Hi,
While the osglogo's outlined text looks much better on ATI with the v1.1 win32
binaries, last night I noticed that the flashing white edge pixels artifact now
can be made to happen on an NVIDIA card.
Eric, did you mention that you had a different overall approach? If so, perhaps
migrate to that?
Cheers
-- mew
Yes, I have introduced 4 different rendering backends: polygon offset,
no depth buffer, depth range, and stencil buffer.
I need you to test all of them. (You've already tested the first.) In
osglogo.cpp, there are the following 4 lines:
text->setBackdropImplementation(osgText::Text::POLYGON_OFFSET);
// text->setBackdropImplementation(osgText::Text::NO_DEPTH_BUFFER);
// text->setBackdropImplementation(osgText::Text::DEPTH_RANGE);
// text->setBackdropImplementation(osgText::Text::STENCIL_BUFFER);
Switch out the one you want to test and recompile osglogo. These are
heavily documented in the osgText::Text header file so you can
understand the trade-offs of each.
In my experiments, DEPTH_RANGE works surprisingly well and I haven't
had any of the issues we've seen with POLYGON_OFFSET. But it's
potentially vulnerable to the same kinds of issues so I need to know
how it works on other video cards/platforms. If this works well enough
for everybody, we should make this the new default.
NO_DEPTH_BUFFER should always work at the cost of losing depth writing
for the text.
STENCIL_BUFFER should be the option that always produces the correct
results and at high quality. But it needs a hardware accelerated
stencil buffer to run at a decent speed. I also am not sure I wrote
the code correctly because there are things in there that don't make
sense to me. (See previous emails.) But it seems to work which puzzles
me. I'm still hoping somebody can audit my code and validate it.
-Eric
_______________________________________________
osg-users mailing list
[email protected]
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/
