On Wed, Apr 4, 2012 at 5:40 PM, Frederic Bouvier wrote:
> The first thing to do it to convert the tree shader in a way it
> outputs correctly normals and colors to the 3 buffers. As long
> as you can see identically shaded objects in all three buffers,
> it's not good.
>
> You may want to look into deferred-gbuffer[.vert/.frag] shaders
> to see how to use gl_FragData instead of gl_FragColor. It's
> important to leave out all kind of shading. If you want the
> trees look cylindrical (put a better word here) and not flat,
> you may bake a normal that is not strictly toward the viewer,
> as it is really a flat billboard, but you can leave that for
> later.

I've managed to migrate the tree shaders to Rembrandt. Pretty
straightforward - thanks for the good instructions and examples.

I've checked in an update to git, but this morning realized that
I'd made a mistake in the effect file (tree.eff) in allowing both
non-Rembrandt and Rembrandt versions to co-exist. Specifically
I've created two techniques with the same n= value, so the Rembrandt
version is being over-written by the non-Rembrandt version.

I won't get the chance to fix this until after the weekend, but in the
meantime at least the Rembrandt performance problems with
random vegetation should be resolved.

-Stuart

------------------------------------------------------------------------------
Better than sec? Nothing is better than sec when it comes to
monitoring Big Data applications. Try Boundary one-second 
resolution app monitoring today. Free.
http://p.sf.net/sfu/Boundary-dev2dev
_______________________________________________
Flightgear-devel mailing list
Flightgear-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/flightgear-devel

Reply via email to