I'm generally happy with the current method. It certainly is quite fast, as far
as diffuse interreflection calculation goes. The amount of parameters is not
that horrific, esp. concidering that only 2-3 need to be changed regularly. And
adjusting the relative scale of radiosity effects is a necessity, so I don't
know how you could get rid of scale dependence.
My only gripe is that - since values are not calculated on a per-pixel basis -
normal patterns cannot very well be taken into account.

Margus

TonyB wrote:
> 
> > The current radiosity implementation is quite good if you know how to use it.
> 
> Yes, I'm sure it is, but most people don't seem to understand how to use it. I
> #include "myself". The new one should improve radiosity outside of a little box
> room, should find a way to simplify the excessive amount of parameters that one
> is required to understand, and should be a less scale-dependant. If they can
> speed it up too, then that would be a welcome plus. Please help us, Mr. Kopp...
> take your time and finish the other things you must do, but please do this
> someday.
> 
> --
> Anthony L. Bennett
> http://welcome.to/TonyB
> 
> Graphics rendered
> by the Dreamachine.

Post a reply to this message