OctaneRender® 1.024 beta 2.48b TEST (lin/win) [OBSOLETE]

[YoonKyung]

I am aware of the fact octane only uses gpu for rendering.
The fact is octane is not the only renderer I use...

You can use CUBIX, a PCIe expansion box.

[pixym]

Yes I know that as well but the fact is I really need more CPU power for my other renderers

[YoonKyung]

YoonKyung, do you mind telling me what kind of mobo and PSU you use for 4x590? Nvidia seems to run late on their kepler schedule, and I might need to upgrade soon to something faster thatn 2x480... 4x590 sound like it might be up my alley

/Regards Peter K

I am using a Lian's case which support 8 slot
+ a moher board, "ASUS P7T6 Supercomputer"
+ a Thermaltake's 1500W PSU + a 600W VGA PSU.
And, I added 4x 14 inch fan for nice air flow.

[abstrax]

ok sorry for bothering but i updated video card driver and cuda ...i see only GTX 470 in my inactive video card and i cannot add it to active list..

What can be wrong ?

It's a (stupid) bug in beta 2.48/2.48b. It will be fixed with the next release. I'm sorry about that.

Marcus

[pixelrush]

a test 'C' coming soon? or we wait for beta 2.5 now?

[abstrax]

Hello, is there anyone working on big scenes (at least 4 million triangles) ? I saw no comment about the trouble with the material picker delay, so I copy/paste here a part of the results of my tests reported on page 4 in this same topic :

Hi Roubal,

we will investigate this issue next week. The problem is not the geometry size, but the output size and we have some ideas to improve things, although the reported times are still too excessive in my opinion.

We are currently fixing some other issues and try to make the Mac build work. Then we are planning to release a beta 2.48c with those changes (hopefully this Friday/weekend or so). After that we plan to investigate the interactivity problems and then either fix them (if it's an easy fix) or do the offline mode work first, if more work is required, and then do those interactivity changes afterwards.

Thanks for your help and feedback, it's much appreciated.

Cheers,
Marcus

[abstrax]

I agree that the aperture doesn't work right in Octane. It doesn't behave like real aperture so maybe they should name it something else.

What do you mean? As far as I can see it works correctly. The aperture is the diameter of the "virtual" lens in centimeters.

Cheers,
Marcus

[Refracty]

I think he meant two things.
1. A change in the aperture should be linked with a change in brightnes. When the hole is smaller it gets darker (simply said).
2. The fstop value is obsolete, because it is the unit of defining the size of the apperture (√). So there is no need to have apperture and fstop simultaniously. But shutter value is needed. So if you rename fstop in shutter and give it a logical number range this is fine.

I agree that the aperture doesn't work right in Octane. It doesn't behave like real aperture so maybe they should name it something else.

What do you mean? As far as I can see it works correctly. The aperture is the diameter of the "virtual" lens in centimeters.

Cheers,
Marcus

[vipvip]

i think the goal is not to reproduce the real cameras ( with all his optical constraints-> i do 'real' photographs for my job...) but just to define right parameters without confusion but flexibility:
Please look at the pdf attached: it would be nice to have such parameters fort controling the 'deepness" of DOF. Do you think it would be possible ? It will be a great help for creating virtual photos with octane . Just this kind of settings should be in conflict with actual aperture/focaldepth parameters... perhaps it should be a solution...

[Daniel79]

There are many factors that affect depth of field. The most important are the focal length, the distance of the subject and the camera's aperture setting.
Now...in octane the only value that changes the DOF is "aperture" setting. this is not correct ok?
You can make sure that the depth of field is managed as a real reflex? (like Maxwell render).