Project

General

Profile

Bug #62

Halo clipping vs. masked walls

Added by skyjake about 19 years ago. Updated almost 10 years ago.

Status:
Closed
Priority:
Lowest
Assignee:
Category:
-
Target version:
Start date:
2003-06-09
% Done:

100%


Description

For example in E1M3, around the nukage pool.

Halos look really bad when they're visible between
the ceiling and the top of the masked walls.

Labels: Graphics

History

#1 Updated by yagisan about 15 years ago

Logged In: YES
user_id=1248824
Originator: NO

bump - whats happening here ?

#2 Updated by danij about 15 years ago

Logged In: YES
user_id=849456
Originator: NO

This is an interesting problem. If we really want to handle this properly we will need a LOS test that can cast a ray through a perspective projected texture. Test to see if the covering texels of the lumobj origin are clipped by an opaque pixel in the masked wall.

I wonder if it might be better to rethink our approach with this. If instead we encoded lumobj origins into the accumulation buffer we could then cross reference with the z buffer and add halos as a post processing step.

#3 Updated by skyjake almost 13 years ago

Wait for 2.0 renderer.

Also available in: Atom PDF