Dear Group
There have been previous post about how to obtain accurate (or as accurate as can be expected) material decriptions for image maps used within scene. The general approach seems to be to scan the material, calibrate it using a macbethcal.cal calibration file, normalize it with normpat and later apply a material with desired reflectances. My question is what would be a valid approach for materials not scanned and calibrated? Many of the image maps that I use I obtain from various sources (i.ow. the internet) such that I can not get the exact reflectance of the material used for my image maps. This would generally apply to the floor materials that I use. What I have been doing has been a similar approach that I have used with Lightscape. What I do is take the image apply it to a material and set the RGB reflectance to a range similar to the recommended reflectance range that is used in Lightscape. I question the validity of this approach. I know that I am in the minority in this group due to the fact that I wish to obtain a visually pleasing image while at the same providing accuracy when it regards the lighting distribution throughout the building. Any help would be appreciated.
Regards,
Marcus
Hi Marcus,
If you know roughly the reflectance of the surface you want to model, you can still use normpat (or more easily "pfilt -2 -e 2") to create an image that has its overall value set to 1. Then, simply apply the desired reflectance as a neutral value to your surface (e.g., 20% reflectance would be "0.2 0.2 0.2") and use the image map you have downloaded.
I hope this makes sense.
-Greg
···
From: "Marcus Jacobs" <[email protected]>
Date: October 29, 2004 10:59:09 AM PDT
Dear Group
There have been previous post about how to obtain accurate (or as accurate as can be expected) material decriptions for image maps used within scene. The general approach seems to be to scan the material, calibrate it using a macbethcal.cal calibration file, normalize it with normpat and later apply a material with desired reflectances. My question is what would be a valid approach for materials not scanned and calibrated? Many of the image maps that I use I obtain from various sources (i.ow. the internet) such that I can not get the exact reflectance of the material used for my image maps. This would generally apply to the floor materials that I use. What I have been doing has been a similar approach that I have used with Lightscape. What I do is take the image apply it to a material and set the RGB reflectance to a range similar to the recommended reflectance range that is used in Lightscape. I question the validity of this approach. I know that I am in the minority in this group due to the fact that I wish to obtain a visually pleasing image while at the same providing accuracy when it regards the lighting distribution throughout the building. Any help would be appreciated.
Regards,
Marcus
Hi,
I use the maps only to vary the brightness of the materials. So I use an
image, apply normpat to it (so the average pixel value is one) and use it as
colorpict on a well defined material.
So the question of defining accurate physical properties remains in the
materials' definitions. I use "calibrated" pictures for these, too, but of
course you could use any RELIABLE source of information for these. In fact,
there are better ways to measure reflectance than calibrated photo scans...
I have no idea about the reliability of the materials defined in
lightscape's libraries. Also you need to find out the definition of
Lightscape rgb (what does it mean as material properties?). That is a real
Lightscape question, the definition of the values that must be supplied in
radiance are well documented in "Rendering with Radiance" as well as in the
included documentation.
The other way to use image maps, I asked about some months ago on this list,
was to use colored image maps with defined materials. It seams to be to hard
too achieve valid results that way too me, so I didn't try any more. The
third is too have calibrated image maps on a neutral material, so that the
reflectance is ONLY defined by the picture values, but as I understood that
is not the way you want to / can go.
Good luck, CU, Lars.
···
Dear Group
There have been previous post about how to obtain accurate (or as accurate
as can be expected) material decriptions for image maps used within scene.
The general approach seems to be to scan the material, calibrate it using
a
macbethcal.cal calibration file, normalize it with normpat and later apply
a
material with desired reflectances. My question is what would be a valid
approach for materials not scanned and calibrated? Many of the image maps
that I use I obtain from various sources (i.ow. the internet) such that I
can not get the exact reflectance of the material used for my image maps.
This would generally apply to the floor materials that I use. What I have
been doing has been a similar approach that I have used with Lightscape.
What I do is take the image apply it to a material and set the RGB
reflectance to a range similar to the recommended reflectance range that
is
used in Lightscape. I question the validity of this approach. I know that
I
am in the minority in this group due to the fact that I wish to obtain a
visually pleasing image while at the same providing accuracy when it
regards
the lighting distribution throughout the building. Any help would be
appreciated.
Regards,
Marcus
_______________________________________________
Radiance-general mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-general