A forum where development builds are posted for testing by the community.
Forum rules
NOTE: The software in this forum is not %100 reliable, they are development builds and are meant for testing by experienced octane users. If you are a new octane user, we recommend to use the current stable release from the 'Commercial Product News & Releases' forum.
justix wrote:2nd hand Gtx 590 are still more expensive than a new Gtx670..
And that should be your main guideline for assessing the card speed (at hardware level, anyway)
6xx cards have 2x (or whatever) more "cuda cores" but at the same time they cost like 1.5 less than 5xx. Unless Nvidia became insane over night, they wont sell you a 2x faster card for half the price, no?
Forget cuda cores, stream processors, yada yada... Look at the price, because nvidia already converted speed into price for you.
Octane standalone doesn't know about timeline and animations, so it doesn't support image sequences. But a plugin / exporter could supply Octane a different image each frame, so this is a question for a specific plugin / exporter.
justix wrote:2nd hand Gtx 590 are still more expensive than a new Gtx670..
And that should be your main guideline for assessing the card speed (at hardware level, anyway)
6xx cards have 2x (or whatever) more "cuda cores" but at the same time they cost like 1.5 less than 5xx. Unless Nvidia became insane over night, they wont sell you a 2x faster card for half the price, no?
Forget cuda cores, stream processors, yada yada... Look at the price, because nvidia already converted speed into price for you.
I don't think Nvidia is even thinking of us..if they changed the architecture is because they went for new memory management / less power consumption etc....but it seems LESS computation power...But If we could run a 670 as fast as a 590 (corewise) would be everyone's target no? I'm not this wealthy to buy a new Motherboard..2 X 5xx or 6xx just in order to achieve a full working 1000cores + that could match a 590 speed or more...now I still need to understand IF is Octane programming that needs to adapt to the new programming architecture OR if the 6xxx is 'as it is' and so is just an hopeless business and need to wait for a crackdown.
Win 7 64 | 2 X MSI AERO GtX 1070| Intel I7-6850K| 32 GB DDR4 RAM | Asus X99 II-A
justix wrote:now I still need to understand IF is Octane programming that needs to adapt to the new programming architecture OR if the 6xxx is 'as it is' and so is just an hopeless business and need to wait for a crackdown.
Octane most likely needs to be optimized for the new arhitecture, but do not expect Kepler to be the same as Fermi, core-wise. It just doesn't make sense. I'm not an hardware expert, but as I said, the price of the card will tell you more about its performance than the number of cores. Octane will get better performance on Kepler, but it wont be the same performance per card, compared to Fermi. (so yes, that means the need for more slots)
What I'm also concerned is: will Octane optimization for Kepler, when the two builds will be merged into one, bring worse performance on Fermi? ie. performance on Kepler will improve, but such build will be less fast than a Fermi specific one would be? If yes, than in a way we are being screwed by Nvidia...
renmaxhb wrote:
1.Option is very slow in maya,and creat LIVEBD, some matierals,some light........
I have not understood, please spell it more clear...
renmaxhb wrote:
2.It is very slow during exporting maya to engine,my animation scence cost 40's at least.....It must Optimizate.
Take a look at "voxelizing..." time in Octane verbose ouput. If most of these 40 seconds are spent there - this question you should ask to Octane engine developers, not to plugin developer. The export itself (that is made by plugin) is very fast in comparison to Octane engine voxelizing process.
renmaxhb wrote:
3.Light sample is very insetresting,
one light,glable sample is 100,light works very fast and good,
two lights,glable sample is 100,lights works slowly and quality of images is bad ,so glable sample must be 1024
ten lights,glable sample must be 3000?............you change lights sample ,the light with lower sample works badly.
The same thing: all that relates to the rendering itself - is area of responsibility of Octane engine, not the plugins, so ask please about it to Octane engine developers.
renmaxhb wrote:
4.It can not suport images sequece,engine is also.... how can i make a animation of water?
A few days ago I already added your feature request about it to the TODO list to one of the first positions. It will be supported in one of the next versions soon. I'm working on it.
renmaxhb wrote:
5.LIVEBD is very good. But I found you shutdown the server and restart it sometimes evrey day? So I can use it at day,I cant sue it at night?
Server is not the plugins developers responsibility, I don't know how it works.
renmaxhb wrote:
6. the passes is not enough.
The same thing: all that relates to the rendering itself - is area of responsibility of Octane engine, not the plugins, so ask please about it to Octane engine developers. I can't add the passes to engine, as in this plugin I just use this engine through SDK API.
renmaxhb wrote:
7.It works slowly when matierals have some transperacy texture.It is very obviou....
The same thing: all that relates to the rendering itself - is area of responsibility of Octane engine, not the plugins, so ask please about it to Octane engine developers...
pixelrush wrote:One thing that has been bothering me a little, and I think this issue has worsened in recent betas, although I might be imagining it, is the smoothing. To my mind the smoothing is not quite functioning as it should.
I remember Roeland tweaked something recently and perhaps this arises from that.
Attached is a close up of a small mechanical part that shows some faceting where I think it really ought to appear smooth. Also attach is the normals render. Clearly smoothing is working to some degree...
What concerns me is the polygonal appearance of the flat surface and the here and there creasing going on around the curved faces.
Has anyone else noticed this issue or thought it is more obvious lately?
I think there is sufficient tessellation in this model for it to appear quite smooth and nicely curved but I could be wrong
Are those the shading normals or the geometric normals? If it's the geometric normals can you post an image of the shading normals?
There was a fix to smoothed materials in beta 2.58e, is there a big difference in shading with earlier versions?
Please make the pasted node appear under the current cursor location in GE. With the current implementation, when copying nodes between tabs, they can appear in unexpected locations completely off-screen.
