47266adb$1@news.povray.org...
> Kewl. Faked, but kewl. ;-)  When changing the IOR in an unbiased render 
> automatically makes a spectrum, let me know. ;-)

From what I gather from the Maxwell manual, materials use "full ior" data 
(ior for each wavelength) that are taken into account for dispersion. They 
even talk about a "performance hit" when using these, which must be 
something...

Frankly, the Maxwell demo is free (crippled at 800x600 with a watermark, but 
one can create, save and render scenes), available for Windows 32 or 64 
bits, OSX and Linux, and comes with a few test scenes. Lots of user-made 
materials are also available for free apparently. It's really worth the 
couple of minutes it takes to download it, if only to see how it works, and 
a little test drive would probably answer some of the theoretical questions 
I see in these threads.

G.

Post a reply to this message

>> You're still thinking in terms of a traditional ray-tracer with separate
>> diffuse and reflection components.
>
>  So does your algorithm you give below.

But it's flexible enough to allow you to insert any BRDF, just fire more % 
of rays in a direction towards the light than you do randomly scattered, 
following the shape of the BRDF.  In a traditional ray-tracer this is 
impossible, in the code below it is trivial.

>  That algorithm makes it impossible to render images like this:
> http://warp.povusers.org/images/sphere.png

Just change "surface_color" to some other variable, or even probably you'd 
want to use some function that gives a physically correct change from one 
colour to another depending on the angle relative to the light position.

Post a reply to this message