With regard to what Wing64 said, now i recall something... Isn't the MRT's added to the SVN already? some of the passes can be done in one step, while you render, perhaps, phong and normal, you also render the depthmap. But yes, 3 renders for reflection and refraction are too heavy
"There is nothing truly useless, it always serves as a bad example". Arthur A. Schmitt
wing64... could i try to run your app??? and i would also like to get the models from it and try to run it with my deferred shading class and post process to see how much my HDR, SSAO, motion blur, Depth of Field, a couple of dynamic lights and 6 split CSM will render with same thing (I use MRT and other tricks)... and i will chuck in water too if i can do it in time (my water doesnt draw 3 times but 1 once for the reflection and also performs depth rendering/volumetric fog and water caustics).
@devsh: Firstly thanks for interested our models. That's a great idea and i need to test my scene too. U can send your deferred shading class and post process routine to me and when i test your code with my models finished i will report to u. thanks again for interested our models.
devsh wrote:i think it would be quicker the other way... saving you time porting to windows and meddling through my deffered shading and sphagetti code.
Guess you won't get access to high quality arts this easily
If wing64 is a commercial developer and has no rights to distribute assets to anyone outside his company, then I guess getting authorisation, writing up a non-disclosure agreement and getting it signed would be quite a slow process.
my submission (doesnt have a chance of winning but i hope it will bring interest)
I'm still working on making it efficient...
ATM it needs 128 samples which is pretty high.. i might drop to 64 but i need to "fake" the sampling meaning produce an artificial sample between the current and the last.
... All of em....
