There is a proverb (?) in France: << La nuit porte conseil >> (which means "let pass some time and things will go clearer by itself"). Another one states that if you are in trouble, better ask some friends about the subject. To get better help, I put them in practice together and have some discussion with a friend of mine last night. She's not a scientist but works at a contemporary art museum and she had several ideas on what makes a texture and what makes an image. So we summarize/formalize the question in this way:
1) Shapes, objects, forms (the whole world in fact), no matter of being bidimensional like for example a picture or tridimensional like for example a sculpture, are seen by the eye because they exhibit a spatial luminance distribution (which may be approximated by saying "a light and shadows distribution").
2) We will define an _scene_ as a lights and shadows distribution which leads to something "intellegible" (a face where lights and shadows helps to identify the nose, mouth, eyes etc., a landscape where we can identify grass, forests, lakes areas, etc.). In a simplified, but operative model, a scene can be decomposed in a set of several _different_ objects, each of them having its own texture.
3) We will define a _texture_ as a lights and shadows distribution which leads to an "unintellegible" image where we cannot really identify what is the object (a close-up of grass may, for example, by as well seen as hairs, a detail of a cloud as a piece of cotton ball, etc.). A texture may be decomposed as a [random] distribution of bright and dark _identical_ elements (this again is a simplified model).
4a) A scene may be "contrasty" or not (a black cadillac drived by Abby vs. a grey car in a misty day). It is the same for a texture (sand is a flat texture but salt and pepper is a contrasty texture).
4b) The contrast of a scene may be quantified by its luminance range, by looking for its darkest and brightest areas. The same for a texture, except that black/white areas are now "micro-areas". Practically, measuring the contrast range of a scene may be done with a spotmeter. Measuring the contrast range of a texture could be achieved in the same way. Just is it more difficult, because the spotmeter has to focus to these micro-areas (difficult because common spotmeters are not designed for this purpose, but it is just a question of technology).
5a) We may easily also define a "scene brighntess" which would be the luminance measured over the whole area. It is the luminance measured not by a spot meter, but by a wide-angle photometer. Brightness is in fact the mean value of the luminances of all the several different areas (so brightness will be low if there are many dark areas with just a few bright ones, etc.) Usually zone system users are not at all concerned by this "integral" measurement.
5b) Texture brightness may be defined in exactly the same way. Texture brightness measurment is in fact exactly what every zone system user does when it measures the brightness in the different areas of interest....
