Category: Uncategorized

Apple will doubtless sell a boatload of the iPod mini, despite it being only $50 cheaper than the 15Gb version – it’s so darned cute. And they’ve ticked all but one of my ‘things I’d really like iMovie to do’ boxes, the most important being ‘rethink the way clips are trimmed and do it better,’ which will make a huge difference to the scope of project it’s sane to do before falling back to Final Cut Pro. Very impressive. Then there’s Xgrid, not mentioned in Jobs’ presentation but the sort of thing that makes the ex-supercomputer researcher in me perk up and take notice. I’ll be interested to see where that goes.

Only a few disappointments: no word on the iTunes Music Store for the world beyond the US, bah! No hardware updates beyond the (much-anticipated) Xserve G5 machines, which is a bit of a pity since I’m in the market for a laptop in the near future.

But most bitterly: Douglas Adams isn’t around to play with GarageBand. Until people get their mitts on it, it’s rather hard to tell, but it looks like it’s around the iMovie sort of level, possibly a bit more advanced – so, complex enough to do real work with without losing approachable simplicity. Limited, sure, but Apple’s genius with the iLife applications was to realise that, sometimes, if you get the balance right, ‘limited’ is enough.

One thing, though: what’s with the neither Aqua nor Metal ‘dark metallic but not quite Final Cut Pro’ interface stylee? Did we really need a fourth ‘standard’ Mac OS X GUI? Harrumph.

[update:] Everybody’s favourite shareware author, Brent Simmons, is clearly excited about GarageBand. ‘I’ve waited 20 years for this,’ he says. John Gruber, meanwhile, points out that it exemplifies exactly what Apple is all about.

[update 2:] Er… they don’t call it the ‘iPod mini’ for nothing. Want to see it alongside an Ericsson T68i? Try here, here, and here. That’s tiny.

YourParty.org (should be .org.uk, but there we are) – interesting. I’ve just completed their pre-launch survey, including a stream of vitriol in the ‘comments’ box about how appallingly-written it is. Talk about forcing people down predefined avenues… blimey.

Have a look-see. It’s a probably naïve but still fascinating idea, but if they’re not going to think it through any more than that, I for one will not be involved. Bah! Market research! Bah!

See also this.

One of these days I’ll get around to writing up some notes on my bike, since lots of people seem to arrive here via a Google search for ‘Ridgeback Genesis review.’ In the meantime – firstly, it’s wonderful and I love it, and secondly, I’m slightly surprised by the new 2004 model range.

For the 2003 models, the choice was really between a Day 1 and a Day 2 – the 2 being a slightly more hardcore ‘fast road’ machine, with a lighter/stiffer frame and better components throughout, notably a much fancier set of wheels, but no mudguard eyes etc. I bought the 2 partly because late in the year the price dropped by a hundred quid, but really because it was available in my size. Which nothing else was.

This year, it looks a little like Ridgeback have decided they’ve got the brand name, so they’re building down to a price. The 2 is pegged at £700: what I paid, and hence cheaper than last year’s model. It gains carbon forks – probably a wise move when the competing Specialized machines all have them, though I don’t find mine’s alu-forked ride harsh – however, it loses the fancy triple-butted frame, which is now exclusive to the top-of-the-range Day 3. Since that model is now more expensive (using a mix of Ultegra and Dura Ace rather than 105), there’s a gaping hole between the two models. The Day 3 still isn’t available in a 60cm frame, either.

The Day 1, meanwhile, remains a cost-effective buy, but continues with what I’d class as marginal brakes – I wasn’t impressed when I rode them, whereas the Tiagra units on mine are better than I’d thought sidepulls could be. The wheels probably play a part here.

Was it worth Ridgeback dropping the fancy triple-butted frame for the sake of keeping the price point? It’s hard to say. Critically, the 2004 Day 2 has mudguard eyes, which to my (unhip) mind is a significant advantage. But I’m slightly surprised there isn’t a ‘Day 2.5’ model, spaced between the Day 2 and Day 3, with 105/Tiagra bits on the Kinesium frame.

Perhaps Ridgeback are planning a range of drop-bar bikes? Or maybe they’re not making any of the triple-butted frames this year, and are simply using up stocks from last season, limiting sales by slapping high-end kit on them? Strange.

Crivens! Vinay’s blog front page usually scrolls so far you can’t see the horizon; I habitually pester him to limit the number of posts rather than the number of days’-worth of posts. But now I discover that he’s not posted for so many weeks, the layout’s fallen apart.

