Comments for Physically-Based Rendering

Comment on Layered Material by D-POWER

D-POWER — Wed, 27 Jan 2010 16:23:12 +0000

Does this mean that you can simulate wet materials ?

Comment on Dielectrics bis by Romain

Romain — Mon, 14 Dec 2009 09:47:30 +0000

Hi Erwin, and thanks for the kind words.
The renderer is currently pretty basic: The renders I posted so far show off everything it can do. If in the future it matures enough to be usable then I would consider releasing it and add more rendering options… Obviously that’s pretty far off at the moment.
I’ll definitely add render times in my next posts as it seems to be a most requested info

Comment on Dielectrics bis by erwin

erwin — Sat, 12 Dec 2009 16:54:45 +0000

keep up the good work!, the images your render engine created looks very nice. how long are the render time for these ?. will there be any option to go biased?. would it be released?

Comment on Dielectrics bis by Romain

Romain — Tue, 08 Dec 2009 22:23:48 +0000

@D-POWER This is a standard Fresnel specular model rendered at 3000 spp using an also standard Monte Carlo path tracer. As you can see there are still a few fireflies, the fact that the box is Lambertian does not help. I have no idea about the time, though (I really should add that info in the renderer's stats). But it is certainly not excessive for such a simple scene.

Comment on Dielectrics bis by D-POWER

D-POWER — Tue, 08 Dec 2009 20:34:50 +0000

Hi ,
Which BTDF are you using to model dielectrics?I also would be interested to know the render time/number of samples to get to this level of quality because i have always troubles with the rendition of this kind of materials.
Keep up the good work!

Comment on Reflections by Romain

Romain — Fri, 27 Nov 2009 09:21:46 +0000

@D-POWER
Oren Nayar takes into account the microfacets distribution and effects like shadowing and interreflections. But as the microfacets are Lambertian, the model does converge to a Lambertian BRDF when the variance parameter tends to 0.

In the render above I used the Blinn distribution function, and a pretty high Ior. So obviously your observation has some truth in it.
(The right-most sphere is pure Fresnel reflection btw.)

Comment on Reflections by D-POWER

D-POWER — Fri, 27 Nov 2009 07:46:20 +0000

I think Oren Nayar is very similar to a Lambertian BRDF ,as a matter of fact I see no difference.What type of material can you model with this BRDF that you can’t with a Lambertian BRDF ?
As for other 4 spheres , it looks to me like the modified Phong BRDF width different roughness.
Anyway ,I like this render .Great Job!

Comment on Convex Quadrilaterals by Romain

Romain — Thu, 29 Oct 2009 21:20:18 +0000

@D-POWER
I recall there are around 1000 samples per pixels and that it took close 45 minutes to 1 hour.
The indirectly-iluminated areas are still slightly noisy, though. It’s difficult to see on the picture (but visible on the front,dark side of the cube) because I scaled down the original render from 700×700 to 300×300 pixels.

Comment on Convex Quadrilaterals by D-POWER

D-POWER — Wed, 28 Oct 2009 15:07:56 +0000

That’s neat !I was wondering what is the number of samples per pixel that you used to get to this level of quality and how did it take to render this image?
Thanks

Comment on ASCII Tracing by Romain

Romain — Fri, 23 Oct 2009 11:59:42 +0000

Hi Fused,

Yeah that’s a nice ASCII render you have there!