Multidimensional Adaptive Sampling and Reconstruction ?

[strangeday]

Thank you very much Dade, very explanatory!, and as SATtva said, thank you for your effort and time involved to make LUX better!

[A-man]

The result isn't very exciting. Again, it isn't the complete filtering process described in the paper but it has very very high disk, memory and cpu requirements and I'm not obtaining very interesting results. At moment, I hesitate to spend more time on this work because I'm not sure if it will take me somewhere.

If you quickly toggle between the input and output images, the progress is actually pretty substantial. It would be interesting to see a test with more than 8 s/p, like maybe 250 or so.

[Dade]

The result isn't very exciting. Again, it isn't the complete filtering process described in the paper but it has very very high disk, memory and cpu requirements and I'm not obtaining very interesting results. At moment, I hesitate to spend more time on this work because I'm not sure if it will take me somewhere.

If you quickly toggle between the input and output images, the progress is actually pretty substantial. It would be interesting to see a test with more than 8 s/p, like maybe 250 or so.

A-man, the problem is that a 256 s/p rendering generate a 23,648MB file (for a 600x600 image) And the processing time is proportional to the file size too (i.e. very long).

I'm quite convinced that, from a "philosophical" point of view, LuxRender/PBRT have too much focus on Samplers and not enough on Filtering/Post-processing at the moment. There have been lately a good number of papers about image reconstruction of MC renderings. I think we need something in this area.

The clustering of samples according position, normal, BSDF, ecc. for filtering is a simple and good idea but we need a manageable approach (in the framework of LuxRender).

[B.Y.O.B.]

I wonder what the authors of the paper did different... AFAIK they also used a modified LuxRender?

[A-man]

Has anyone actually contacted them and asked for their code? Or did we just assume they didn't want to share?

[weltbesiedler]

I read a few pages before that Dade wrote a letter to them, right?