Random Thoughts

From PEL Wiki

1 Graphics Card Motion Blur
2 Soft White Lights
3 Muffler
4 Trade Deficit
5 Handheld Projector
6 OLIN
7 Brain Computations
8 Simulations
9 Doesn't fit
10 Conference Tidbits

Graphics Card Motion Blur

I've always thought it would be fun to create motion blur instead of continually bump up frame rates. One reason why this is attractive is because of the comparitively slow refresh rates of LCDs.

Could you implement motion blur with a decaying alpha channel? How would you designate the polygons to be blurred? Would each one need an ID?

Are there simpler ways such as having motion vectors for each polygon?

Would it actually be less overhead than rendering at 85 fps? At least on average?

[edit]

Soft White Lights

If you partially silver a light bulb, does it produce an efficient soft light?

[edit]

Muffler

Can you create a muffler that has multiple chambers that cancel each other out with resonance?

[edit]

Trade Deficit

There's an interesting relationship between the national debt, trade deficit, and the worth of my assets. I don't understand it as well as I'd like to, but here's a link to Warren Buffet's perspective on continued deficit spending. Squanderville Article

[edit]

Handheld Projector

Using random vibration + modulation of the light source can you make a cheap handheld projector out of a laser pointer?

[edit]

OLIN

There's an interesting university experiment called OLIN in the Boston area. They are trying to give a good engineering education starting with projects instead of starting with theory. There was an article in IEEE Spectrum - May 2006 Table of Contents

If nothing else, it sounds fun. Since that's their aim, I think it's great.

[edit]

Brain Computations

In the same issue (May 2006 Spectrum Table of Contents) there was an interesting article about mimicking the computation of nature. They claim that we do digitization too early, and waste resources because of it. If we use analog processing, then digitize the results, a cochlear implant becomes truly implantable. Pretty cool. It seems like there are lots of areas where we could encode information more usefully and efficiently than we do if we take into account our biology.

[edit]

Simulations

I think simulations are intriguing in that they can produce a lot of data, but it is very difficult to verify their results. I think that would be a fun challenge to tackle in an abstract sense.

I think it would be fun to develop a simulator for pruning, which would allow you to progress through several seasons and see the effects that different pruning styles could have on your orchard.

[edit]