Axxlog

It can now exclusively be revealed that I write scripts for audio description.

Yes! It’s true. And it’s been a publicly-available fact for over a year – if you happen to watch one of the movies I described and stick around for the credits.

So far, I’ve done a mere two, both of them large-format films: Santa vs. the Snowman and Imax NASCAR [3D]. You quite possibly may have missed Santa vs. the Snowman, but the NASCAR picture is gonna be running for a while. You won’t be able to watch the movie with descriptions in Canada, since no Canadian Imax theatre has a MoPix system. Only some of the U.S. theatres will run it, and it is impossible to find the whole list of theatres online. You’ll have to phone your local house, unfortunately.

At any rate, I wrote the scripts at the request of my friend and esteemed colleague Nancy Harvey of Nancy Harvey Productions. She’s overseen language dubbing for Imax and other productions for many years and has occasionally been responsible for captioning and description. She’s been a big booster, actually, and hired me for this purpose.

Methods

I used to think that videotape remains a tolerable source medium for any kind of accessibility treatment. But I suffered through the minor shifts in read/write heads on even the best of my four VCRs for Santa, which forced me to rewind to get just the right timecode for a description. This time I gave up and went digital. I digitized the NASCAR VHS tape using an EyeTV box and, through some jiggery-pokery, saved it as a slim, tidy 519 MB QuickTime file, which allowed me to stop and start on a dime.

I decided to live dangerously and just run the thing and describe on first viewing rather than watching it in totum first. I am not convinced that a dry-run first viewing is necessary. It’s certainly not necessary if done with sound only and no picture. The ostensible rationale is to make very clear to sighted describers exactly which moments in the production aren’t understandable from the soundtrack alone. But describers can do that while they’re watching the picture for the first time. They’d only write the descriptions when watching the movie anyway; an initial viewing for review is a waste of time.

I did a lot of reading aloud against playback to ensure the descriptions would actually fit, and took into account the fact that I talk faster than narrators. I nailed the time allotments pretty well, as it turned out.

The trickiest part by far was describing one sequence after another of cars racing around tracks. I decided to stick to basics and describe what I observed – not just cars around a track, but specific cars around specific tracks in certain weather with certain crowds, crews, and onlookers.

Keeping the example of 2 Fast [&] 2 Furious quite prominently in mind, I assumed that many viewers attending a NASCAR movie are gonna be car buffs, so whenever possible I identified make and model of vehicle, which in some cases required extensive Googling. Under better conditions, printed sources would be available. (Nancy later asked one of the producers to look a few things up for us, which he did. Fortunately, we had time for that step.) I have a small fear that I flubbed the identification of a few of the cars – if so, for the love of God tell me – but that is merely a fear; I have reasonable confidence that every car I name actually is that model. I’m a car buff myself, you see, despite being unable to drive.

A few visible details were too small to discern in VHS and certainly too small in QuickTime; we had to double-check against the Beta tape in the recording session. Beyond that, so help me, it wasn’t that hard. Watching 130 hours of cinema audio description, plus description on TV and home video for the better part of a decade, has taught me a few things. (But you’ll be able to fact-check my arse; see below.)

Interestingly, this picture, with a 46-minute runtime, took slightly less than six hours to script. That would be considered monumentally fast for one leading description provider I know (where getting through a mere ten minutes per workday is the standard) and possibly a bit slow for another (where turning around a stack of tapes in two or three days from start to finish is the standard). My work time was much faster than with Santa. Using online media helped, but I think I’m just getting better at it.

My triumphant moment: Describing a dualie pickup truck as such – because that’s what it is.

Script mechanics

I wrote the whole script in HTML. I do everything I possibly can in HTML.

I used markup as follows, which you are free to duplicate. The definition list is, by spec, the correct semantic markup for dialogue.

<dt class="T"><span class="hours">01:</span><span class="minutes">03:</span><span class="seconds">42:</span><span class="frames">06</span></dt>

<dt class="lead-in">green flag drop on race day</dt>

<dd class="DX">Shirtless guys in shorts and gals in halter tops watch the race from the roofs of their RVs.<br />

[end before women’s voices]</dd>



I should have used another dd for extros like [end before women’s voices]; I’ll do that next time.

Timecodes were a bother and are possibly a reason not to use semistructured HTML for this application. The tape I was given was 48 seconds off the final tape; I had to add that increment to every timecode. With a great deal of text massaging in BBEdt, whose steps I cannot now recall, I was able to import the text file into Excel and use Belle-Nuit’s timecode macro to add 48 seconds. (The documentation is crap. I know how to use it now and can help you out if need be.) The results were not always correct – in three or four places, our timecodes were off – but the day was saved.

Another downside is printing, with widows and orphans everywhere. Importing into Word was a possibility, but it did not read the indented stylesheets properly. We lived with the widows and orphans; perhaps next time I’ll try harder.

Nancy and I spent an hour on the phone cleaning up the script and flagging parts for fact-checking (“verificationism”).

Recording

We eventually found a narrator, after a circuitous process: Jeremy Harris, the voice of Toronto 1.

