[JDHill]

Micha :: this makes perfect sense. From the ground, you are not able to look 'down' on any droplets, therefore the rainbow ends as a semi-circle. From the air, this is not true, and you can see the full-circle. Also, a smaller version is easily simulated with a lawn sprinkler.

mmm.....summertime.....

[Thomas An.]

JD is right,

• http://www.popsci.com/popsci/science/12 ... drcrd.html
  Since each color is bent through a specific angle -- red light comes from the sun and is reflected back at an angle 42 degrees away from its original direction, while blue light bends only 40 degrees -- each color appears at a different place in the sky. Red, say, denotes all those locations in the sky 42 degrees away from an imaginary line connecting the sun to the back of your head. Together, these places trace out an arch. Since blue appears only 40 degrees from this imaginary line (which also connects your head to its shadow), the blue arch of a rainbow is always below the red.

Here is a trick question for you:
If we had a square sun, would we also have a square rainbow ?

[JDHill]

Thomas :: thanks for the science! I was really just taking an educated guess.

That said, here's my attempt to answer your question:

The shape of the Sun would make no difference, as the shape of the rainbow is determined by the fact that the water droplets are spherical. Each photon does not 'care' about the shape of its' source.

[edit: To clarify, the Sun>droplet>eye triangle is sufficiently acute to render the Sun>droplet vs. Sun>eye sides effectively parallel. Factoring in that the 2 major sides of this triangle measure around 93 million miles, the Sun could also be several times its' width, as well as various shapes, with little to no perceptible difference to the observer.

Taken to an extreme size, I theorize that the observable effect would be a 'thicker' rainbow, that is, each color would be viewed as being physically wider.

I'll stop now. I'm having fun, but I'm a little out of my depth, and just winging it at this point. What's the old saying, "Better to remain silent, than to open your mouth and remove all doubt." ]


~JD

[Thomas An.]

Hi all,

To bring a conclusion to this discussion... I have discovered that beam arching occurs in nature too.


Basically this is because, the ray coming from the center of the rectangular emitter is the only "perfect" ray which hits the prism at the designed angle... all others rays do not refract the same because they hit differently on the prism surface (slightly different angles for rays traveling diagonally from the lower corner of the emitter and exit from the upper corner of the collimator apperture) and thus they bend in two directions (at composite angles).

JDHill helped with this explanation which I feel is quite plausible.
(The sky-rainbow physics is a seperate matter)

Solution:
We can avoid arching by adding more divisions in the collimator:

[JDHill]

Say 'collimator' again. Go ahead, brother. Try it.


_JD

[Thomas An.]

Say 'collimator' again. Go ahead, brother. Try it.


_JD

You bet !

I was looking for this word .... so .... now that I found it, I will spam you with it

(possibly jack threads randomly too; just for the heck)

[tom]

So, what about onions Thomas

[Thomas An.]

So, what about onions Thomas

Well its the "wavefront" theory versus the "geometric optics" explanation.

if you look at the geometry involved in the wavefront approach, it boils down to the same conclusions (you are going to use geometry to analyze it anyway). So both approchses basically discribe the same thing in different words.

[tom]

If you go with the "wavefront" theory then it should not matter how many divisions there are inside the collimator.

Why not? Collimator would straighten any incoming ray, isn't it?

[Thomas An.]

Hi all,

Since some people from time to time send mail asking about the setup of this scene... Here is a sample mxs file for Maxwell 1.0 to get you started.

http://rapidshare.de/files/20124664/Sam ... 0.mxs.html

[-Adrian]

Thanks a lot Thomas, very generous of you.

I mirrored the file so people don't have the hassle to wait for Rapidshare.

[Thomas An.]

Thanks a lot Thomas, very generous of you.

I mirrored the file so people don't have the hassle to wait for Rapidshare.

Thanks Adrian

Yes, Rapidshare is a pain

[Thomas An.]

Dispersion physics are vastly imporved in V1.1 versous beta 1.2.2a

[ivox3]

I'd say.

How long has that been in the oven ?

[4 HeRo]