LuxRender Competition: Splash Screen (Closed)

[A-man]

a fast question: does SPPM support volumetric scattering (homogeneous volumes)?
In general, which are the limitations of features of SPPM compared to BiDirectional?

No volumetric scattering. Glossy materials are also problematic right now.

[SATtva]

Glossy materials are also problematic right now.

Convert them to matte+glossycoating to fix that.

[zeealpal]

Are we allowed to tonemap the output for the competition in another program? If we save as a HDR image, and convert to LDR?

[A-man]

Are we allowed to tonemap the output for the competition in another program? If we save as a HDR image, and convert to LDR?

That's probably pushing the boundaries. I know once MS Paint was allowed to overlay a border-rendered region, but other than that, using another program is taboo I think BinaryCortex will be able to make the final call though.

[B.Y.O.B.]

5. The only post production allowed is what is built in to LuxRender. (We are trying to showcase LuxRender, not Gimp or Photoshop.)

AFAIK the tonemapping is part of the post production... or could be considered in the case of physically based renderengines, where you don't get a final image without tonemapping.

[J the Ninja]

5. The only post production allowed is what is built in to LuxRender. (We are trying to showcase LuxRender, not Gimp or Photoshop.)

AFAIK the tonemapping is part of the post production... or could be considered in the case of physically based renderengines, where you don't get a final image without tonemapping.

BinaryCortex has forbade external tone mappers in the past, I've asked. It's all kind of a slippery slope of "where do you draw the line on post pro?". The true of it is, generally 3D artwork does not spring fully formed from the render engine. Sometimes you can nail the lighting JUST RIGHT that it does, or the image might be ok with straight render output. But there is always that extra touch from fancier tonemapping algorithms, blending multiple tone maps, levels and hue/saturation tweaks, screened-on bloom lighting layers (bloom in lux does not have a mask). Pretty much all the video footage you see done professionally has been color graded, and photos are always tweaked and airbrushed, why should you expect renderings to be better? "You can control the light and colors exactly" isn't a valid reason when you get down to it. Trying to get those things just right in the initial render is usually a LOT of fine-tuning and trial and error, whereas a lift/gamma/gain control will get it perfect, with real time feedback and no re-rendering.

[zeealpal]

Fair enough Turns out I made a new one straight from Lux that looks better now

[binarycortex]

Sorry for the restriction on post pro, but this is a LuxRender competition not a Gimp or Photoshop competition. My best work today wabi-sabi looks much better after even some mild post in The Gimp. But, then I could airbrush out fireflies, or photoshop in something from a real photograph and cheat. With that being said...I would like to make a challenge to all the developers out there. Lets see if we can add some of the open source tonemappers that exist out there. Or how about a simple levels control. I find the levels dialog in Gimp to be quite powerful. That alone could drastically improve some renders. Because, much like a real camera, LuxRender is subject to creating color casts, or having a less than ideal contrast every now and then. My wabi-sabi render had both.

There you have it. The gauntlet has been thrown. We need a levels adjustment in LuxRender. But make sure its applied post render, like the film response. Which in some cases provides a similar result.

[B.Y.O.B.]

I have a question: is the position of the LuxRender logo in the top right corner fixed or does it get adjusted to fit to the splash image best?

[jeanphi]

Hi,

It is adjusted to match the picture (look at the splash.xcf file from the repository).

Jeanphi