OctaneRender™ Standalone v1.0 beta3.02 [obsolete]

[abstrax]

New specular material BRDF they really should not miss the sun?

No, the sun still doesn't get through specular materials.

[Rickky]

Can you show an example with index for the fresnel effect, for the moment I don't catch it ;(

[Elvissuperstar007]

Can you show an example with index for the fresnel effect, for the moment I don't catch it ;(

+1
I also do not understand!

[abstrax]

Can you show an example with index for the fresnel effect, for the moment I don't catch it ;(

We can do that tomorrow, but you can easily try it out yourself:
- create a glossy material and set its diffuse pin to black
- set the specular pin to 1
- render the material ball with the glossy material
- now play with the index slider (at the bottom)
- as long as the index is >1 there is a Fresnel effect, i.e. the ball reflects stronger at shallow viewing angles (at the edge of the ball)
- if you set the index <= 1 (e.g. 0) there is no Fresnel effect, i.e. the reflection strength is independent of the viewing angle, more like a mirror

The cases without Fresnel represent materials like metals, the cases with Fresnel represent materials with a transparent coating, like furniture. If you don't use Fresnel, then you mainly control the reflection strength via the specular pin. If you use Fresnel, you control the reflection strength mostly via the index (of refraction).

Cheers,
Marcus

[PeterCGS]

Hum.. When using a old scene, my viewport is completely empty. I tried numerous scenes but everything is blank. I see my mesh in the project and materials etc. But the viewport is just blank. If I load in the .obj it works as normal.

Seems to happen only to scenes that use a environment map :O I will test it further..

[Jaberwocky]

having a moment to test it, i've observed something that you all might be interested in.

Not sure if a few of you could corroborate this.

I ran a speed test using an online Procedural texture and a home made texture using a .jpg map.

please see the attached results below.

[funk]

Thanks guys. So this needs a Cuda 5 driver for Windows?
Quadro 305.93 will do I guess..

In theory a CUDA 4.2 driver should be enough, but for some reason the texture mapping doesn't work correctly with 301.x drivers. At least on GeForce. Only the latest seems to work correctly. We havem't figured out yet, why.

Cheers,
Marcus

Well I hope you guys figure it out. I installed 306.23 and my PC freezes whenever I render in octane (all versions). I had to go back to 301.42

[matej]

Here are my findings so far:

Running the benchmark scene with a GTX 570 (stock speed) on Linux:

1.) with drivers ver 285.05.09 and CUDA toolkit 4.0.17

- Octane 259: 2.68 Ms/sec
- Octane 302: 2.19 Ms/sec (19% slower!!)

2.) with drivers ver 304.51 and CUDA toolkit 4.2.9

- Octane 259: doesn't work anymore, illegal instruction on start
- Octane 302: 2.08 Ms/sec (additional slowdown..)

Upgrading the drivers & cuda obviously made things worse. Are there any suggested drivers / toolkit to use on Linux?

EDIT: downgraded the cuda toolkit, 259 now works:

3.) with drivers ver 304.51 and CUDA toolkit 4.0.17

- Octane 259: 2.71 Ms/sec (slightly better than with older drivers)
- Octane 302: 2.08 Ms/sec (worse than with older drivers)

conclusion: This new build is 20% slower (at least on the benchmark scene) for me than the old non-instances non-kepler build and the new drivers make it additionally slow down.

[abstrax]

Here are my findings so far:

Running the benchmark scene with a GTX 570 (stock speed) on Linux:

1.) with drivers ver 285.05.09 and CUDA toolkit 4.0.17

- Octane 259: 2.68 Ms/sec
- Octane 302: 2.19 Ms/sec (19% slower!!)

2.) with drivers ver 304.51 and CUDA toolkit 4.2.9

- Octane 259: doesn't work anymore, illegal instruction on start
- Octane 302: 2.08 Ms/sec (additional slowdown..)

Upgrading the drivers & cuda obviously made things worse. Are there any suggested drivers / toolkit to use on Linux?

One problem is that the benchmark scene renders now differently, so you can't compare directly anymore. Without the BRDF changes the latest build would actually run slightly faster. That's obviously hard to prove...

Btw, you don't have to update the CUDA toolkit with beta 3.02. You can leave it at 4.017 or deinstall it. If you leave the 4.0 toolkit, you should still be able to use older versions.

Cheers,
Marcus

EDIT: And one more thing: Could you set up a similar render target, using PMC instead of PT, then switch to this render target and then back? How does it behave? Does the path tracing run slower, as fast or faster than before?

[matej]

I edited my previous post with additional info.

Could you set up a similar render target, using PMC instead of PT, then switch to this render target and then back? How does it behave? Does the path tracing run slower, as fast or faster than before?

I hope I did it right: I started with PT, then changed to PMC, then back to PT - in the same (default) render target, and by creating two separate render target nodes. In bot chases PT runs exactly the same after the switch.
(so no speed-up "bug" for me )