OctaneRender® 1.024 beta 2.47 (lin/mac/win) [OBSOLETE]

[mib2berlin]

Hi, get it to work with beta driver 275.09 and Cuda toolkit 4.0.17 but with PMC i get the same like face.

Hmm, mib.

Pathtracing

PMC

[face]

Ha, im not alone

face

[Draydin_r]

Turns out it was the normal map, note to self: focus the normal map detail when faced with mixed light sources. Thanks to both Kubo and Face for the advice!

[kubo]

I was testing my alien shore entry for the past contest on this built, since there is a huge ice chunk I thought it would look neat with dispersion, as it turns out, it does, but... I got the same "error" than face and mib2berlin, the thing is that everything looks rather cool and balanced in pathtracing (althou I've had to pump up the lights A LOT from the saved scene, not eficency, which was at .25 since it was done with a previous version, but the power of lights from 70 to 5000) but when I swich over to PMC the screen looks ok for the first 8 samples and then it goes bright and looks exactly like the examples posted by face and mib2berlin. So I went and play with the camera values and if I upped the fstop to 65! no less then I would see the scene exactly as it was seen in PT. I'll post some test shots later, just to let them that they are not alone and that it somehow is related to the exposure values in PMC.
@ Draydin_r, glad you found out, it looks really good now btw

[Daniel]

Got it going!

Render time: 53 minutes
Max samples: 16,000

[Jaberwocky]

As requested Radiance

1 dispersion render test on a glass pyramid using PMC

Aprox 1 hr - 12000 samples

[Jaberwocky]

Oh yes... i had a play.....


Inspiration:

Pink Floyd's Dark side of the moon Album cover.

[kubo]

well, here is my failed experiment, not only from sucessful ones we learn . This is my alien shore scene. I wanted to add disspersion values to the ice blocks. Since there is 2 builts in between I had to upped the eficency for the moon and the hand held "torch", that was expected, what was not is that yet with those values even in PT (in which was originaly rendered) couldn't achieve the same exposure, I was getting almost dark shot in PT, so I upped the power of both lights almst to ridiculus values 6000 and 1500 each, but the surprise came when I saw that in PMC, it was totally overexposed giving that blue noise problem, if I rised the fstop to 60 it looked "ok" yet if I lowered the power of the lights I would get no shadows at all. Also, quite curious, the lamp in the alien's hand was changing colors all the time, from cyan to red to several other hues in PMC. See for yourself, any ideas anyone?

[andrian]

For those who are interested, I have created a physically based diamond material, based on a study of the reflection and refraction of light in a diamond by Marcel Tolkowsky. You can found it in LiveDB under Organic/Stone. Test it and give some nice renders with it

[bepeg4d]

For some scenes there can be quite some bias in the image after the first few samples. Is the brightness in your scene the same as with path tracing after 2000 samples?

--
Roeland

yes, as already said by kubo, with dl and pt there are no varietations in the exposition during the render but in pmc after the 16 sample there is a costant increase of the exposition. since it was a scene created with 1.24.6b, i have saved all the necessary nodes and i have recreated the scene in 1.24.7 from the beginning and now all it's ok and also the noise seems to clear faster.
here is the result afer 4800 sample
ciao beppe