vote".

When captioning golf, it is truly awkward when the ball is struck with a "sand wench".

Lots of industrial action this week, but I don't really think Brazilians are striking for higher "way Jews".

I saw this corker from a colleague. The popular Asian surname Ng? Well, they'd made an unfortunate autocorrect rule which made this fellow "Mr nanogram".

And in a complete "what the...?", the phrase "self-explanatory" was once rendered as "Celtic blood trick". The mind bloggles.


Disclaimer.

Monday 9 June 2014

Meet The Captioners

So earlier I waxed philosophical about, like, what it means to be a captioner, man. This is a companion piece of sorts, mercifully rooted far more in tangible reality. I wanted to talk a bit about what sort of people we are. Who is drawn to this kind of work, what skills and backgrounds do they have, what are they like as a group?

Super friendly.


Well, captioners trend very literate. Compulsively, fanatically, instinctually, neurotically literate. We dot eyes like Charles Schultz, and we cross teas like a chamomile Earl Grey chai. An attuned instinct here is actually of more use than a more conscious commitment to precise spelling and grammar. Quite often it needs to be automatic, because pre-air editing must routinely be hammered out extremely quickly (many news rundowns only become available minutes before our 15-30 minute stints) and a respeaker’s on-air editing and corrections can literally never command more than partial attention (we must always also be listening, speaking, checking the TV screen, and often googling names and managing cued text as well, and that’s if things are going right). We don’t perform elegant, refined surgery upon the language, but quick and dirty field medicine with a corkscrew and a rubber band. So anyway, if syntax is a cabaret old chum, then we’ll be the ones singing Tomorrow Belongs To Me.

Which is not to say we're prescriptivists <shudder>


We also trend highly news-literate. Of course there’s a chicken and an egg fighting for temporal primacy here, but I think most voice captioners were oft-enraged news-obsessives even before we started gargling with it for hours every day. The sort of people with hours-long think-piece-backlogs to read, and insurmountable daily Twitter feeds. Sport is a little more selective – many of us are enthusiasts of one shade or another, but many are not. And given the dozens of sports we cover, most will discover a few sporting passions they didn’t know they had, whether it be for WWE or Nordic Combined skiing. Obviously more generally, we also approach an unusually large volume of television with an uncommon degree of concentration, so a fair bit of television literacy is also kind of a given. We then have to be able to digest that volume of content without fatigue. Taken altogether, it boils down to us being a pack of big nerds.

In retrospect, delicious segregationist lollies really sent an ambiguous message.


The need for very clear and consistent diction also skews the demographics. Much-lauded in all discussions of voice captioning is the ability of Dragon or equivalent software to be customised to the idiosyncrasies of the user’s voice and accent. But it remains the case that some idiosyncrasies are more quickly trained in than others. More common and conventional accents begin to be recognised sooner than unusual hybrids, which inevitably limits the cultural diversity of our profession. But many different forms of Australian, Kiwi, Canadian, American and English Received Pronunciation all feature among the accents of the talented voice captioners with whom I work. The inclusiveness of live captioning is also compromised by the practical need that we be completely free of speech and (perhaps ironically) hearing impairments. The pipeline between our headphones and your screens must be as unobstructed as possible.

Perfect.


We trend a bit to the political left. Captioning may be one of the more steadily growing branches of the mainstream media, but it’s far from a goldmine and there are quicker ways to live out your sweaty Randian wet dream of capitalist world domination (plus HR refuse point blank to pay us in gold bars delivered to Galt's Gulch). We’re in it to help people. One of the major attractions of the work is getting to help include those who may otherwise find themselves excluded from cultural life. Deafness is just an impairment, but inaccessible media can conspire to make it a disability, and it's nice to be working against that. We’re also left of centre because, you know, see above re: literate.



We’re also slightly disproportionately young and unmarried, though there are plenty of exceptions. The hours can be a little weird. Our busiest news times are during the breakfast chat shows and the evening news, with most of the sport concentrated at weekends. So it helps to be flexible and fancy free enough to say yes to the generous penalty rates at evenings and weekends once in awhile. We’re basically a few standard deviations more likely to have cats than children.



So how did we get into this gig, what were we before we were captioners? Well, all kinds of things. 1980s rock musicians, postgraduate scholars writing on medieval literature, or television, or fantasy novels, law students, podcasters, teachers, journalists, actors, those high school kids who hung out in the library and formed an anime club, TV studio staffers, linguists, scriveners of all sorts. For a languagey (languid? linguical?) person, it’s not a bad thing to be.




Disclaimer.

Tuesday 3 June 2014

After These Messages: Captioning Advertising

There’s something a little topsy-turvy about the business model of TV closed captions. Or rather, perhaps, something right-side-up about it.

Accessories sold separately.


Commercial TV blazed a new-ish trail, which has since grown fairly ubiquitous with the proliferation of information economies. We all know roughly how it works; rather than selling spectacle to eyeballs for a ticket fee (bread for circuses), commercial TV purveyors first capture wild eyeballs with arresting content, and then on-sell them to the powers behind the spectacle. Rather than buying the shows we want to see, therefore, we instead agree to view them interspersed with advertising we don’t want to see.



