Shmunter
Member
Whichever way the tech is implemented, the tantalising aspect is the ability to bring in such detailed assets in real-time as needed. . The camera zooms up to the character and this is what we see. Zooms away, a reduced LOD is swapped in - or whatever equivalent future tech to LOD there is.Just because they can push lossless assets doesn't mean they shoul. Actually, they can reduce the triangle count of this sculpt and utilize textures and normal maps with a minimum to zero effect on quality.
Also AFAIK, the complex topology of these lossless sculpt may not allow for good animation.
It no longer matters if the model in view takes up 90% of the vram, because it can be flushed out and replaced with world models in an instant. Only the render view is what matters. Game sizes and compression ratios notwithstanding.
I've often used when describing to people the prospect of mass streaming, that gameplay should look as good or very close to cutscenes.
Last edited: