William F Pokorny <ano### [at] anonymousorg> wrote:
> On 2/11/21 2:12 PM, Kenneth wrote:

> > I don't think the failed AA is simply the result of a difference
> > in contrast/brightness with surrounding darker pixels.

> At the time AA is done it knows about the color channel values and nothing
> more. With color values >>1 there simply is a massive difference between
> adjacent pixels. If you have say emission 10 even with some sort of
> bloom / blend, you'll need consume 10 or more pixels around the ultra
> bright regions to get into a range were you'd get AA on the '<=1.0' edge.
>

Yes, I understand that now; and I was wrong to say that the surrounding darker
pixels have no effect on the 'lack of AA' of super-bright ones. They obviously
do, from visual tests. Making the surrounding pixels brighter helps to improve
the failed AA.

And the idea of simply 'clamping' very-bright pixels to <1,1,1> is rather naive,
I admit. There are probably many lighting/color situations where only one of the
color-vector values ends up >> 1.0-- in which case, simply clamping the highest
vector value to 1.0 may shift the intended (expected?) color, unless some other
'equalizing' scheme is brought into play to shift the other two values in a
subtle way. [This is just my 'thought experiment', based on nothing but
intuition ;-) ]

Post a reply to this message