This means that in all facets of the commercial television economy, the advertising is what matters – the signal is the noise. In monetary terms, a station outage during 30 seconds of actual programming is far less costly than the same outage which knocks out two 15-second “words from our sponsors”. As viewers we don’t often think of it that way, but that is the hierarchy. While networks will typically offer showrunners the creative freedom required to build compelling programs, any programming which in some way damages the attendant advertising would have to achieve dizzying success to justify its continued existence. Or be on HBO.

Buy War Bonds.


If you keep all that in mind, you might be a bit puzzled when your closed-caption viewing cuts to commercial. Because, truth be told, it’s a little sparse. Plenty of ads with no subtitling, and sometimes the overhanging captions from the program are even allowed to finish up in the ads. So why wouldn’t broadcasters caption what they see as the most important part of the show? Well, it’s all in the unusual background to the captioning business model.



Captioning now is as lucrative as it has ever been. We have cheaper technical and training costs; an ageing Baby Boom generation shifting wealth and political influence onto older (and thus more hearing-impaired) demographics; a recent focus on the impact of closed captions upon literacy, particularly for younger viewers (teach them to read while they watch cartoons); crowd-sourced amateur captioning filling in some crucial accessibility gaps; and hell, even increasing population density, which will probably spark a rise in both muted TVs and tablet-based viewing on crowded trains.

They can't all be "working on their novel".


But this post isn’t about arguing why today’s broadcasters should adopt more thorough captioning policies. Smarter minds than mine are already doing that. My point is that while captioning now makes excellent commercial sense, before it was thus, it was a legal obligation. The recent Ted Talk on caption accuracy gives a useful overview, but basically, large broadcasters were first required to provide closed captioning on just a few prime-time shows, and some news content. Then gradually more and more hours per day, then more of the smaller broadcasters were brought into the fold, then quality standards began to be introduced. A few publicly-funded, inclusively-minded or entrepreneurial broadcasters exceeded their requirements, the rest just waited to be compelled. Or, indeed, fell short and got fined, as ESPN recently were over their audio description.

In fairness, some moments truly defy description.


But despite this tediously slow growth, there are some upsides. The compliance mindset means that captioning is uniquely, compared with broadcasting as a whole, targeted towards what viewers want and need. If you’re a broadcasting corporation doing the bare minimum to satisfy statutory requirements, you won’t go captioning something as superfluous to the requirements as advertising. You’ll be covering programming, starting with the most important.

A slayer slays, a captioner watches.


As it stands, many captioning companies are under large contracts to caption big blocks of programming for their client networks. But advertising is usually left for the brands themselves to ask for and pay for. It’s an interesting financial calculation for them. As ad hoc, one-off captioning jobs, 15 or 30 second ads can be proportionally quite expensive to caption. After all, if you were a broadcasting company with four channels, each of which were captioned 20 hours a day, we could probably cut you a pretty sweet deal for the 560 hours a week of business you’d bring us. But less so if you’re an advertiser with a 15 and a 30 second ad, each of which we’ll caption once, and then replay endlessly for three months. But advertisers want their content seen, it’s how they do, so the increased exposure might be worth the investment. If you’re a small advertiser though, and you want accessible ads, you will have to ask for it. Obviously they have to be captioned offline, as a six or seven second delay would virtually always kill the punchline of an ad. Then again, many ads are designed to be effective even with just audio or just visuals. They reinforce their message in a few semiotic ways – the logo and colour palette, the characters and celebrities, the onscreen slogans, not to mention the products shown, are typically as much a part of the brand as what is being said, and closed captions may at times be superfluous. Our usual gig is reinforcing narrative, but there may indeed not be any narrative to reinforce.

There are some exceptions.


It also means that as live captioners, we treat ad breaks, rather like viewers do, as dead time. We stretch, sip our coffee, chat, cough. We also prepare for the next segment, retrain any words that were playing up in Dragon, hand over to our co-pilots, resolve technical problems. We take a screen-break – I suspect it’s similarly a relief for many captions viewers to be able to stop reading 150 wpm scrolling words on a screen for three minutes, although of course advertising can still represent an important part of the intellectual life of a culture, and universal accessibility may be preferable.

All viewers deserve this.


There is one small additional aspect of captioning in the ad breaks to consider – news updates. Most Australian commercial networks play 30 or 60 second news headlines hourly or half-hourly during ad breaks, along with an increased concentration of them in the lead-up to their newshour. This actually does need to be captioned, as it technically counts as news content, which is singled out by statute as an area of captioning importance. Although it’s only a minute here and there, it can actually be a full-time role. It needs to be prepared for as diligently as any other news, so the captioner of the unrelated program can’t do it. It also typically requires separate headlines for each state or major city, and can be either pre-recorded or live, so the updates captioner remains on her toes.

Help! Mila Kunis keeps making me do news updates!


And with that, I can hear the ad break finishing up. Gotta zip.


Disclaimer.