[Radiance-general] RE: Lamp colour

[Gregory J. Ward]

Hi Anthony,

> From: "Anthony J. Farrell" <anthony.farrell at dit.ie>
> Date: May 24, 2005 9:19:34 AM PDT
>
> Thanks for the very helpful advice. Unfortunately I cannot access  
> the link
> you gave me as it says 'object not found', what's the name of the  
> paper?

Sorry about that -- I miscopied the link somehow.  Jack got the right  
one in his response.

> Apologies if my email was confusing, to simplify, I wish to  
> illustrate to an
> audience the dynamics of artificial light in providing a warm and cool
> environment relative to mixing light from 2000 to 6000K as is  
> presented by
> Philips
> (http://www.lighting.philips.com/gl_en/index.php? 
> main=global&parent=global&i
> d=global&lang=en under' a new concept in office lighting ' and then to
> compare that with actual warm (morning or evening) and cool  
> (midday) light
> from a window.

Oh dear.  Mixing illuminants is even more complicated, and not  
something Radiance does very well.  For this to be done correctly,  
you really need a full-blown spectral lighting simulation.  The  
problem is, there aren't any.  You can trick Radiance into doing it  
by running multiple scenes and choosing different spectral sample  
wavelengths for RGB each time then combining the results, but it's a  
hassle and difficult to do right.

> The lamp colour defaults are 0.907 0.863 0.4762

Is this RGB or Yxy (or ???)

> I now understand that the ies2rad program translates the colour  
> from the -t
> and -c options into the .dat file which is used in generating the
> distributions. The .fmt file must be one just used by DRAD, so not  
> to worry
> bout that.

Once again, what you'll get out of this technique is a simple color  
cast in your renderings, which will have almost nothing to do with  
how people might perceive the actual space, and doesn't tell you a  
thing either about how colors are rendered.

> Now when I consider the human eye things get complicated. There is no
> spectral power distribution of the fluorescent illuminant and the  
> spectral
> reflectances of your objects are also not available. It is a very  
> simple
> office scene with angled desks, yellow chair, and light coloured  
> partition
> dividers. Can you suggest where I can get generic data of this  
> kind? Or can
> you suggest the best way to approach what I am trying to simulate?

If you don't have spectral data, then this is an exercise in  
futility.  You can't learn anything by modifying the source colors  
independent of the surfaces.  Your results won't have any value or  
meaning.

-Greg