It is not an easy task. For this kind of production (essentially a guy movie about guys racing cars for guy spectators), a male narrator is preferable. Key here are clear enunciation and a voice that is mostly unbiased yet prosodically varied enough that you can understand the actual words. Further, you need vocal qualities that are not in the least bit kooky or marginal. That latter requirement immediately disqualifies half the voice talent in town, who are accustomed to “acting” in kooky, marginal radio commercials. It was very easy to listen to MP3s and eliminate candidates right off the bat. That could be a kind of selection error: If all you give us are sound samples of kooky, marginal radio commercials, that’s the only kind of work you’re going to get.

It is difficult to separate kookiness and marginality from the intrinsic qualities of a voice. I am not at all expert in this area, and indeed all I did was suggest names to Nancy, who has an ear for this sort of thing. The criteria may be hard to put into words, which is something I find frustrating, but a good ear can tell at once who works and who doesn’t. Our instincts were correct, and Jeremy gave us a good read.

We made very few corrections to the script, and only a couple of deletions.

Studio session

Now, what went on in the recording studio? Well, you’ve come to the right place: You Are There!

We recorded the description track at a Major Downtown Studio with a sound technician and an assistant who wrangled ProTools on a dual-monitor G3 running OS 9. I brought one of my laptops and a raft of printouts. We had to reboot ProTools and our producer mike didn’t work, but we had no other technical glitches. (We could still talk to Jeremy using the room mike.) Jeremy stood – the entire time, by choice – in the quintessential soundproof booth. I was amazed to learn that his distance from the microophone could be detected by Richard, our technician. Twice without looking, he told Jeremy to move back to his original spot.

Since we knew we weren’t recording a home-video mix that would require ducking and restoring the main audio, we were free to custom-craft a monophonic voice recording meant solely for listening through headphones. (There was no chance of Jeremy’s voice clashing with that of the NASCAR movie’s narrator, Kiefer Sutherland. You the description listener always hear our guy through headphones and Sutherland everywhere but headphones.)

We had the right atmosphere in the room: Time is money, but if anyone heard something amiss with a take, we spoke up. (All of us heard something at one point or another – even Jeremy.) The flaw may not actually have been there – a couple of playbacks proved those takes were fine – but if you don’t speak up, mistakes will slip through. That may nonetheless happen even in high-profile productions; I’ve heard Miles Neff croak here and there in DVS Theatrical. Still, you have to develop an environment where anyone in the room may speak.

(At the one and only DVS recording session I attended, the producer heard us muttering on the couch at the far back of the room. Right into the microphone, he muttered back “Everyone’s a describer.” Indeed we are!)

Now, how long did it take us? A long time. Really quite a long time. Longer than we’d have liked. We could have shortened the session by a small amount in retrospect, and I already have a checklist going that will streamline the process for our next such production, but in all fairness, we were working at Imax levels of sound production and we cut no corners whatsoever. The technician, assistant, and Nancy also did a post-recording mix to remove stray breaths, modulate the volume to match the original sound level, and smooth pacing in a couple of cases. To simulate the actual theatrical experience, they listened to description over (fabulously-high-quality) headphones and main audio through room speakers.

And ironically, unless I fly to Houston or something, I will likely never hear the finished product.

Photos

I told you already: You Are There!

Jeremy, plum tuckered

We’ve just finished the entire recording session. What better time to snap our narrator’s picture than when he’s pooped? My fave detail here: The folded and discarded script pages on the floor!

View from the big chair

Nancy and I sat on a platform at the rear, from which I snapped this photo. (My picture of Nancy is so hideous it’ll never see the light of day.) That’s Richard, the sound technician, at the controls; he’s wearing a Molson Indy T-shirt for the occasion. On the monitor at top left is the credit sequence, which took years off our lives.

Listings

To return to the question of listings: The NASCAR film was captioned by WGBH for the Rear Window® system, but they didn’t do the description. In cases like these, WGBH likes to pretend the movie barely even exists. You can find Imax NASCAR on their Descriptions of First Run Movies Shown with MoPix and All MoPix Films pages, both of which list the movie as captions-only. That was true until last week. Let’s see how long it takes WGBH to update the listings.

Fact-check my arse

In the grand tradition of Web publishing, I hereby invite readers to fact-check my arse. I put two years into watching DVS Theatrical productions, griping about every warble, cough, and malapropism – and laying down for prosperity all appropriate praise for the good work. I told it exactly as I saw it, good, bad, or otherwise. It would be hypocritical to expect listeners of my description track to refrain from doing the same.

I’m merely the writer, not the producer or narrator, and I’m making the following offer purely in that capacity. If you’re at all interested, see if you can figure out where the NASCAR picture is playing in your neck of the woods. Watch and listen. Jot down anything that comes to mind. (I learned the hard way that notes are essential. Just how a blind person takes notes in a movie theatre I haven’t worked out yet, but not all listeners will be blind.) Drop me a line at description at joeclark dot org. All issues will be addressed.

Or you could go all the way and post your comments online. Heck, I do!