What would be the correct way of converting color value from float to byte? At first I thought b=f*255.0 should do it, but now I'm thinking, that in this case only the exact 1.0 will be converted to 255, but 0.9999 will already be 254 which is probably not what I want...
It seems that b=f*256.0 would be better except that it would have an unwanted case of making 256 in the case of exact 1.0.
In the end I'm using this:
#define F2B(f) ((f) >= 1.0 ? 255 : (int)((f)*256.0))
I believe the correct is floor(f*256), not round. This will map the interval 0..1 to exactly 256 zones of equal length.
[EDIT] and check 256 as a special case.
1.0 is the only case that can go wrong, so handle that case separately:
b = floor(f == 1.0 ? 255 : f * 256.0)
Also, it might be worth forcing that f really is 0<=f<=1 to avoid incorrect behaviour due to rounding errors (eg. f=1.0000001).
f2 = max(0.0, min(1.0, f))
b = floor(f2 == 1.0 ? 255 : f2 * 256.0)
As the simple:
int value = (int)(f*255);
isn't giving you the result you require, any other linear conversion probably isn't going to cut it either.
What you'll need is a non-linear function something that gives you more values at the extremes and compresses the middle range. There are probably spline functions you could use - I'll need to do more research.
Notes on the simple conversion
This will convert 1.0 to 255 (barring rounding error) and 0.9999 to 254.
If rounding error on the 1.0 case is an issue then handle it explicitly.
The only issue might be the resolution. 1.0 - (255.0 / 254.0) returns 0.00392. Is this good enough?
Why not try something like
b=f*255.999
Gets rid of the special case f==1 but 0.999 is still 255