November 2008 Archives

Tonight, I won’t be at the Children’s BAFTA ceremony. It’s a long while since I’ve been, and I never quite managed to struggle to a nomination (harrumph). However, this year’s nomination list provides some indication of the state the industry is in.

Animation: all CBeebies/CBBC shows.
Drama: all CBBC shows.
Entertainment: all CBBC shows.
Factual: the perennial Nick News, plus three CBBC shows.
Presenter: all CBBC/CBeebies.
Writer: all CBBC/CBeebies.

Now, there are some caveats. Several of these nominations are for shows produced by indies (looks like a good night for Tiger Aspect, in particular). Also, other categories produce a stronger showing for Five, particularly, and Nick does well in the Short Film category (though… who else makes short films for kids?). Plus, one can argue that BAFTA nominations have always over-represented BBC shows, historically.

But is this still a picture of the dire state of the children’s television industry? Heck, yes. It’s entirely dependent on BBC money, and from where I’m standing there’s simply no way back from that.

I know I’ve banged on about this here before, but I keep running into people whose reaction to the story behind SciCast is disbelief. On a couple of occasions they’ve been almost belligerent about it.

They’d likely regard such a BBC-dominated nominations list as evidence that their license fee is being well-spent. But the fact is, in most of these categories, the only contest is which BBC shows are nominated. There is no competition.

Thus, there’s no longer a viable career path in this industry. That is why children’s TV is dead. It’ll take five years to really show — maybe ten, if we’re lucky — but it’s now inevitable.

[Update 1^st December: Winners now posted on the BAFTA site. Ironically, a good night for the few remaining indies.]

This should be a good way of wasting an afternoon: Auditorium.

Interactive Video Object Manipulation from Dan Goldman on Vimeo.

I’ve a sneaking suspicion that the processing overhead for handling this sort of thing — particularly with high-def footage — is ‘non-trivial.’ Which is to say: cripplingly hard. See also these previous stories, ditto.

Dual-processor laptops are one thing; eight-way desktops with hyperthreading, matched with stream-processing video cards, all coordinated via OpenCL implemented at the system level: well, that’s going to be something else when it comes to heavy-duty video processing.

Personally, I find it more than a little scary. We’re barely used to still photographs being routinely and heavily manipulated; processed video is something we expect of Hollywood, but not home movies. I suspect we’ll adjust less quickly, culturally, than the technology will propagate.

I’ve waited ten years to get a grip on how the various bits of the International Space Station (didn’t it used to be called ‘Freedom’? Ah, OK, no.) fit together. Now, for its tenth birthday, USA Today have made a lovely animation of how it’s grown.

I miss the canteen at Yorkshire Television.

(inspired by this)

Further to the previous post, I think what I’m trying to say is:

“The audience”

how old media projects from old media organisations refer to their most important constituency.

“The audience”

how new media projects from unreconstructed old media organisations refer to their most important constituency.

“Audiences”

how new media projects from old media organisations that ‘get it’ refer to their most important constituency.

“Community”

how new media projects from new media organisations refer to their most important constituency.

“Communities”

how new media projects from new media organisations that are so cutting edge they have no need for this sort of key refer to their most important constituency(/ies).

“Advertisers”

the people who are really the most important constituency for all the above, unless (media organisation) == ‘BBC’.

FindYourTribe is an online survey that claims to work out what social grouping you’re part of. It’s a bit of a giggle. It’s from Channel 4, despite lack of appearance on the front page (roll over the thinly-grey ‘About this game’ to learn that crucial snippet).

I rather enjoyed taking the survey. It’s wittily done, doesn’t take itself too seriously, and some of the questions are unexpected and quite clever.

I also think the whole thing is massive nonsense, and not in a ‘just a bit of fun’ reasonable sort of way.

My impression is doubtless skewed by two main problems: firstly, I’m entirely the wrong age for this thing (though there is a category for ‘35+’. Gee, thanks). Secondly, the three options it gave me at the end for ‘my tribe’ were all defined principally by music choice and hair style. Er… what?!

