In a comment on Women are not 1-dimensional… Michael wrote:
you go from needing maybe seven different sizes (even numbers 8-20) to… Well, with just four variations in each dimension that’s 4,096 different sizes
And he is of course correct.
What I was thinking of was something like this, only much less expensive. The linked system builds a 200,000-vertex 3D model of its customer then, with a bit of manual help, works out how to adapt a suit-design to best effect for that subject. They’re using some funky light pattern techniques to get high accuracy, but I reckon a sufficiently accurate model could be made from a few images from different angles recorded on a reasonable-resolution camera. Some kind of ludicrously high-powered vector-processing hardware would also be useful here.
Given the model, and information on how a given style is meant to fit, it does not seem unreasonable that a sufficiently good computer program could calculate the pattern and assembly instructions for a garment which fits.
Aha, you say, how will shops work: they’re not going to like it if you keep getting things made for you then not buying them. I suspect that they would need to keep the existing size system around, and you would try one of those on to see how you liked it before having one tailor-made: whether contractually or by social pressure you’d have to buy the tailor-made one unless it was clearly defective.
So, what is required to build this dream? And how much of it do you, my collective readers, and I already have? I reckon we’d need:
- A reasonable-resolution camera (check: I have a Canon EOS 300D lying around doing nothing useful).
- Some kind of ludicrously powerful vector-processing hardware (check: PS3).
- Some genius programmers to code this up (check: Mark, Michael, etc).
- Women of varying shapes and sizes (check) who’re dissatisfied with the status quo (check) and who don’t mind being photo’d wearing not very much in the interests of science (doubt it).
- An automatic device which converts cloth and patterns into clothes. These do exist, but I don’t know anyone who owns one.
- Someone who knows how to design clothes that fit.
- Someone who can explain that to a programmer who can get a computer to do it.
I think the last two are the trickiest.
Anyway, what do you guys and (more importantly) gals think? Should we give it a go?
Nah.
Once the women had gone through all that, they’d still take the item back the following week.
Hehehe!
I think it’s quite a good idea for more “special” or specialised bits of clothing, but is probably a bit more trouble than a T-Shirt or jumper is worth…
However. How about a standardised set of measurements made by a system like this that could be used for any item of clothing? You’d only need to be measured once, and then Example Woman could go to the shop whenever, see a bunch of things she’d like, buy them and give the shop her size data, and then a whole batch could be made at the same time.
Maybe that was what you meant.
A disadvantage is, I suppose, that people tend to change shape over time, and clothes fitting very close to perfectly could exaggerate the imperfections caused by normal shape-changing (even if it’s quite slight and shortlived, like post-Christmas podginess or whatever), so it could reduce the lifespan of clothes considerably.
Oh, and if you need something to wear tonight…
So not a total replacement to off-the-peg clothes.
I think recording sizes for later use is a good way forwards; it could even allow your average bloke to buy clothes as presents for his special lady without giving away the surprise or coming home with something hopelessly wrong. Well, maybe that’s not true, but at least it’d have a good chance of fitting properly.
Clothes which once fit properly and are now wrong due to body-changes should look at worst as bad as off-the-peg clothes which sort of fit now; it’ll certainly seem much worse by contrast, but I don’t think it’d be worse than currently.
There’s no reason to limit this system to women, of course, other than their current situation being worse than ours. I suppose I could test my body-modelling software on myself if it came to it…