OctaneRender® 1.025 b2.55 test release [Obsolete]

[PeterCGS]

SSS for diffuse materials are broken..

Nothing happens, and Octane doesnt start from "zero" when you change the sliders in the medium node, as it does when using it connected to a specular material. So I suspect the medium node is deactivated when used with diffuse :/

[telemmaite]

SSS for diffuse materials are broken..

Nothing happens, and Octane doesnt start from "zero" when you change the sliders in the medium node, as it does when using it connected to a specular material. So I suspect the medium node is deactivated when used with diffuse :/

+1 !

[pixelrush]

hi guys, thanks for the test release.

1.I notice scenes are darker for some reason now. sunlight needs a wee boost to 1.2 or so? texture power 1.5 or more?? or perhaps colours are more saturated??? not sure what it is, cant seem to attribute it to any particular setting..maybe its my monitor...
EDIT: seems to be due to camera response preset not loading correctly initially.
I have Agfa HDC 100 plusCD. change to another and back and it is brighter HTH

2.there is something going on with specular materials given scattering and emission.
see attach not clearing up getting worse.. ground is diffuse btw

3. scenes seems to load/voxelise noticeably faster

will post more as/if I find it

[Zay]

SSS for diffuse materials are broken..

Nothing happens, and Octane doesnt start from "zero" when you change the sliders in the medium node, as it does when using it connected to a specular material. So I suspect the medium node is deactivated when used with diffuse :/

Yep, can confirm that.

Also using specular makes textures more white now.

[radiance]

SSS for diffuse materials are broken..

Nothing happens, and Octane doesnt start from "zero" when you change the sliders in the medium node, as it does when using it connected to a specular material. So I suspect the medium node is deactivated when used with diffuse :/

Hi guys,

SSS or absorption medium DOES work on diffuse transmission materials,
but you need to make your diffuse material into a diffuse transmission material first, by connecting a texture node float value or colour to the 'transmission' (2nd from the left) input pin first.

Once it's a diffuse transmitter, light does transmit into the object and as such allow you to use a medium node on it for absorption or SSS.

Radiance

[radiance]

Hi guys,

Regarding the default tonemapper problem (the image is darker than in previous versions), this is due to internal changes to the nodegraph system.
You just need to resave your default scene after starting this version, and then you will have the correct defaults again and an identical response as previous versions.
I added this note to the first post.

Radiance

[pixelrush]

I am not sure this applies to this 'test' only but if you have a spec material - the example I am looking at is a car windscreen - when you vary the index the visible amount of reflection varies too.
I wonder if this is correct? since index is internal and reflection is surface... I would have thought they would be independent

I ran the reflection slider up to the max but had to set the index high to get a decent amount of reflection which had the inside view excessively distorted - if you follow what I mean.

[roeland]

Wow, great work ! This is really a huge improvement ! Thank you Refractive Team !

Gabrielefx means the simulation of a thin thickness without haveing a thickness at all. (Simulating SSS on a single surface).

I just see the quote above. I wanted to ask the same question about the use of SSS on thin things like rose petals or tree leaves : As SSS is supposed to be due to refraction in the thin layer inside the material, under the surface, is it necessary to give an actual volume for the petals ? Are they extruded with some thickness ? Or is SSS faked on simple surfaces (faces) without thickness ?

To use SSS you need to create a volume, it will not work correctly on simple surfaces.

--
Roeland

[roeland]

2.there is something going on with specular materials given scattering and emission.
see attach not clearing up getting worse.. ground is diffuse btw

Hi,

Subsurface scattering in specular materials will clear up very slowly or not at all when using very small or distant light sources. In these cases a material with diffuse transmission will work better.

--
Roeland

[pixelrush]

if you say so but I think its broken atm or it seemed to be - just got more and more snow on the ground
I am not sure people would use this combo, sss+emit, but I tried it and it didnt seem right, dunno.
I think we have seen a case of unseasonal pixel fall once before.... could be wrong though

other stuff > tried out 8.5 million face scene - loaded ok. took 12 minutes but there was a fair bit of paging to disk - could do with 12gb ram for 2gb vram - I'm not sure but I think its faster, didnt time it last time but I seem to think it was more like 15-20 mins then.
havent noticed any artifacts

EDIT: managed to stuff in 10m faces - didnt have enough room to change to run PT, just as well DL is so cool these days - 1902 out of 1917mb used with HDR and 1280x1024 res.
load/voxel seemed smooth, everything well behaved, possibly used even more memory than previous betas? not sure, might have used 15gb commit all up with this scene which is more than I remember before, still runs well though, responsive to UI changes. no artifacts noticed.
during load/voxel filled my 8gb ram in about 2m45s, finished in about 13.5 min. guess if I had 16gb ram it would have been ready in <5 min but its ok dont often have scenes that big. an SSD may help

Based on this example only, the new method seems to take up less vram space for geometry (~17.5% less), seems faster to load (~25%? incl. paging), but is more ram hungry??