Ray tracing is a conceptually lazy and computationally expensive. Fire off as many rays as you can in every direction from every light source, when the ray hits something it gets lit up and fires off more rays of lower intensity and maybe a different colour.
Sure you can optimize things by having a maximum number of bounces or a maximum distance each ray can travel but all that does is decrease the quality of your lighting. An abstracted model can be optimized like crazy BUT it take a lot of man power (paid hours) and doesn’t directly translate to revenue for the publisher.
The only downside of raytracing is the performance cost.
The downside is the wallet cost. Spreading the development cost of making a better conventional lighting system over thousands of copies of a game is negligible, requiring ray tracing hardware is an extra 500-1000 bucks that could otherwise be spent on games.
Ray tracing is a conceptually lazy and computationally expensive. Fire off as many rays as you can in every direction from every light source, when the ray hits something it gets lit up and fires off more rays of lower intensity and maybe a different colour.
Sure you can optimize things by having a maximum number of bounces or a maximum distance each ray can travel but all that does is decrease the quality of your lighting. An abstracted model can be optimized like crazy BUT it take a lot of man power (paid hours) and doesn’t directly translate to revenue for the publisher.
The downside is the wallet cost. Spreading the development cost of making a better conventional lighting system over thousands of copies of a game is negligible, requiring ray tracing hardware is an extra 500-1000 bucks that could otherwise be spent on games.