PBR

[chris1515]

That's exactly the thing - PBR is not something you could recognize, because theoretically one could get very realistic results even with all kinds of cheating going on; different shader variations for different environments and so on.

PBR's main advantage is in trading off math and complexity for artist hours. There's a much lower number of knobs in the shader with more limitations on the settings, but the one final setting should work in all possible circumstances and it could be re-used for all similar materials and so on. Once you get the artists trained, they should become a lot more productive and create generally higher quality content as well.

It is funny because people always think than more power mean more artist work and inflated cost. It can save time and cost like with PBR or without a rendering farm for baked GI calculation..

When the studio could use a good real time and complete GI approximation it will save time and cost. No crazy work on shadow maps, no SSAO or HBAO or other AO approximation or other local approximation like screen space reflexion . And I think use the high model polygon in Zbrush and don't work with normal maps can save artist time. And time is money...

And I read asset creation is the biggest cost in game production.

On the subject I think it will be very interesting to see PBR implementation by DICE in Frosbite. Battlefront?

 

[AlNom]

huh... apparently there was a PBR system of sorts for H2A (MP at least).

 

[Davros]

Crysis 2 used Oren Nayar and Cooke Torrance -- both physically plausible models but they never announced it as a buzzword.

That's because they left Crytek shortly after to pursue a career in world music....