GPU+CPU support for Bi-directional mode

[Googog Mogog]

Hello guys!

When we gonna have GPU accelerated bi-directional Path tracing support in Indigo render.

Because up coming Lux Render version 0.9 has support for GPU accelerated Bi-directional path tracing.

[Zom-B]

Seems you mixed up here something, but I'm open for any proof: link?!

[CTZn]

There's nothing impossible, lycium stated that OnoSendai had it working on an early GPU implementation. For the least, as a proof of concept am I guessing.

Sorry, I'm too lazy to retrieve the link to that statement ;P

[Zom-B]

I know that it is possible, and also has been done (link), but afaik it isn't build in in any availible unbiased renderer atm.
I follow Lux development a little, and think I would stumbled on such a feature.
I didn't want to sound arrogant with my first commend, just interested

[dougal2]

It's true, LuxRender does have the beginnings of a hybrid CPU+GPU bidir integrator, however although it seems to work, it isn't finished and there are some known problems with it still.

[Zom-B]

Kudoz to the Lux team, missed that feature totally! thanks for the Info doug.

[lycium]

It's true, LuxRender does have the beginnings of a hybrid CPU+GPU bidir integrator, however although it seems to work, it isn't finished and there are some known problems with it still.

Well, we definitely haven't forgotten about GPU bi-dir, but judging from that screenshot even our CPU single-dir has nothing to worry about

[Googog Mogog]

Seems you mixed up here something, but I'm open for any proof: link?!

Hello,

As a prove here is the link for Lux Render 0.9 new features!

http://www.luxrender.net/wiki/New_in_0-9

I hope you will believe now "Nothing is impossible"

[lycium]

They also seem to think SPPM is unbiased

Recently I made a joke-post on Reddit about how people (incorrectly) seem to think that anything with noise/randomisation is unbiased, which is applicable here: http://www.reddit.com/r/computergraphic ... of_hdr_in/

I tried to make it like the HDR in photography vs in games comparison: http://img194.imageshack.us/img194/7391 ... 383293.jpg

[Googog Mogog]

Also i want invisible Light emitters features in Indigo render just like maxwell render

[Googog Mogog]

They also seem to think SPPM is unbiased

Recently I made a joke-post on Reddit about how people (incorrectly) seem to think that anything with noise/randomisation is unbiased, which is applicable here: http://www.reddit.com/r/computergraphic ... of_hdr_in/

I tried to make it like the HDR in photography vs in games comparison: http://img194.imageshack.us/img194/7391 ... 383293.jpg

Hey lycium nice discussion u have there.

But According to me no unbiased renderer is true light simulator. The real world light behaves very differently then any unbiased renderer. What happen in real world is that a light source emits light and when this light hits the surface of an object the surface which is visible to both direct and indirect light becomes light itself which means light reflect itself from the surface of an object which is based on the reflective properties of an object and light intensity which is coming from direct light source.

As an example take a torch and just for experiment go in dark room with some objects. Turn torch light direction to white diffuse wall of that room. When u do that u will see that light will reflect from wall which is of course indirect light and this indirect light cast nice shadows onto objects just like Light source.This is reflective light . And this is what real world do. Not to simulate any photons ,bounces what so ever in real world Light got no particles. In real world as light emits it will reflect same as its from the Direct light source but with different intensity and properties based on the surface of an object.

This is my observation.

[lycium]

Unbiased has a very specific technical definition, it's not a matter of opinion It's about the correctness (on average) of a statistical estimator, which is what Monte Carlo renderers fundamentally create.

Please see: http://www.cgafaq.info/wiki/Bias_in_rendering

As for "true light simulation", if we restrict ourselves to Classical Geometric Optics (no quantum phenomena, no interference etc.), then Indigo is pretty much correct.

[Headroom]

Googog,

With all due respect to your observations, please educate yourself before posting!

There are people that would somewhat disagree with what you observe as the real world behavior of light. Such are for example Richard Feynman who has received a nobel price for his work in the area of Quantum Electro Dynamics. You mentioned Maxwell earlier. Has it ever dawned on you why this software is called Maxwell or are you really not familiar with the work of James Clerk Maxwell ?

There are a good number of people that have worked for decades of their academic careers on lighting simulation algorithms an have compared these with real world measurements. Phong, Blinn, Ward .... sounds familiar ?

If you dive a little deeper into Indigo's abilities and read through the forums you will find that Inigo is indeed quite accurate in simulating the behavior of light.

[ENSLAVER]

inb4 "it only exists when you observe it"

[lycium]

in b4 allegory of the cave as counter-argument

edit: woohoo 666 posts