Post Your Irrlicht Screenshots / Render Here.

[hendu]

Such rhythm, much beat.

[Mel]

Nice! (i am an anime freak, i can't help it XD)



I know, it is not very impressive, but it is uses deferred shading, four channels in the GBuffer: diffuse, properties, normals+selfilum and depth.

[Granyte]

so long without a proper update







and some more there

http://granyte.blogspot.ca/

[REDDemon]

Cool

[Mel]

Parallax mapping on a deferred shading pipeline

[chronologicaldot]

Nice job, Mel!
Now if only the pipes didn't make it look awkward. X) Do you think you could do proper shading on the parallax-mapped surface? (I mean the pipes' shadow.)

[Mel]

Oh, there are no shadows there yet. That's just a spotlight.

[devsh]

our testers wanted 120 FoV, so I worked up this

Its a PoC, wont be implemented until all GPU pipeline upgrades are done, if implemented at all
The reason is that it messes up deferred rendering quite a lot, and the only way to counteract is to store extra data (At least 32bit-64bits more ppx in the GBuffer)



it effectively makes large FoVs playable (up to 110 before the projection starts being circular)
If I ever do expand on this, then the fisheye equisolid lens will be implemented for FoV>110


compare with this



and more pics

[Mel]

Nice job, Mel!
Now if only the pipes didn't make it look awkward. X) Do you think you could do proper shading on the parallax-mapped surface? (I mean the pipes' shadow.)

You mean something like this?

[devsh]

dont forget to modify gl_FragDepth for your surface to actually do depth testing properly (and not break the illusion where the pipe meets the parallax surface)

[Mel]

That is what i do (well, in am using HLSL, not that i couldn't use GLSL instead, but i like the good old DX9C ^^) that way the whole process remains the same. If i modified also the shadowmap generation, the parallax mapping could cast shadows on itself properly for instance, and besides, aditional effects making use of the depth buffer would take into account the depth alteration, and react to it properly (namely screen space ambient occlusion, depth of field blur and more...)

[wing64]

Character test rendering on ARSA Framework.

[Granyte]

Some interesting texture blending i'm trying on the dx11 driver with texture arrays

[chronologicaldot]

Nice job, Mel!
Now if only the pipes didn't make it look awkward. X) Do you think you could do proper shading on the parallax-mapped surface? (I mean the pipes' shadow.)

You mean something like this?

[ent1ty]

our testers wanted 120 FoV, so I worked up this

Its a PoC, wont be implemented until all GPU pipeline upgrades are done, if implemented at all
The reason is that it messes up deferred rendering quite a lot, and the only way to counteract is to store extra data (At least 32bit-64bits more ppx in the GBuffer)

I have asked in #irrlicht and I have asked in ##OpenGL, but nobody knows - how exactly does higher field of vision mess up deferred rendering? Is this something to do with your specific implementation?