OCTANE DEVELOPERS READ THIS THREAD TO SEE OCTANES ISSUES Tweet

[Notiusweb]

I like the list you have for Octane updates, particularly the Denoise one. The CPU one and the 'non-Titan' don't mean too much to me with cards I have.
Do all of these features exist in Arnold and VRAY presently? (like the ones about Volumetrics, AO, and OSL?)

[Tag12345]

I like the list you have for Octane updates, particularly the Denoise one. The CPU one and the 'non-Titan' don't mean too much to me with cards I have.
Do all of these features exist in Arnold and VRAY presently? (like the ones about Volumetrics, AO, and OSL?)

Its a mixed bag when it comes to Arnold - Vray - Thea - Redshift


have tried them all to one degree or another.


Redshift has capabilities to have Light Linking and stylistic lighting.

Vray has really good memory management -- denoise system too - and can use CPU/GPU in Vray RT

Thea is an underrated one as well, and can use GPU and CPU really really well.

Arnold has many of the Volumetric aspects I can see being changes in Octane. They use a Volume Material as a Global controller, and allow for light tags to handle individual volumetric settings with each light object.
Arnold doesnt use GPU yet, but they have teased it, but I do know its pretty quick on CPU.

As far as AO and OSL, im not too sure redshift has that yet, i may be wrong. I know Thea could calculate AO since it had boost from CPU, and not sure if Vray RT already does that.

[Tharso]

http://www.cgchannel.com/2016/04/otoy-r ... of-octane/

OctaneRender 3.1: due later in 2016
Further off in 2016, OctaneRender 3.1 will include a number of interesting new technologies.

I am waiting and waiting and waiting ...............

[Goldorak]

Latest roadmap here:

https://www.facebook.com/groups/OctaneR ... 643308749/

[Vejzovic]

I thing that Goldorak is tired of us (octane customers). It is not our fault you promise too much and then can´t deliver. Next time for octane4 do more and promise less.

[Notiusweb]

I like the list you have for Octane updates, particularly the Denoise one. The CPU one and the 'non-Titan' don't mean too much to me with cards I have.
Do all of these features exist in Arnold and VRAY presently? (like the ones about Volumetrics, AO, and OSL?)

Its a mixed bag when it comes to Arnold - Vray - Thea - Redshift


have tried them all to one degree or another.


Redshift has capabilities to have Light Linking and stylistic lighting.

Vray has really good memory management -- denoise system too - and can use CPU/GPU in Vray RT

Thea is an underrated one as well, and can use GPU and CPU really really well.

Arnold has many of the Volumetric aspects I can see being changes in Octane. They use a Volume Material as a Global controller, and allow for light tags to handle individual volumetric settings with each light object.
Arnold doesnt use GPU yet, but they have teased it, but I do know its pretty quick on CPU.

As far as AO and OSL, im not too sure redshift has that yet, i may be wrong. I know Thea could calculate AO since it had boost from CPU, and not sure if Vray RT already does that.

Like as far as Ambient Occlusion, you want a greater set of variables to tweak, like an amount depth, sharpness, blend sort of thing, as opposed to a single AO distance node? Isn't Octane giving us Ambient Occlusion already?

I'll be honest, unless my CPU can crank out speeds faster than my GPUs can, I'd rather direct my CPU doing compiling as fast as it can! But I know you'd like the option, at minimum.

The emitters to me are the golden trophy, if they could glow and cast light in an emulative sort of way, as opposed to having to cast light 'true'. I wouldn't care if it's not accurate, as long as I could make it look accurate. Like, I could render a single frame with 'true lighting' on, and then shut it off and use a false-lighting mode. I would use the real lighting render as a reference-benchmark, and then I would emulate that dynamic best I could with the fake emitter lights.
Like a fake emitter lighting system that would not need to be computed accurately, just be superficial, as a short cut, for faster render speeds. I guess instead of going after the materials to save on computation (like PBR) I would go after the lights themselves. It might look innacurate and messy if such lighting was done wrong, but if you could benchmark one scene off real lighting and then emulate that to make it look right, could be nice!

Yes, that is my wish - Fake Emitter Lights!

[coilbook]

not sure why some complain about AS. For animation it is great. Render time went from 8 minutes to 2. The only noise issue we have is speclular in DOF and volume fog. Hopefully they will fix it

[Goldorak]

I thing that Goldorak is tired of us (octane customers). It is not our fault you promise too much and then can´t deliver. Next time for octane4 do more and promise less.

Not at all. User feedback is important for us and it is incredibly helpful when it points out something we missed and that we can do better - including dates and roadmap communication..

Octane 3.x included two very ambitious features - OSL GPU compiler and AMD support - that took longer than we thought. The latter is done, after a year of working with AMD on driver stability, and literally building our own CUDA cross compiler for AMD GPUs. We still have a ways to go on getting AMD GPU perf tuned, but we made huge progress this year, to the point where AMD has fully QA's Octane for AMD independently of us.

Completing OSL will be our primary focus in the next few updates after 3.07 is done, along with many smaller improvements, some announced, others not.

We've committed to one high level feature in V4 - that Brigade will be integrated in Octane, so that speed/interactivity (especially in scene graph layer) will be significantly improved. A lot of other things will become possible in V4 with this as a foundation.

[Tag12345]

I thing that Goldorak is tired of us (octane customers). It is not our fault you promise too much and then can´t deliver. Next time for octane4 do more and promise less.

Not at all. User feedback is important for us and it is incredibly helpful when it points out something we missed and that we can do better - including dates and roadmap communication..

Octane 3.x included two very ambitious features - OSL GPU compiler and AMD support - that took longer than we thought. The latter is done, after a year of working with AMD on driver stability, and literally building our own CUDA cross compiler for AMD GPUs. We still have a ways to go on getting AMD GPU perf tuned, but we made huge progress this year, to the point where AMD has fully QA's Octane for AMD independently of us.

Completing OSL will be our primary focus in the next few updates after 3.07 is done, along with many smaller improvements, some announced, others not.

We've committed to one high level feature in V4 - that Brigade will be integrated in Octane, so that speed/interactivity (especially in scene graph layer) will be significantly improved. A lot of other things will become possible in V4 with this as a foundation.

How long (estimated) before we see V 3.07?

As well as maybe v4?


Honestly if v3.07 has all we need when it comes to faster and more efficient rendering when it comes to having Noise reduction, then thats definitely a start in the right direction, and can hold most of us off for Render engine reworks -- Im guessing most of this will be done in v4 though since Brigade is taking an approach similar to how Unity and other game engines work .

But like I said, it would be nice to have your guy's team at least shine light, in a sticky thread about any delays and maybe a timeframe for certain versions. I feel, by knowing this stuff -- even if late than the expected date -- fans and users would be more lenient of the long waits in progress. Right now, it feels that communication aspect is distancing slowly, and yu guys only respond when need be, but dont actually interact by taking the initiative to ask the community questions. Maybe what people think. Feedback,etc.

Please start by making a sticky thread and then go ahead and let people know any challenges you guys are having/ or had. Connect with us, and let the community feel connected to you guys.
I understand you all are devs, but you guys are people too, so dont be afraid to speak out and let us know.

Anyways, thats just my suggestion. Very eager for 3.07. I havent had a great time fiddling with octane lately since speeds with things like DOF,SSS,AO,Displacement -- slow it down with larger scenes tremendously. And thats with 2 GTX 1080's non -SLI


Just feel like that amount of power, should do wonders. Shouldnt need 4+ to get things working the way I need.

[promity]

Another good idea is for implementations in the Octane

Vray_GPU_001.jpg

I.e. access data between GPU's.