Semi-Translucent Colored Object Filters Light?

[Twinkle]

The new "laser" objects are definitely fun to play with. I was wondering how hard it would be for objects with opacity<1 and saturation>0 to filter the various colors of light based on hue and value. For instance, a very red, translucent object could filter out much of the other colors, especially depending on how long the light stays within the object.

Or, if the other colors would be reflected more, we could demonstrate sunsets with a bunch of air particles that reflect blue and transmit other colors. Shining white light into the particles should disperse blue in random directions while allowing for a "yellow" sun beam to pass through. The deeper the beam has to travel through the material, the more orange/red it would become.

The acts of absorbing certain colors and reflecting vs. transmitting certain colors actually seem to be two different variables to consider. I'm not sure how they could be combined, so I guess this is really two different suggestions.

[admin]

Well, this filtering does occur and should work in moste cases. Start with white light and you should find that different wavelengths are affected in different ways. We use a generic model for this, that may not representative for all materials of course.

However, some phenomena, like when light passes through small particles (or the size of the wavelength of light) in the atmosphere (dust, pollen, smoke, water) are real hard to simulate since they require a really large number of internal reflections to give the correct result (this is referred to as "Mie scattering theory" and it relies on a infinite sum of reflections).

We certainly want to improve all kinds of things to get more precise physics, including light and optics. The concept of a "model" is essential in physics, and there is no model that is even close to reproducing all phenomena at all scales and under all circumstances, but we will always work on improving our models!

It would be really helpful if you could create a scene that is supposed to reproduce a certain physical property of light and matter, submit the scene to algobox, and describe what goes wrong (bug) or what is missing from our physical models.

[Chronos]

viewtopic.php?f=16&t=97 this has been suggested

[Twinkle]

However, some phenomena, like when light passes through small particles (or the size of the wavelength of light) in the atmosphere (dust, pollen, smoke, water) are real hard to simulate since they require a really large number of internal reflections to give the correct result (this is referred to as "Mie scattering theory" and it relies on a infinite sum of reflections).

I would think an approximation would suffice. Kids can extrapolate rather well.

The way I see it, there are three basic effects that can be simulated: reflection, refraction, and absorption. Each of these effects can be dependent on the wavelength of the light involved. Right now, refraction is modeled exactly like that... hence the spectra from prisms, etc. Absorption as a function of wavelength, as Chronos mentioned, has been suggested elsewhere. Reflection is modeled, but not as a function of wavelength.

It would be really helpful if you could create a scene that is supposed to reproduce a certain physical property of light and matter, submit the scene to algobox, and describe what goes wrong (bug) or what is missing from our physical models.

Here is the best I can do so far:


Author: Twinkle
Title: Why The Sky is Blue
File Size: 98.45 kB
Rating:


The example has to use refraction instead of reflection, which is a sort of all-or-nothing way to redirect the light (I notice that your reflections can be partial... allowing some light to be transmitted). Because of this, I've had to use many laserpens of the same color. I originally made two different kinds of material and red, green, and blue sources, since the amount of reflection should be a function of wavelength. However, I think that since I'm using refraction in the example (which is already a function of wavelength), that wasn't really necessary.

The better implementation would be similar, but with a single laserpen of white light. As the light struck a particle, a portion of it, based on wavelength, would be reflected (some of the blue, a smaller amount of the green, none of the red). The rest would be transmitted. This would keep the transmitted wavelengths going in the same direction (more or less), becoming more and more yellow, then red. The reflected wavelengths would demonstrate why the sky is blue. If the light passed through a thin layer of particles, it would come out the other end yellow. If it passed through a thicker layer, it would come out orange or red.

[pixartist]

Author: pixartist
Title: Colorfilter for suggestions
File Size: 42.01 kB
Rating:

[KarateBrot]

Author: KarateBrot
Title: Gravitational redshift
File Size: 34.44 kB
Rating:


There's a physical effect to change the color of the laser depending on gravitation. It's not as cool as real filters but anyhow you can get the red or the blue color of any color you choose.