CPR Weblogs
© Copyright 2004 Lewis Downey  .
|
 |
Wednesday, August 27, 2003 |
| |
5:59:15 PM 
NY Times Dust and Deception...under White House direction, the E.P.A. suppressed warnings about indoor pollution. Scattered evidence suggests that as a result, hundreds of cleaning workers and thousands of residents may be suffering chronic health problems...
|
| |
5:57:44 PM
I am enjoying an excellent discussion with Marc Torzynski about defining a definition for textures. Marc writes:You hit exactly the point! Let me first fully agree with your interpretation of texture as a result of the interaction between light and a three dimensional surface. As you point out, the appearance of a surface will depend on the light properties (very directionnal light from one side -> highly textured surface, diffuse light coming from above the surface -> nearly no texture at all).
Your interpretation of the "strongness" of a texture in terms of lighting ratio between bright areas and shadow areas is also very useful.
Back to the beginning: when I read your Zone system tutorial for the first time, I mean, the paragraph saying something like "zone 0 is full black... zone 2 is the first suggestion of texture, zone 3-7 are fully textured, etc.", all this was perfectly clear and this description seems to me a very rigorous one. But when I really start to thing about practical cases, I fell in trouble and I quickly understand that the definition of (for example) zone 1 is partly arbitrary: for example, if a highly textured object (a concrete wall illuminated from one side by the sun) just suggests its texture when placed at zone VIII, the same object, once poorly textured by a more diffuse light will show no texture at all when placed, once again, at zone VIII.
I hope this example shows clearly why I am trying to define a "standard texture". According to your description, we have to define this "standard texture" by specifying the ratio between shadowed and illuminated areas. But, as you mention it, the strongness of the shadows areas depends on the multiple reflections between the bumps. It is nearly impossible to do any computer simulation because there are too much parameters. So I have absolutely no idea about the "stop range" of a textured object. Maybe should I modify a spot meter to make close-up measurments on several textured objects (wall, stones, etc.). Or use your idea about the "scalability" of the textures...
Marc is responding to a quick note with a drawing
I will post more if the conversation seems to go anywhere.
|
| |
5:53:29 PM
I do not know that we must define textures by the
luminance range, but it might be a useful element in a definition.
It seems to me there ought to be some way to model an extremely simplified texture. Maybe Photoshop or a 3D rendering program would be helpful in figuring this out.
To get a handle on the basic properties, try reading a surface with shadows on it
-- both the shadowed and illuminated parts. This subject could be a driveway
and your shadow. Watch out for those specular reflections you alluded too; asphalt
is full of them -- concrete much less so. I got slammed by the specular reflections in asphalt while calculating reciprocity corrections for a nighttime exposure.
It definitely seems better to work at human scale to me; although I do not know how to model the thing. Maybe we could start with a bunch of little pyramids evenly distributed over a surface -- either touching or close to touching. Then consider the light -- first using directional light later diffused light. Even if the first model is grossly simplified, there must be some way of doing it, or do you think that the effort can never achieve practical results.
|
|