Of course, now I can’t go back and check what all the other tribes were, but I’ve a bigger problem with this: while I accept that youth culture is, to an extent, tribal, isn’t ‘tribe’ just a politically-correct synonym for ‘stereotype’?

Profiling audiences as a route to understanding them is, of course, entirely reasonable. But there’s something insidiously self-selecting about the presentation here, as if I’m supposed to funnel myself down into one of the predetermined stereotypes, or rail against the system for — OMG! OMG! Worst day of my life! — putting me in the wrong one.

Hmm.

Actually, I think what frustrates me most is that we’re seemingly stuck with blunt instrument tools like this, as we explore the intersection of centralised media and dispersed, interpersonal audiences. This feels like a tool from a previous generation. Wrapped in neat design and carefully-appropriate language choices, to be sure, but structurally the sort of thing the BBC might have done with a clipboard in 1986.

The problem with ‘the audience’ is that, to its members, it’s not ‘the audience.’ It’s ‘me, and my mates, and a bunch of people I don’t know but with whom I apparently have something vaguely in common, apart from that guy over there who’s obviously a tosser.’

And that’s only for physical groupings: for broadcast or web media, ‘the audience’ is usually, as far as I’m concerned, me. Just me.

While much of this new social media revolution might be about connecting me with people slightly like me in new and interesting ways, it’s still experienced by individuals. Lots of them. All alone. Simultaneously, but not together. To your servers they might look like ‘the audience,’ but in their heads they’re not.

Tools like FindYourTribe might be useful after all, if they help spot patterns of behaviour, broad groupings, and give a sense of the individual variation within groups. They can work as ‘pull’ models, where the media producer uses the stereotype labels as affordances to help understand audiences, and grapples with how their media might affect people within those categories.

Too often, however, the stereotypes become the targets for ‘push’ models, where one assumes the stereotype works and mercilessly tunes one’s media and its delivery to suit. That worked for broadcast TV, where the stereotypes were very broad-brush and the audience scale was immense; we have neither of those factors on our side with the sorts of things FindYourTribe is meant to inform.

It’s high time I introduced the regular reader to another project of mine, namely this blog: Why is science important?

Well, I say ‘mine.’ Really I mean ‘Alom’s’, since the project is his. I just threw together the blog. Alom’s a TV producer and physics teacher (odd combination, roll with it), making a film trying to explain to his students why science is… er… important. Funded by the Wellcome Trust, so there must be something in it.

The list of contributors so far is quite impressive, with more to come.

Join in.

Having been impressed by this ‘behind the scenes’ video, I found myself a little underwhelmed by the advert itself. Partly because I found it rather hard to track down. Go to this site, click ‘explore the site’, then ‘view the ad’ in the top menu.

Damned clever, but too clever, maybe?

[Update: good links and discussion on this at Metafilter, which led me to this article about kinetic computing.]

Following on from my previous post about the MinoHD, I happened to write this in an email for Vinay, who’s about to do some video policy work for an NGO. It might be useful more widely too:

The major limitations/trade-offs with the Flip concept are:

1. No microphone jack. You have zero options for good sound, unless you count off-camera recording, clapperboards, and post-sync.

2. Fixed screen means you more-or-less have to shoot from head height. You can’t easily shoot sitting subjects, or people interacting with props. It can be done, but handling in these situations is poor.

3. Close focus is limiting; big close-ups will always be blurry.

The Flip can do what it does in part because it makes a huge assumption: that 80% of what you shoot will be somebody speaking to you or the camera. That is, a mid-shot of some sort. Most of the time, this is a reasonable guess.

But it’s not a valid assumption for demo-based filming — the story is often told by the detail of the prop you’re showing, and the Flip starts to break down here. There are good reasons people use cameras costing £500, £5,000, or £50,000, and they’re not all about raw picture quality.

