I wonder if it is possible to use the transmittance/absorbance spectra of a
liquid in povray? I've got some data for some liquids that I would like
rendered.

Also ... is there a way to use standard illuminants, such as d65, C etc?

Thanks for your help

Post a reply to this message

> I wonder if it is possible to use the transmittance/absorbance spectra of a
> liquid in povray? I've got some data for some liquids that I would like
> rendered.
> 
> Also ... is there a way to use standard illuminants, such as d65, C etc?
> 

Not internally managed by POV-Ray, but perhaps these CIE includes can be 
useful for what you want:

http://www.ignorancia.org/en/index.php?page=Lightsys

Regards,

--
Jaime

Post a reply to this message

"epidot" <nomail@nomail> wrote:
> I wonder if it is possible to use the transmittance/absorbance spectra of a
> liquid in povray? I've got some data for some liquids that I would like
> rendered.
>
> Also ... is there a way to use standard illuminants, such as d65, C etc?
>
> Thanks for your help

Uh... I have no idea about "standard illuminants", but I guess what you're
trying to do would need a render with more spectral "bands" than just "Red",
"Green" and "Blue".

An idea how to achieve this would be a multi-pass render with a "false-color"
approach. Who said that the three components in a color vector are "Red",
"Green" and "Blue" anyway? Could be any spectral "band" you choose, e.g. three
particular "Fraunhofer lines" or whatever you like.

So wherever you'd normally use a color, you could instead use a macro with, say,
12 or 24 parameters representing spectral components, which picks a subset of
three depending on the render pass.

This way you'll end up with a bunch of false-color renders of your scene, which
you can then post-process in gimp, photoshop, or a properly designed
orthographic scene, by splitting up the three channels of each image into b/w
separate layers, multiplying each with the corresponding spectral band's
contributions to percieved RGB, and mixing finally mixing all the resulting
layers.


Alternatively, you might simplify the post-processing step at the cost of quite
some rendering time, by designing the macro so that it picks just a single
spectral band per pass, and already computes an RGB value from it. (Given that
R and B don't overlap too much, you might also be able to "cheat" a bit and
process two bands per pass, one in the "R/G" domain and one in the "G/B"
domain.)

Post a reply to this message

Jaime Vives Piqueres <jai### [at] ignoranciaorg> wrote:
> Not internally managed by POV-Ray, but perhaps these CIE includes can be
> useful for what you want:
>
> http://www.ignorancia.org/en/index.php?page=Lightsys

Duh! These are all single-pass renders?

I guess I want those in POV-Ray "natively"...

Post a reply to this message

>> http://www.ignorancia.org/en/index.php?page=Lightsys
> 
> Duh! These are all single-pass renders?

   You mean the demo images on the Lightsys page? Then, yes, these are all
single pass radiosity, and in fact with fairly dull rad settings.

> I guess I want those in POV-Ray "natively"...

  Yes, having the raytracer to handle internally spectral data instead rgb
will lead to more "physically correct" lighting and texturing, but I think
it's a bit too much work. And CIE.inc can be used for the same purposes, if
you know what you're doing... ;)

  Regards,

--
Jaime

Post a reply to this message

"epidot" <nomail@nomail> wrote:

> Also ... is there a way to use standard illuminants, such as d65...
>

I had to look that one up...

http://en.wikipedia.org/wiki/D65

Fascinating. So would the idea be to use (or compare) D65's 'standard'
color_temperature of 6504-deg. Kelvin to POV's rgb <1,1,1> in a light source?
I've been wondering about such a thing, too.

Ken W.

Post a reply to this message

Kenneth wrote:

> Fascinating. So would the idea be to use (or compare) D65's
> 'standard' color_temperature of 6504-deg. Kelvin to POV's rgb <1,1,1>
> in a light source? I've been wondering about such a thing, too.

They are very different things. D65 describes a spectrum, which
is a physical property of the light. The rgb value is a color in
a 3D color space, which is in effect an artefact of human evolution
(color vision with receptors for three wavelengths common on Earth).

The details usually just confuse me. Using the above mentioned CIE
includes, you should basically be able to generate an RGB value which
is perceived by some defined observer to be the same color as a D65
light source, after you did some worrying about white balance.

Post a reply to this message

Christian Froeschlin  wrote in message <497f5baa@news.povray.org>:
> (color vision with receptors for three wavelengths common on Earth).

A small correction here: the three sorts of cones are not sensible to just
one wavelength each, they are sensible to a whole range of wavelengths. Each

wide, and there is a lot of overlapping. Otherwise, we would not see
rainbows as a color gradient but as three thin colored stripes.

The reg, green and blue of sRGB are not artifacts of the human eye, they are
technological choices. To cover a reasonable range of the set of visible
colors the cheapest way possible, three primary colors, one around the reds,
one around the greens and one around the blues or the violets, are
necessary, but the actual precise value of these three colors is not a
physical or physiological property, it is just an arbitrary choice by the
industry.

Post a reply to this message

Nicolas George wrote:

> The reg, green and blue of sRGB are not artifacts of the human eye

yes, the specific base colors are not, but I think the
3-dimensionality of color spaces in general is.

Post a reply to this message