OctaneRender™ Standalone 2.10

[kavorka]

I was just checking these new features. Undo works like a charm
Coherent mode gave me a nice speed improvement on the scene I tested it in.

I was checking out the render passes and I really like how it is implemented. I did get a crash when I was looking at the AO pass and started to adjust the distance slider.

[funk]

Thanks guys!

I noticed a bug with the diffuse / glossy / specular render passes. If you use hotpixel removal, it actually changes the look of the pass instead of just removing hot pixels. It seems to be removing dark parts of the pass and making them bright. Really odd.

[funk]

It might also be handy to give us the compositing math

Which passes add up to the beauty pass?

[FrankPooleFloating]

OMFG! It figures this would come right when I am in the middle of a JS programming gig, and won't be able to touch Octane until this weekend at the earliest! D'oh!

Someone tell me this: Do hot pixels only exist in certain passes?.. If so, they will be much easier to clean up in PS. I am probably one of the few that would rather manually clean some HPs (even a couple hundred - methodically going left to right, down a chunk, right to left...back and forth, using clone tool and spot healing brush) than do a 10hr cook.

[Seekerfinder]

Looks like a great release, thank you.

How many undo cycles are allowed?

Seeker

[smicha]

Someone tell me this: Do hot pixels only exist in certain passes?..

Passes with HP: IL, Refl, Spec, Diff.

I have't tested scenes with SSS yet.

[riggles]

Good to see the speed return. Looks like 1.5 users who were unhappy with sizable 2.0 speed loss can feel better. In a simple interior test, 1.5 still had higher Ms/s than 2.1 (10ms/s vs 9), but after pausing the render 2 minutes in, there was less noise in 2.1 (term power set to default .3). Raising the term power up to .5 increased Ms/s to 12, but left some more noise in the shadows at the 2 minute mark. This could be very different story in a scene less dominated by shadows.

Also, coherent mode is a must if you want to see speed on par with 1.5, according to my test.

Still my personal view on speed is that focusing on Ms/s is fun and easy for all of us to do, but it's deceiving and not totally relevant in production. It's about balancing the best final render quality in the time allowed. In this admittedly simple test, 2.1 seems to have improved over 1.5, not just 2.0.

best viewed at full size

[bepeg4d]

Super great thanks for the big improvement
here's a couple of screenshots of render passes in action with normal beauty and sss of the classic cube fog:

wonderful
Unfortunately i wasn't lucky with install_demon script in both mac and pc
after installing, all the network goes down and i had to uninstall the demon and fight a bit to reset everything, here is the mac terminal window:

OR2-10firstopen-demon-mac.txt

(4.66 KiB) Downloaded 330 times

i'll try again as soon as i can, anyway the other features are fantastic, thanks again

[Sakalakapaka]

Big thanks for this great update

Octane Team.
Any words about poor performance of the new nvidia 9xx cards on Octane Render ? Is it a drivers problem, Octane engine problem or these cards are so bad in GPGPU computing ? I must buy a new graphic card, and don't know which I should choose, GTX 780Ti or two GTX 970.

One more thing. PMC kernel seems to be much slower in 2.1 version. In the Trench scene I only get 2.00 Ms/sec, compared to 3.45 Ms/sec in the 2.06 version. In another project speed decreased from 2.45 Ms/sec to 2.25 Ms/sec. GTX 580 3GB.

[smicha]

Big thanks for this great update

Octane Team.
Any words about poor performance of the new nvidia 9xx cards on Octane Render ? Is it a drivers problem, Octane engine problem or these cards are so bad in GPGPU computing ? I must buy a new graphic card, and don't know which I should choose, GTX 780Ti or two GTX 970.

If I were you I wouldn't even think about 970