POV-Ray : Newsgroups : povray.programming : Ray collision export, spectral rendering, and dispersion at relective grati= : Re: Ray collision export, spectral rendering, and dispersion at relective grati= Server Time
25 Sep 2021 11:16:19 EDT (-0400)
  Re: Ray collision export, spectral rendering, and dispersion at relective grati=  
From: clipka
Date: 5 Apr 2018 12:10:48
Message: <5ac64a88$1@news.povray.org>
Am 05.04.2018 um 15:22 schrieb Alexander jahn:

> a) is it possible to export text of the individual ray collision and
> diffraction?


In POV-Ray, by design raytracing is only employed as a means to create
images, not as a goal in and of itself.

You /might/ be able to coax POV-Ray into writing /some/ rudimentary
ray-object collision "log" by using photons and writing the photon map
to a file, but that log would be far less detailed, IIRC have some
binary format, and would also require some tinkering with the scene to
set it up for this purpose.

Adding a dedicated "ray logging" mechanism to the code may be possible,
but probably non-trivial, especially since contemporary POV-Ray uses
symmetric multithreading so you can't just naively dump ray-object
intersection information to a file "on the fly".

> b) Is it possible to apply the attribute "wavelength" to an individual ray?
> In my setup, the wavelength of the light is very important, but most ray-tracers
> do not include that info afaik.
> The response to this question is important for question c)

For some mechanisms POV-Ray does indeed tag rays as monochromatic with
an associated wavelength, namely whenever dispersion comes into play.
(Note however that the tag and wavelength are only evaluated in
subsequent dispersive refraction events; colour computations are
/always/ done on an RGB basis.)

> c) Is it possible to simulate a reflective refraction grating?
> Example
> https://www.thorlabs.com/images/TabImages/Littrow_Configuration_D1-780.gif
> The main issue that I see is that it is not a particle effect, but rather an
> interference effect, using a ray-model of light is not the best way I guess.

At present, POV-Ray provides no way to model such a thing. I'm sure such
a feature could be added (given enough time and energy), but probably
only via a stochastic approach. (Someone would also have to come up with
a neat syntax for the feature, so that it fits nicely into POV-Ray's
current system of material properties.)

Post a reply to this message

Copyright 2003-2021 Persistence of Vision Raytracer Pty. Ltd.