[deadalvs]

hi there.

" what is pure white in digital lighting ? "

* * *

i wanted to ask a few thing about digital image files. (that might interest some other people too)

* * *

at latest since .hdr lighting and the displacement feature/fea(r)ture, hehe, exist, everyone knows now that images and the technical stuff behind is not just that simple to understand.
now.
what i still don't understand is the relationship between bit-depth, the represented floating point scale and the brightness.
for example a floating point TIF file. --> can the color/brightness info exceed "1" in such a file or is it stored only between exactly [0,1], with the subdivision of either 2^8, 2^16 or 2^32 steps ?
an .hdr is said to store the "real" sunlight intensity, but how would that relate to a "100%" light in any 3d software? are there general agreements about that or how are the relations ?


so for an other example: if i use an .hdr map of a sunny sky (a photograph directly against the sun in front of a very deep coal mine gate) and use it as a displacement map.
let's say the brightness values go from 150'000 cd/m2 (sun) to 20 cd/m2 (cave). which values are stored in the .hdr?
--> if the values exceed 1, will the map be scaled down to the interval of just [0,1], scaling the whole logarithmic light intensity?

how does the lin/log value storing style change these values ?

*puzzled*


i guess this will clear a few things up.