OctaneRender™ Standalone 2.16

[shawnfrueh]

I think there is a big misunderstanding here. Octane's Fstop and ISO have never worked like a real camera. A "technical" user who knows how a camera works could never have used them to produce "correct" results. Removing those pins hasn't removed any "accuracy".

This should be something that is then addressed by the Octane team. Instead of simplifying an attribute they should be staying true to their statement and make the Fstop and ISO, with the addition of a shutter speed, work correctly:

Octane Render doesn’t sacrifice accuracy in order to provide a 10x to 50x speed increase over traditional, CPU based unbiased ray tracers.

The camera parameters are part of the final calculation and thus I consider it part of the "accuracy." They can still keep their simplicity and have that single slider change the values within Fstop, ISO etc.. and vice versa.

[enricocerica]

Hi,

I have two GTX580 in my cubix expansion box, it works fine if I activate only one card but if I activate both of them, it crashes very often with a BSOD. I discovered that if I set the Minimum display samples to at least 2, I have no problem, as soon as I set it back to 1 it crashes again.

Any idea why I got this problem and if there is a way to avoid it as I don't always think to set it to 2.

[augdalmo]

I think the section dedicated to the camera should be developed and implemented as it should.

Currently there is no possibility to change the focal length, and even to change the diaphragm, in an independent manner, and then the depth of field does not reflect a true optical. The adjustment of the opening should not be removed but implemented with the ability to adjust the shutter speed.

Clearly the commitment to develop these functions is significant but still possible and desirable.

Seems to me that Maxwell render has implemented a series of optics with different focal lengths that are based on real physical composition of the lens and then automatically you get adjustments focus based on the depth of field effect and not fictitious. The light passing through a lens is regulated by several factors, however these are currently free will of those who use the program and therefore do not reflect an actual behavior of optics.
I do not understand what it's for a camera with the parameters blocked, when varied the scale automatically changes the opening! so I can not change the focal length, it seems logical? Very limiting. Not at all real.
With C4D I can adjust various parameters of the camera, I can not say the same for Octane Standalone.
What is the sense that the software has the default opening of 2,8F? Which is almost the maximum for a common optic?

If we boast a render engine unbaiased, why the camera is not real?
Think about it ...

[Elvissuperstar007]

ctrl+z not work
Standalone 2.16

[kavorka]

ctrl+z not work
Standalone 2.16

It is working for me in Win7. What actions are you trying to undo?

[Rikk The Gaijin]

Hi, I just found a problem with the normal maps.
As you can see from the image below, the rocks should all point outward, but it seems some are pointing inward.

error.jpg

The same scene in Maya looks correct, all rocks are pointing upward.

maya.jpg

The normal map has rotated parts. Normally this should not be a problem, but it seems Octane fails to calculate the normals correctly.

normal.jpg

The problem exists in 2.16 and 2.15 as well.

[grimm]

Hey Rikk,

I think that is an optical illusion. The light in Maya is very directional, while the Octane light is non-directional. Have you tried to look at the scene with daylight at a low angel?

[Rikk The Gaijin]

Hey Rikk,

I think that is an optical illusion. The light in Maya is very directional, while the Octane light is non-directional. Have you tried to look at the scene with daylight at a low angel?

No, it's not an optical illusion. And the light is directional in both Maya and Octane.
To demonstrate that, I changed the normal map with another one full of spheres. It's clear that Octane renders them wrong.

test1.jpg

Again, in Maya they are displayed correctly.

test2.jpg

Note how all the spheres are shaded in the same exact way in the normal map

norm.jpg

[grimm]

Normals flipped for some reason?

[Rikk The Gaijin]

Normals flipped for some reason?

No, the maya Viewport 2.0 would displayed them black, if that was the case.
Really, i think the problem is in the UVs. Some UV islands are rotated, and that seems to confuse Octane.