PMC dispersion bolts test

[radiance]

Hi all,

Here's a little test render with the new pmc kernel in 2.47+,
using dispersion on the glass bolts. (scene is by dave62, it's in our new testsuite)
It's raw output from octane, no post-proc.
The cauchy B parameter for dispersion is too high though, 0.1, it's not something any realistic materials would have, I just wanted to see as much dispersion as possible.

Radiance

[DayVids]

Good lookin, how much is the render time affected (if at all) by this?

[funk]

That looks great man.

What sort of render times should we expect once you've optimised it? Will it be faster than path tracing?

[hadouken]

that looks really clean.

[radiance]

I left it running overnight, and in the morning it was finished at 16000/16000.
I don't know when it was done though, and forgot to check... (should have taken a screenshot too... )
It looked pretty clean after on hour though, i'm sure 2 hours would have been more than enough, and I used one GTX460 GPU.

We're still working on optimizing the PMC kernel, at this time it's difficult to predict, but I'm hoping we can get 2x speedup after the optimizations are finished,
in a few weeks with the big beta 2.5 release.
Bidirectional pathtracing (and PMC) should make renders like this one converge MUCH faster,
and bidirectional pathtracing + PMC is something we will be starting work on next, after 2.5 is out.
PMC + bidirectional will be ideal, it will be as efficient as the popular standard in CPU based unbiased engines (MLT+bidir),
and this combines with the power of GPUs should really take things to a new level.

Radiance

[Jaberwocky]

I left it running overnight, and in the morning it was finished at 16000/16000.
I don't know when it was done though, and forgot to check... (should have taken a screenshot too... )
It looked pretty clean after on hour though, i'm sure 2 hours would have been more than enough, and I used one GTX460 GPU.

We're still working on optimizing the PMC kernel, at this time it's difficult to predict, but I'm hoping we can get 2x speedup after the optimizations are finished,
in a few weeks with the big beta 2.5 release.
Bidirectional pathtracing (and PMC) should make renders like this one converge MUCH faster,
and bidirectional pathtracing + PMC is something we will be starting work on next, after 2.5 is out.
PMC + bidirectional will be ideal, it will be as efficient as the popular standard in CPU based unbiased engines (MLT+bidir),
and this combines with the power of GPUs should really take things to a new level.

Radiance

That sounds great.......More speed.......I love speed........

Careful radiance...don't feed the zombies.

[radiant]

Bidirectional pathtracing (and PMC) should make renders like this one converge MUCH faster,
Radiance

So is bidirectional path tracing like an add on or a modified version of path tracing?
What is the difference between the two?

[tchoa]

That is very promising!
Cheers
F

[radiance]

Bidirectional pathtracing (and PMC) should make renders like this one converge MUCH faster,
Radiance

So is bidirectional path tracing like an add on or a modified version of path tracing?
What is the difference between the two?

It's a more complex algorithm, it's also unbiased, produces the same result (if implemented properly),
but is more efficient for caustics and hidden lightsources, as it traces rays from lights and from the camera,
and interconnects them.

It's quite complex to implement, and requires a complete foundation, which we now have.
It can also be combined with PMC to make it even better.

Radiance

[tungee]

That sounds so good!