Example: want to film a mosquito? With a Flip you’ve no chance. The best camera I have for this is the £225 FS100, which close-focuses down to about 1cm (which is bizarre, actually, but I’m not complaining). It captured the image you see attached to this post. You can tell it’s a female, for heaven’s sake.

As the regular reader will be aware, I’m a big fan of the Flip Ultra camera, in part because its relatively large sensor offers surprisingly good performance in low light. The Flip is rarely a great video camera, but it’s a decent one in a wider range of conditions than anything else for the money. Kodak’s Zi6, for example, can produce some terrific images at much higher resolution than the Flip Ultra/Mino, but its anti-shake is lousy and its low-light performance sucks rocks.

Enter the Flip MinoHD. Andy Ihnatko has his mitts on one, the lucky git, and meanwhile there’s speculation that it might not share the Zi6’s gloomy grain problems.

Not so fast. The specs page does indeed quote better low-light sensitivity than the non-HD Mino, but those figures may not mean what one might assume. The HD has a physically smaller sensor, note, which means the individual pixels are dramatically smaller — if the ‘2.2µm’ figure is the length of a pixel’s side, they’re about one sixth the area of the previous model’s. That’s a lot less light per sensor pit.

So how can one explain the greater claimed sensitivity?

Gain. Which I’m guessing is what they mean by ‘automatic low light detection.’ No point detecting it unless you’re going to do something about it, and with just shutter speed to play with otherwise, tweaking gain is the only exposure control in hand.

Now, this may not be a bad thing. We’ve seen from Panasonic’s HMC150 that even quite high gain levels can give smooth images with H.264-derived codecs. We also know that the Flip folks have consistently made smart decisions in their software. But I’ll be waiting for the reviews before leaping for joy over this — finally, physics may not be on the Flip’s side.

Two other things to note:

1. H.264, yay. However, this may be the final nail in the coffin of Windows Movie Maker. The lack of reliable, compatible, and up-to-date entry-level video editing tools on Windows is baffling at this point. For all the grousing about iMovie’08, it’s a wonderful tool for lashing together footage from these sorts of cameras, and for getting it on YouTube, fast.

2. Without a microphone jack and a properly usable screen, this will still be a troublesome camera for filming things rather than people. For all its faults (notably: terrible low light performance), the Canon FS100 is still the cheapest flash media camera that’s properly versatile.

Mars Phoenix is, most likely, gone from us. I heard it from the lander itself via Twitter — its last tweet ‘Triumph’. In binary, obviously.

If you look, right now, at what people on Twitter are saying about the mission and the lander, you’ll see fond farewells and people tearing up more than a little. Examples: 1, 2, 3.

Meanwhile, here’s a terrific interview with the writer behind the tweets.

She’s very interesting when she talks about ‘getting into character’, and the interviewer’s exactly right to suggest that the way she’s approached it is highly cinematic.

Great stuff, and absolutely my favourite STEM Engagement project of the year: quick, relatively cheap, starts discussions and conversations, personal, large audience, emotional connection… the only problem is the heavily-skewed audience. But hey, Twitter geeks deserve STEM engagement too.

Google Maps has a new(ish) trick: directions for walking from place to place, rather than traveling by car. It’s unlikely it was intended to cover 400 mile trips between Glasgow and London, however, and the suggested route is, at first, somewhat alarming.

This reminds me of an old routing system I had on my first PowerBook, which covered all of Europe. In mere minutes it could calculate — with split-second precision for the journey time — the optimal route from Glasgow to Venice. By bicycle.

Since the average speed of a ferry is pretty good, it took ferries wherever possible, starting with Troon to Larne. It then struggled, rather, to work out how to get from Larne to Dublin without taking the motorway, and mysteriously added a loop around the perimeter of Northern Ireland before striking out to the republic. From Dublin to, I believe, Cork (a phantom ferry which I doubt exists, though I dare say one could arrange passage on a coaster if necessary), Cork to Swansea, …

Fifteen days (eight hours, thirty-seven minutes, and twelve seconds) later: Venice. Via Morocco.