Colour me worried.

An amusing reference spotted on Joi Ito’s blog, with the title, “Oh no! My Tivo thinks I’m gay!” I have similar sorts of problems with Amazon, in that it can’t seem to get its vast electronic brain around the combination that I (a.) have given up being a hardcore web programming geek, at least for the moment, (b.) am fascinated by graphic design, especially fairly esoteric typography, (c.) buy the novels of a few specific hard SF authors, (d.) look at a fairly wide range of history books, mostly monographs about weird things like waterwheel design, or the phlogiston theory, and (e.) buy the sort of contemporary literature that gets reviewed in the Guardian.

Don’t get me wrong – Amazon’s recommendations are at worst amusing, at best downright useful. But it strikes me that most of us are not a single person – or at least, should not be considered such for the purposes of profiling. Amazon, Tivo et al should be looking for instances where choices fail to fit the profile, and forking off a new persona rather than trying to integrate the two. Maybe they do, I don’t know – it doesn’t seem to work all that well, anyway.

Hmm. I could get rather interested in how the profiling is parameterised, come to think of it. I wonder what Amazon would make of that.

Tonight was the opening of Glasgow’s newest cinema, the Tantallodeon. In other words, I had some friends round to watch a film, but we had trailers and adverts and supporting features and everything – iDVD is a seriously nifty piece of software.

The feature presentation was provided by Colin, who brought along Roman Polanski’s 1974 masterpiece Chinatown. I’m surprised I’ve never seen it, but there we go.

A fine film; fine company; fine… well, fairly decent wine; extreme cheese on toast (with very fine truffle sauce). Oh, it’s a hard life. Meanwhile, I’ll be blogging more regularly again once How2 is finished, in a couple of weeks’ time.

I’m fed up doing the same calculations every six months, so I’ll post them here in the hopes I remember where I left them. I’m also fed up with whoever decided that television pixels would be rectangular, and whoever decided not to include any conversion system in QuickTime, but there we go.

Conventional 4:3 aspect PAL transmissions are taken to be 720×576 pixels. Which the Carol Vordermans amongst you will instantly recognise as being 5:4, not 4:3 at all, so to make a full-frame PAL capture look right on a computer monitor, it should be scaled to 720×540 pixels.

Now, a 16:9 widescreen capture is still stored at 720×576 pixels, simply stretched more at the display (which is the sensible approach when you can’t change the transmission bandwidth, to be fair). So to make a widescreen frame display right on a computer, scale it to 720×405 pixels.

Some notes: I’m assuming the pixels on the target computer are square – It used to be the case that Macs would only do square pixels, but now we have more choice of monitors there’s more scope for the user to screw things up. 1280×1024, for example, isn’t the same aspect ratio as 1024×768, so it’ll look squished on (4:3) CRT monitors; 1280×960 is a ‘better’ resolution. Windows users have been able to mess things up like this since SVGA came along, of course.

Also, you may wonder why I’m only scaling the vertical resolution – why not interpolate horizontally rather than throw away vertical data? Well, because the vertical data isn’t really there in the first place; the highest-resolution capture format that’s common is (so far as I’m aware) DigiBeta, which is what most of my shows originate on. DigiBeta record decks typically handle about 500 vertical lines, maximum. When running widescreen, many cameras simply subsample the (4:3) CCDs, giving an even lower ‘true’ vertical resolution. Which sounds ghastly, but bear in mind that people happily watch VHS recordings, and VHS staggers to only about 240 lines vertically; even DV is only about 360 lines, I think. And funnily enough, all the stuff I’m playing with is DV-sourced; so scaling to 720×405 is still interpolating upwards from the true capture resolution.

There. I hope that’s clear. The magic numbers are 540 and 405; on my monitor, they look right. Thanks.

Merry Christmas!

Back in a few days.

Have spent the weekend thus far recovering from the rigours of TV production. Which for the most part means that I’ve been asleep.

I am starting to catch up with things, though. I have a very long ‘ToDo’ list (very long), and the number of unread entries in NetNewsWire is down below 800 for the first time in about three weeks.

Happy find so far: this entry at Typographica, leading to the fonts Gararond, Moonbeam, and Salmiak.

It surprises me that the BBC doesn’t do more of this sort of reporting online. It’s a terrific combination of world-class journalism and the immediacy of the web, the former being particularly significant in a mediasphere increasingly under threat from ill-trained bloggers who make up in enthusiasm what they lack in rigour.