is there somewhere explaining the difference between them?
I started with visibility based with the new update. but i ended getting a few cuda out of memory crashes (on 4gb). and seems to take longer and generally crash more. so i’ve stopped using it.
photoconsistency seems to be quicker and i haven’t had the same issues with it crashing.
but would be good to know difference. and if i can work in a different way that doesn’t crash visibility based so often, then that would be good too.
chris
Hi Chris
I started with visibility based with the new update. but i ended getting a few cuda out of memory crashes (on 4gb). and seems to take longer and generally crash more. so i’ve stopped using it.
= can say what precise CUDA error you get ? so we can rule out issues…
Visibility based is faster and not so compute resources hungry
Photoconsistency based do much more computations to for overall texture quality. mostly helpful when have quite different imgs quality ( brightness fluctuations and etc )
will prepare some examples so can see it in action…
but would be good to know difference. and if i can work in a different way that doesn’t crash visibility based so often, then that would be good too.
never get crash on this step
go to ucp.php?i=profile&mode=signature
and ad PC config description so can see what GPU are you using, hope have newest WHQL drivers installed