OctaneRender® pre-Beta 2.3 v4 (win/lin/mac) [OBSOLETE]

[kubo]

a minor bug, if I use a film with and height over 4096 it'll be ignored at the command line and render for whatever that was saved in the file. I haven't test it in previous builds.
Also I like your test 8Eggar8.

[Elvissuperstar007]

you crazy developers ))))))) two days and a new release ) Oo

[radiant]

god i turn my back for two days and you have another one.

[matej]

Let me first say congratulations & thanks to the dev team. The new update version of pre2.3 is really stable (quite more than 2.2, for me) and is a big improvement. Octane Live, outliner, emissions, new nodes... really great! :drooling:

Now a probable bug:
I get this jaggies when applying a bumpmap (procedural or image). This shading bug happens only in daylight lighting model (in pathtracing or directlighting).

Some issues & minor annoyances:

* If a node is detached form a material parameter, the corresponding parameter disappears from the Node Inspector. It has a 'blank' circle representation in the Graph Editor and it's not even accessible from the Outliner anymore. I thik that in this case its value should revert to the default.

* If I create a new Diffuse material node the transmission parameter is not accesible from Node Inspector (blank circle in Graph editor, as above)

* If I change the normalize parameter of emission node from 'bool' to 'bool input' Octane crashes. I'm not sure why I would change this, but I was just randomly pushing buttons & toggles and this produced a crash.

* If I set a HDR for IBL, change to daylight model and then go back to texture environment, I have to reload the image. This is a minor usability annoyance, but still.

[brecht]

Let me first say congratulations & thanks to the dev team. The new update version of pre2.3 is really stable (quite more than 2.2, for me) and is a big improvement. Octane Live, outliner, emissions, new nodes... really great! :drooling:

Now a probable bug:
I get this jaggies when applying a bumpmap (procedural or image). This shading bug happens only in daylight lighting model (in pathtracing or directlighting).

This is a typical problem that happens when the normal doesn't match the actual geometry anymore, there is no ideal solution to this unfortunately. Bump mapping is a trick, and at a certain point it breaks down and you will see the actual geometry that doesn't actually have the shape that the bump mapping pretends it has. There's a few things you can do to reduce this issue:

• Tweak ray epsilon for the kernel
• Reduce bump mapping strength
• Subdivide geometry more

[er_beppe]

hello,
I wanted to make a light bulb, but I ran into this problem:
the emitter does not pass the glass ... I maybe wrong to make glass .. you can not have an explanation on the glass? or is it a bug?

[radiance]

hello,
I wanted to make a light bulb, but I ran into this problem:
the emitter does not pass the glass ... I maybe wrong to make glass .. you can not have an explanation on the glass? or is it a bug?

It's been in the release notes and said many times that direct light does not pass through glass and opacity yet,
so you need to wait a bit more for 2.3 with the new alhgorithm.

Radiance

[Phr0stByte]

I played briefly last night with the linux release, but on my first try, Octane closed by itself after the render was well on it's way. On the second try, Octane rendered for a while, and then froze up. I imported an OCS from 2.2, so that may be the cause, I will play more tonight with a freshly exported OBJ and report back.

running latest nVidia driver and Cuda 3.0

[matej]

This is a typical problem that happens when the normal doesn't match the actual geometry anymore, there is no ideal solution to this unfortunately. Bump mapping is a trick, and at a certain point it breaks down and you will see the actual geometry that doesn't actually have the shape that the bump mapping pretends it has. There's a few things you can do to reduce this issue:

• Tweak ray epsilon for the kernel
• Reduce bump mapping strength
• Subdivide geometry more

No, something is definitely wrong here.
Same scene, exported from Blender, same bump texture used with same scale. Left pot glossy material, right diffuse material:

pre2.3_v4, bump on left 1.0, bump on right 0.2, daylight model

pre2.3_v4, the same render as above, both bumps at MAX, with IBL

[radiance]

Hi,

it's not a bug at all, it's normal.
it's the raytracing 'self-intersection' problem.

set the ray epsilon value to 0.0 in the render kernel parameters to solve it.
However, if you do this for other scenes, like interiors, you might get light leaks,
so our current default is best for most practical scenes.

it's just not good for bump mapped material balls, but you don't sell those images to your customers i assume

Radiance