Hi Everyone
My name is Johan Hammes, and I basically joined the list because someone
told me that there is a shadow discussion going on here. More on that later.
I started my own company a year ago that develops real-time algorithms, and
one of my products does exactly what you are discussing currently, except
that by just adjusting gamma ramps, you can never simulate the range of
light that the human eye is capable of seeing.
On the SIGGRAPH article, my stuff s based on that, but they did a non
real-time implimentation, unless I missed an article.
They human eye can see 1.5 log units of light at one moment (log (cd/m^2)),
but can adapt (mainly neurologically) to over 7 log units of light. Of this
only one log unit is achieved through the widening of the iris, the rest is
neurological. To simulate this wide range, you have to simulate the eye in a
lot more detail.
One of the nice things of an algorihm like this is that it can also be used
to simulate other adapting sensors like video cameras.
Have a look on my website for the screenshots on how this work. I know you
probably dislike advertisements, but all of this is for sale to incorporate
in commercial engines. For those of you who are just enthusiasts, if you
preasure me enough, I will release it as open source for non commercial use.
Johan Hammes
Nineteenseventyfive
http://mzone.mweb.co.za/residents/jhammes/main.htm
|
