[Radiance-general] An open materials library?

[Rob Guglielmetti]

On 2 Aug 2002, at 15:59, Peter Apian-Bennewitz wrote:

> two thoughts:
> How does a user know the real visual appearance of the "light
> yellowish wood" he finds in the library ? Plus variants like
> "polished" "slightly polished" etc. One may supply photographs of the
> real thing, but I'm not sure that's good enough. Secondly, measuring
> materials (spectral and/or angular) is, even with automated and fast
> machinery, some work, data processing and modeling are too, especially
> for an easily increasing number of materials, such as in a catalogue. 

Well, it's still a good start, for the newbies (like me) especially.  
Learning by trial and error is so time consuming; a "starter set" of 
materials that are maybe from a common pallete, that could be 
used as a starting point for experimentation could help the trial and 
error process move along a bit more quickly. 

> It'll be very valuable to get feedback from users how they would use
> any form of catalogue - education or project work, for filling in data
> when no materials are given in a project or for matching a given
> material in a project ?

Here's my perspective, coming from the Lighting Designer-who-
uses-Lightscape side of the fence: we tend to use "value-only" 
materials in Lightscape (ie, hue=0, saturation=0, value=.xx, where 
".xx" is the "reflectance" of the material).  The limited reflection & 
transmittance capability of Lightscape does not allow for really 
complex material definitions.  In addition, most other lighting 
design calculation programs used in the US (Lumen Micro, AGI, 
etc) don't support anything more than a reflectance value.  So long 
ago I adopted this simplified method of material definition in 
Lightscape.  Our models are greyscale representations of spaces.  
We like to think of them as "electronic foam core models". We shy 
away from hyper realism in the interest of time, and also in the 
interest of reducing the number of variables.  

But I've always wanted to learn Radiance, and the latest debacle 
with Lightscape going the way of the dinasour (and the advent of 
OS X) have pushed me into the radiance camp at long last.  So 
now I have this wonderfully capable calculation program, and 
suddenly have all these possibilities in terms of materials and 
reflection models.  It's exciting, but daunting.

So, a catalog of materials with known properties, perhaps tested 
with your gonio-photometer would allow me to create Radiance 
models that I could stand behind.  Even if it's not all-inclusive, at 
least if there were a selection of common materials with known 
performances, we could use them and then extrapolate as 
necessary, until a more fully-developed library is created. 
 
> I had the idea of compiling a reference sample catalogue, around 8-12
> material samples (e.g. 3x6cm) initially, plus their measurement data
> and fitted Radiance model. Would be quite handy for projects and
> feasible too, using my 12 year old, still used and fully automated
> gonio-photometer at FhG-ISE or a newer one, but I doubted there'll be
> enough interest. 

Well, I for one *am* interested!!  I actually have been meaning to 
ask you if your device is available to test submitted samples, 
Peter.  I'm involved in a 1/2-scale gallery mockup right now, costing 
the museum client hundreds of thousands of dollars, because 
Lightscape was in no way up to the challenge of testing the 
glazings.  Radiance probably could have, but not without good 
BRDF data.  Only way to obtain that is with your gonio-photometer. 
 Would this be an option in the future, perhaps for a fee?  We send 
you samples, you send us data files?  

:::::::::::::::::::::::::::::::::
Rob Guglielmetti
http://home.earthlink.net/~rpg777