The Daily Grind

There are times when the BBC looks like a blundering behemoth, not so much intent on world domination as accidentally trampling the rest of the industry as it fails to see beyond its own nose. Then there are times when it demonstrates that it really is the global centre of excellence one would hope and expect. There’s no dichotomy here, of course: the BBC is vast, it has many heads, and at any given time a statistically significant number of them will be facing precisely the wrong way. So it’s pleasing – and not a little scary – to come across a situation where the BBC isn’t merely looking in the right general direction, but is striding onward, leading a charge into the bold new (media) future.

Thanks to Tom Loosemore, here are the BBC’s Fifteen Web Principles, as signed off by the BBC board. There are several things to note:

1. None of the things listed will come as a surprise to people who’ve read the Cluetrain Manifesto, follow Euan Semple, chuckle at Hugh MacLeod, marvel at the talented Mr Hammersley, and generally hang around getting excited about blog-type stuff.
2. It’s nevertheless nice to see a concise summary that’s also practical.
3. This was, apparently, signed off by the BBC board.
4. I’ll repeat that, for the hard of nuance:
5. This was signed off by the BBC board.

I’m currently working with a small handful of fairly sizable public bodies, and last week I was talking to a whole bunch of visitor centres and museums. And you know… the BBC gives every impression of being waaaay ahead of pretty much everyone else on this stuff. There are occasional flickers that suggest other institutions’ interest is piqued, but there are serious language and conceptual hurdles.

As Euan says, it’s not really about people who ‘get it’ vs. people who don’t. It’s about those of us who think we see something – glimmering in the light, just out of reach – managing to express that clearly enough for others to catch a glimpse of our perspective. Then it’s about holding onto that vision for long enough to define and build something around it, so the real decisions can be made.

After all, it’s not clear that the BBC is right. As it strides boldly on it could well be flattening perfectly good alternatives underfoot, or leading us all over a precipice. And then we’d look proper silly.

But until then, documents like the Fifteen Web Principles are tremendously valuable. They’re pragmatic, straightforward, plain-speaking, and reasonable. What’s clever is that implementing them requires a fairly radical approach to engineering, organisation, and editorial management. If you accept these principles, doing something about them – doing it properly – is challenging.

Case in point: with SciCast, I reckon we get a partial score on less than half of the principles. Which is better than it might have been, but not as good as it needs to be.

The unlikely-named Steve Spangler is big in the world of Diet Coke and Mentos explosions – I believe he was doing it before the Eepybird boys – and he managed to persuade the company to spring for a high-speed camera. The result? Mentos in a fizzy drink, filmed rather artfully and at 2,000 frames/second. Nice. Video here.

[hat tip to Paul for the link]

Stewart Copeland writes about the second night of The Police’s new tour. Well worth a read; anyone who’s been on a stage will recognise and understand.

Given that YouTube is stuffed to the gunnels with lip-syncing videos, I probably shouldn’t have found this one quite so entertaining. I think it’s the way-better-than-average production values, and the clever use of props. Besides, how many lip-sync videos include catapult take-offs?

Sheep racing, duck trials, and shearing exhibitions compete with the delights of ‘ewetopia,’ at The Big Sheep. It’s big in Devon. Apparently.

Tee-hee. I’d completely forgotten the run of drafts in which the lecture about food allergies and hazards was based entirely around the different ways in which a Christmas Pudding could kill you. The very opening of the lecture involved us dropping a giant Christmas Pudding to land SPLAT! right where Krebs had been standing a split-second earlier.

Cartoon violence rocks.

The use of ‘mike’ instead of ‘mic.’ It’s a shortening of ‘microphone.’ Tsk.

(example in this article, which I’m reading because it was linked from Robert Sharl’s post here, on which I can’t comment because I don’t have a Google account, or something. Robert: you’ve seen the video from TED of the Microsoft SeaMonkey/Photosynth demo, right? If CoreAnimation makes that sort of thing easier to implement, I’m all for it. Yes, we’ll get pointless crap too, but we have that now, it’s just uglier than it will be.)

I’m wading through piles of paper trying to find a specific document that I must have mis-filed/incorporated in a pile of ‘not worth filing’ stuff/generally misplaced/etc. Instead, I’ve found a bunch of documents pertaining to the 2005 Christmas Lectures. Specifically, early drafts of the scripts, with scribbled notes from rehearsals and run-throughs.

Captured in the margins are moments of insight. Such as: when the opening line of Lecture 3 changed from ‘Look at your arm!’ (no, really – and no, I didn’t write that) into ‘You’re all turkeys!’. Which I think ended up as ‘You’re all vegetables! And some of you are turkeys!’, but you get the point.

