You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I did some tests and I think that after merging #1788 now edges are rendered in a much nicer way when log. depth buffer is enabled than when it is not!
@xeolabs, Would you be interested in introducing the gradient technique also when log. depth buffer is not enabled?
The main changes I see that would be needed and could impact performance is to propagate the Z value from the vertex shader into the fragment shader, and the calculation of dxf and dxdy from the fragment shader.
Other than this the change is conceptually very easy 🙂
The text was updated successfully, but these errors were encountered:
Hi Toni!
Yes I think it would be good to use the gradient technique in non-log depth mode. The performance hit is OK, because users often use FastNavPlugin to hide edges while moving the viewpoint.
@tmarti This would be a very nice improvment too and could be interesting regarding performance compared to the log. depth buffer. Do you already have a draft implementation of this which you can share?
I did some tests and I think that after merging #1788 now edges are rendered in a much nicer way when log. depth buffer is enabled than when it is not!
@xeolabs, Would you be interested in introducing the gradient technique also when log. depth buffer is not enabled?
The main changes I see that would be needed and could impact performance is to propagate the Z value from the vertex shader into the fragment shader, and the calculation of
dxf
anddxdy
from the fragment shader.Other than this the change is conceptually very easy 🙂
The text was updated successfully, but these errors were encountered: