We Need an Aesthetic for Technical Interview Video (hurry, my eyes are bleeding)

OK, done? I'll be interested to see the percentage of those who watch to those who don't. I see a lot of these developer-centric videos as I'm accumulating potential links for the java.net front page, and sometimes I'll use them when I think the topic or speaker will be interesting to our audience at that site.

But jeeez, I can't stand to watch the stuff myself. And I can't understand why it's spreading.

Well, OK, I can sort of understand the spread, from the producers' point-of-view. The YouTube-powered acceptance of Flash as a low-hassle web video delivery mechanism has people thinking about video who wouldn't have considered it as a content offering 18 months ago. There's also a popular belief that video, being a "richer" media type than web pages or audio podcasts, is somehow inherently more valuable or just better. It's easy to walk into this by applying the "a picture tells a thousand words" analogy.

Sure it does, and that's the problem. Video is not just higher-bandwidth in terms of CPU resources or Ted Stevens-esque "pipes", it also demands more of the audience. Text can be scanned, audio attended to while doing something else, but watching a video requires the near-exclusive attention of the viewer. It potentially delivers a great deal of visual information. Unfortunately, a static shot of someone talking for 20 minutes doesn't deliver information. After 15 seconds or so, I'm pretty sure I know what Joe or Jane developer looks like, and really don't need the video channel anymore (experts may see an interesting angle with the predictability of boring head-shot video as being a perfect example of low information entropy, and in communications, entropy is good and predictability is bad).

When thought of in the terms of the viewer's self-interest, a lot of these videos come up really short. I can get a lot more out of reading Romain Guy's Curious Creature blog for 10 minutes than watching an unedited single-camera video of him for that length of time. Sure, the video gives me a sense of his personality, his accent and style, that the blog doesn't... but is that enough to merit the investment of time? I don't think so.

I think that a lot of people are using video for the wrong thing, and that's why these videos are so boring. To figure out what we should be doing instead, I look to an interesting aesthetic theory in Scott McCloud's Understanding Comics that breaks down any art into six "steps":

1. Idea/Purpose - The ideas, philosophies, emotions, etc. of the work
2. Form - The specific form of the work (a video, a comic book, a blog entry, etc.)
3. Idiom - The genre of the work within its form (exposition, parody, narrative, etc.)
4. Structure - Composition, ordering, what to leave in and take out
5. Craft - The practical knowledge/skills of creating the work
6. Surface - Production values, finishing touches, superficial appearance

As McCloud points out, almost nobody starts out at step 1 by saying "what do I want to say, what form should that take, etc.". More typically, people see step 6 ("ooh, shiny online video") and dive in, with absolutely no idea of the underlying principles. For McCloud, this is the teenager slavishly copying a picture of Wolverine; for our purposes, it's pointing a camera at someone (probably using ambient light and a camcorder's built-in mic, since awareness of lighting and sound implies step 5 skills), recording them for 10 minutes, compressing it and uploading it.

Given that this is where most web exposition is today, can we at least start to move back through the other steps? It's starting, but it's a long road. The NoFluffJustStuff tour recently posted a truly strange Groovy/Grails Discussion video in which the participants are shot one standing in front of and aside the other, in front of a blue wash. The lighting is atypically good for web video (the light on Venkat's shoulders and hair suggest he's properly back-lit), and the staging shows an awareness of depth-of-field. But on the other hand, what's the point? The staging is wildly unnatural, and having the non-speaking participant just standing there in the shot, staring forward, seems pointless. The producers haven't made it back to step 4: understanding the structure of effective video.

Not that this problem is limited to web video. If your cable or satellite provider wastes bandwidth on the G4 channel, take a look at the Cinematech program sometime. This show consists of edited two-minute segments of videogame footage. The proper question to ask is: what's the point? What is the value of watching game footage when you could just as easily be playing those games? Is it meant to be promotional, or artistic, or what? It feels like an acontextual data dump, just meaningless. I don't know who watches it or why it's on (if there is a point to it, please enlighten me). Again, they know enough to capture video, assign some kind of structure (step 4), but I'm not sure there's any thought about idiom (step 3, how the viewer is being addressed), or the purpose of the exercise (step 1).

For my money, the rare example of web video that seems to have thought through the entire series of steps is ZeFrank's TheShow. You get the impression from this that he made an explicit choice to deliver his messages through the form of online video, used a somewhat-affected (effected?) direct-address idiom (his jump cuts of himself might just be remembered as the "ZeFrank effect"), he understood what he could and couldn't achieve with the equipment, skill-level, and time available to him, and built his shows around that, and he wrapped it in a handy RSS feed. I'm not saying I want all developer interviews to look like ZeFrank, but things will get better when creators better understand the creative process and value-proposition of video, and think things through.

A good example of walking the six steps is actually Sun's Level Up! video series. You don't have to like it -- the reviews on iTunes are quite mixed -- but at least a competent set of decisions has been made. The purpose of the show is to discuss the games industry and advance Sun's products within that industry, the idiom is an interview talk-show wrapped by news bits presented by the host, the production is lavish by web standards (a real set, well-lit, tight editing), and the final product is quite polished. It clearly borrows much of its inspiration from television idioms, and it might not be a good example to follow because it is clearly an expensive proposition (I count 15 people in the end credits). But at least they've thought it through, and the result is atypically watchable.

One thing Level Up does right is to play up the personality of its host, Chris Mellisinos. This is something a lot of "serious" web videos choose to avoid, and that's probably a huge mistake. Perhaps the most pathological case is that of Robert X. Cringely's NerdTV. Despite the fact that Cringely was telegenic enough to host his own PBS specials on TV, he hides off-camera in his own web videos, remaining an inadequately-mic'ed off-camera questioner in this single-camera interview series (which often run over one hour per installment). What a bizarre choice! He's good enough for TV, but not for low-budget web video? Is there something in his contract or union rules that keeps him off-camera? It's crazy!

If part of the value of video is that the viewer can have a better sense of the personalities involved, then it seems almost obvious we need to see those personalities on-screen. In fact, it's fairly typical in other media for the host to be more of a draw than the guest, particularly in content formatted as a series. You don't hear too many people tuning in to The Daily Show because "they've got former Secretary of State Lawrence Eagleburger tonight, and Todd Rundgren later in the week!"; no, you watch because Jon Stewart makes the interview entertaining. Instead of hiding off camera, our technical interview video producers need to either find some good on-screen talent to use as interviewers, or become that talent themselves. And I think Amanda Congdon's already taken, sorry.

We're at an early stage of web video, with YouTube as a modern day nickelodeon, showing little novelties like what happens when you put Mentos in Diet Coke. It took a while for film and television to develop a visual language for exposition, too. But let's at least agree we're on the road to something better, and are determined to get there, because right now, the fact that technical discussions and interviews are available as online video is little better than novelty, and that won't last.