Then there’s the musing where James asked, of an entire lecture, ‘What’s the question we’re addressing here?’, and I recall sitting bolt upright and saying ‘Not quite the right question. It should be: what’s the perspective shift we’ve inspired?’, that led to refocussing the entire lecture around the rôle evolution has played in shaping our diets (even later drafts defocussed the lecture again, but that’s another issue).

There are the parenthetical comments that note the first appearances of The Goat, and The Squad of TA Soldiers (both eventually dropped again), and the moment I first dreamed up having a mad machine serve Krebs ‘the pill on a plate’ so beloved of 60s Sci-Fi, that ended up as the opening of lecture 5.

This is how I like to remember the creative process. Rewrites are painful, but the moments of inspiration, of throwing out cruft and fixing something, of refactoring entire trains of thought – those are magical events.

Writing at that sort of level, with that sort of care – agonising over both individual words and the shape of sentences, about how paragraphs flow, and about driving driving ever onwards whilst simultaneously coaxing contrasts of intense grandeur and amicable ease – it’s hard.

And I like it.

I should do it more.

One of the problems we’re facing with SciCast is that just as video is becoming fairly universally accessible, the manufacturers are retiring the only type of low-end camcorders that really make any sense – mini DV. Sony, JVC and Canon each have remarkably few consumer DV cameras on the market, as they push towards supposedly more modern formats like recordable DVD-video and, particularly, hard drive recorders.

They’re all perfectly reasonable acquisition formats, but if you want to make films rather than force people to suffer through every second you shoot, you need to be able to edit the result. And this is where the non-DV formats rather fall down.

DVD-video is a lousy post-production format but still arguably better – at least for now – than the format used by Sony and Panasonic hard disc recorders, the snappily-named AVCHD. (just to be really confusing, there are also AVCHD camcorders that use data DVDs as the storage medium, or even flash memory cards)

The trouble with AVCHD is, ironically, how fantastic it is. It’s an MPEG4/H264 variant with an extremely large frame size. Which is, in short, crippling for current PCs to decode. Anecdotally, systems that can handle four streams of the MPEG2-based, DV-derived HDV format can just about handle one of AVCHD.

Playing back the video on your PC isn’t where your problems end, however – edit system support is sadly lacking. Sony’s Vegas editing package is supposed to handle it, and I hear rumours that Ulead Videostudio can too. What we really want, however, is support in iMovie, Final Cut Pro, and Adobe Premier. None of which is in place at the moment.

Right now, the best option is to – I kid you not – pull in full uncompressed HD via an HDMI capture card, but that’s hardly realistic for a supposedly domestic format. There’s an interesting thread here about forthcoming AVCHD support in Final Cut Studio, but it’s equally clear that the strain of cutting AVCHD material is going to drive demand for übermulticore desktop computers well into the end of the decade. Right now, the only viable option is to re-render captured footage into some intermediate codec for editing. Which isn’t something you want to talk your granny through. There’s no word of support in the likes of iMovie, either.

So we’re left in the faintly ridiculous situation that editing video is today harder than it was a couple of years ago. We’ve a plethora of mutually-incompatible formats, excessive render times even on the beefiest of desktop PCs, and still not many cameras with proper microphone inputs.

Personally, I’ll take cheap, reliable, and decent-quality DV over any format that’s theoretically HD but practically unplayable except in unedited form, direct off the camera. Harrumph.

(though I wouldn’t mind me some HDV…)

My hate/hate relationship with my N95 continues. I was showing it to Matt tonight, and it crashed within about 30 seconds of my starting to demo it. He was actually quite impressed; from the way I’d been talking, he’d expected more like 15 seconds.

Anyway, catching up on feeds tonight (via the wonderful new NetNewsWire 3.0), I find that Scoble has blagged himself an N95. One of the problems I’ve been having is finding sensible software repositories for Series 60 apps to try out; one of Scoble’s comment threads helps there.

Incidentally, I’d agree with Scoble that the phone’s camera quality is excellent – noticeably better than the not-at-all-sniffy K750, and Nokia have toned down the somewhat crude over-sharpening their previous high-end Zeiss-lensed überphones applied. However, there are two problems with the N95 camera; shutter lag is pretty bad, but more of an issue is startup time. I count it at about five seconds, which is extreme. I’ve started flicking the lens cover open while the thing’s still in my pocket, as I’m walking up to where I think there might be a shot. Once I’m in place, I still have to count to three before the screen’s up and stabilised. Ugh.