I missed this news from IBC this year: JVC were showing a prototype/design study/mockup of a new very small/light SD-card camcorder. Small, that is, but still with ‘proper’ XLR audio jacks. Think ‘Sony A1 without tape’. Very interesting indeed — particularly since it’s badged ‘ProHD’, which implies MPEG2/HDV streams, not MPEG4/AVCHD.

Thread at the Panasonic discussion forums; video segment from MacVideo.

Also interesting: JVC have licensed SxS from Sony, for their high-end cameras. So broadcasters now either go Panasonic and P2, or Sony/JVC and SxS, with a range of cameras from either.

• Obama: 349
• McCain: 162
• Twitter: 538
• BBC sound engineering: -4096

I was very conscious, as I watched the BBC’s coverage, of three things. Firstly, that I should have gone to bed after Ohio declared. Secondly, that they were having horrific problems with audio, all through the night. Barely a link was survived without echo, mysterious noises off, one-way connections, a contributor or correspondent yanking their earpiece to avoid howl, microphones not being faded up or down correctly, or a remote cue being entirely missed.

Couple the all-pervasive audio problems with some distinctly squiffy video quality (some amazingly bad chromakey, very chunky compressed stuff in places, and some footage that was plainly transmitted in the wrong aspect ratio), and you have a bad night for the BBC technical department. Something of a rout, in fact.

The dodgy technical quality was mirrored, to some extent, in shaky production values. There were some excellent, fiery and insightful studio guests, and the BBC’s correspondents were, mostly, thoroughly professional (Katty Kay and Matt Frei, take a bow). However, the new touchscreen analysis battle room set they were trying out was, frankly, ill-conceived. Let’s not mention the ghastly sound effects and touch…touch…touch dammit touch screen, let’s rather point a scornful finger at the basic problem of using such devices on TV: the presenter spends most of their time with their back to camera. Gee, thanks, that’s lovely.

Couple with Dimbleby rather bizarrely moving the show along immediately there was any sniff of interesting discussion — usually to something banal, or with no audio feed — and the whole package was, frankly, pretty shaky.

Now, for the most part it wasn’t bad as such. But the show rarely rose to the heights we expect — nay, demand — of our national broadcaster.

Which brings me to the third thing of which I was critically conscious: this was, I believe, the last election night I will experience exclusively through ‘old’ media. I’d intended to lug my laptop into the lounge, but it was busy compressing video and my broadband was on the fritz anyway. So I didn’t.

And then I started to miss Twitter, and the US political blogs.

And then I got all nostalgic. I don’t remember even thinking about following the 2004 election online — I was BBC all the way. Now, suddenly, I’ve crossed whatever mental barrier was in place, and my starting assumption is that the BBC’s coverage is just another feed to place alongside all the others.

I realised, at that point, that I’m unlikely ever to watch election results the ‘old’ way, again. So I watched as much of US’08 as I could with a slightly forlorn sense of time being marked.

If it’s possible to be nostalgic for something while it’s happening, that was my frame of mind. Hence, I rather liked the technical goofs. They were part of that package, part of that way of doing things, part of how the world used to be.

Now we have fail whales and “500 Server Error” and spam storms. They’re just not the same.

Matt Webb likes plastic cows. I don’t know why. Neither does he. But if he receives 100 as gifts (specific kinds only), he’ll donate £500 to a choice of charities. Details here.

Yes, he really is this barking. Excellent.

Here are some clever maps of US voting. I remember admiring this guy’s work after the 2004 election, too, though I still find the red more visually-dominant than the blue, which skews my perception of the red/purple/blue renders.

More election maps collated by Kottke, here, who also links to some terrific photographs of Obama.

Chessboxing. Obviously.

Tractor, originally uploaded by Quernstone.

Perhaps my favourite photo of the year. A gift of a scene — the sort of moment where you look at your camera and wonder if it’s going to see what you see, because it doesn’t seem plausible.