On 28/04/2014 08:59 AM, scott wrote:
>> Presumably what you do is make a translucent layer which has inclusions
>> inside it that refract the light. So the outer surfaces are smooth, yet
>> you still get colour. Hard to see how you could do that cheaply though.
>
> Well you could just make it in two parts, so long as the two materials
> had different refractive indices. A bit like how a CD is made. Maybe you
> could use some photo-etch process on the first layer, borrowed from IC
> manufacturing?

That sounds quite plausible. Although probably not as cheap as just 
spraying some dye onto things, which is what we do now.

>> Has anybody invented a UV coating that actually works yet?
>
> Yes, people rely on them not to get skin cancer or go blind! Also the
> one on your car paint works. I'm sure you can find spectral response
> curves if you are interested.

Interesting. I was under the impression that UV-resistent coatings are 
just overly-optimistic marketing, and in reality none of these coatings 
actually work.

>> I'm sure things have changed. Just perhaps not as rapidly and
>> dramatically as the original revolution.
>
> You could argue that in some ways things are changing even faster now.
> Continuing with your plastics example, at the start hardly anyone used
> plastic and development was slow. Today plastic is used in almost
> everything and there are millions of different types. I'm sure if you
> plotted a graph of the number of new plastics invented against time it
> would be going upwards.

Really?

OK, really ancient plastics weren't very good. But the plastics I see 
today and the plastics I saw 30 years ago seem pretty much identical in 
every respect. What's changed?

>> This is very interesting. From what I can gather, the primary problem is
>> the absurd amount of computer power required.
>
> Indeed, it looks like they had to spend most of the effort trying to
> reduce the computing power needed whilst maintaining a decent image
> quality. Maybe in 5 or 10 years someone will come back to this
> technology and it will be able to take off.

Much like how we had CDs for decades before DVDs were invented. Clearly 
a lot of people *wanted* DVDs, but the necessary computer power was 
simply unavailable.

Then again, we're trying to simulate the interaction of billions of 
photons in a large 3-dimensional space. Maybe there are 
quantum-theoretical reasons why you can't do that any faster than the 
real world can? (At least, not without using a computer many times 
bigger than the real-world system it's trying to simulate!)

Post a reply to this message