[Radiance-general] Color reflectance values and materials..

[Thomas Bleicher]

On 7 Nov 2008, at 09:17, steve michel wrote:

> I am 100% in agreement with idea of using  photospectrometer  
> validated readings of 'standard' color charts. Let me add that  
> perhaps paint manufacturers may already have this data ready or be  
> keen on the approach.

I have here a list of 50+ RAL and xKxxx (like 2K216) colors.
I don't remember where I got those from and if I'm allowed
to share them in public. I also do not know if the Radiance
RGB values are validated or 'just' calculated.

On the subject of material conversion in general I have
to confess that my own attempts to use the equations on
easyrgb.org failed and I really don't know how it should
be done. I was only concerned about the monitor RGB
to Radiance RGB bit (for Sketchup exports) but even here
I felt that a simple linear scale by 0.85 gave better results
than jumping from colour space to colour space. I finally
found a mail from Greg or Axel stating that these two RGB
spaces are so close that a big convoluted conversion is
often not necessary. And so I left it at that.

The library idea sounds appealing at first but i think the
practical value of validated colour charts is just as big as
a good conversion from the published RGB values. The
real world problem I see is that the colour changes it's
properties depending on the material it's painted on.
Take wall paint on plaster for example and compare this
with the same paint on a dense wood fibre board (HDF
or whatever you call it). The texture of the material will
change the properties of the paint.

So instead of a big honking official library a nice and
small peer-reviewed  HOWTO on the conversion from
manufacturer's RGB data (and colour space) to
Radiance values would have more practical value.
I assume that there is such a conversion procedure.
If we have to measure before we get any realistic
values then a library seems to be the thing to do.
Nice job for a student in the need of a thesis ...


Yours humbly,

Thomas