OTOY Forums

OctaneRender™ for Blender plugin. Windows x64 version.


The bundle consists from two parts:
1. OctaneRender™ Server
2. Blender - OctaneRender™ edition.

The first part (server) should be run on the same computer as the second one. To stop the server - just right-click its icon in the tray.
After the server is run - you can start rendering (or open a rendered view) from inside Blender. Before the very first rendering you should fill in the server address (only 127.0.0.1 at the time), login and password (your license key info) to activate the server. After the first rendering the server will remember this activation info, and you will not need to enter it again.

You should use only OctaneRender materials inside Blender scenes - the plugin knows nothing about Blender specific materials and textures (about any Blender nodes, generally speaking).


FIXES AND NEW FEATURES SINCE LAST RELEASE

• Added MS redistributable v120 to installer.
• Fixed the F12 mode hair bug.
• Minor fixes and improvements...

After starting OctaneServer it will take some time before the LiveDB tree and LiveDB material previews will be available from Blender - as OctaneServer caches the LiveDB data from remote server right after starting. But only the materials groups tree will be re-cached every time after server start, LiveDB material preview images will be cached only once - so only the very first OctaneServer start will need some more time before LiveDB tree will be available in Blender.


DOWNLOAD (both distributives must be reinstalled, they depend on each other)

Blender Octane Edition - 9.5 (2.75) (67.7MB autoinstaller file)
Octane Server - 2.24.1 - 9.5 (10.4MB autoinstaller file)

DEMO is available. Use the same Octane edition of Blender for DEMO version of OctaneServer.

Octane Server DEMO - 2.23.2 - 9.5 (9.5MB autoinstaller file)

For network slave and daemon please download standalone installer here:
viewtopic.php?f=24&t=50202

JimStar wrote: Fixed the F12 mode hair bug.

THANK YOU, THANK YOU, THANK YOU!

I'm in tears!

nice work~~but still need hide emitter feature.........please add it ......

Awesome Thanks!

Hi! Do you plan to merge the "Octane render layers" and "native" Blender layers? "Native" Blender layers is very understandable and intuitive, you can move the object by pressing the "M1", "M2", its quickly and comfortable. If i want to move object to octane layer i must open geometry properties and find a little button of number layer. This quietly, if you have one object. But if i need to move hundreds or thousands objects, I'll have to spend a lot of time on their movement.

Another argument in favor of the merging - the possibility of using a "native" blender laysers at compositing, which is currently supported only for octane layers.

So sweet, thank you!!

Hi Jim,
I'm sorry, but the hair problems got worse in the latest update:

When I F12-render any image - except hair - the new version works as expected. But when I try to render a 'hairy' file, the render gets stuck here: "Elapsed: 00:17.01 | Render-target evaluation on server..."

At this point, Blender freezes permanently in a way I've never seen before: You can't force quit, and you can't start the task manager by clicking Ctrl-Alt-Delete.

So the only way out is to reboot.

I'll attach a screen-shot of my hair settings.

The latest Octane/Blender version where hair renders correct is 7.7. During the next versions, more and more hair systems disappeared from the 12 render (but they rendered fine in the view port).

I should add, that I have 8-10 different hair systems in each file.

justavisitor
This is not the bug.
You've just run out of memory resources during rendertarget evaluation due to huge amount of hair data.
Blender uses for live-view display the amount of hair-children set in "Display" field with max 5 vertices per each hair and per each child. When you F12-render them - Blender uses the amount of hair-children set in "Render" field with max 65 vertices per each hair and per each child. In your case only for hair-children this is 20 x 5 = 100 children-vertices per each hair for display (that your computer still handles), and 100 x 65 = 6500 children-vertices per each hair when you render it. That is - 65 times more. If you have the big amount of hairs consisting of 65 vertices each - having in addition 100 children per each hair consisting of 65 vertices each will definitely hang your machine if you have not enough memory.

JimStar wrote:justavisitor
This is not the bug.
You've just run out of memory resources during rendertarget evaluation due to huge amount of hair data.

Sorry, but it is a bug:

Again, version 7.7 F12-renders the image correctly in 3.5 minutes without any problems, what so ever!

The next versions also F12-render the image in about 3 minutes, but they leave out more and more hair-systems. And the latest version doesn't F12-render the image at all.

EDIT: I should add that I tried 3 different sets of Octane/Blender versions today - the correct one (7.7); the one I currently use (9.3); and the latest (9.5), one after another - to make absolutely sure that no other settings or factors could affect the results.

EDIT 2: You might ask: "Why don't you just use version 7.7 since it renders the file correctly?" Well, my problem is that I have to render animations. And version 7.7 has some animation bugs that you fixed in the following versions. So if I have to use version 7.7, I'll literally have to render thousands of stills manually, one by one. And that's not exactly my idea of fun...

EDIT 3: I don't know if this helps, but when I render the image in vs. 7.7 or 9.3, I can hear that the machine is working because the fans start during rendering. However, the fans never start at all when I try to render the image in version 9.5...

justavisitor wrote:it is a bug

justavisitor wrote:version 7.7 F12-renders the image correctly in 3.5 minutes without any problems, what so ever!

If rendering 5 vertices per hair instead of 65 presented by current versions of Blender is "correctly" and "without any problems" - then you know anything better than me.
Your "bug" have appeared exactly due to the reason this was fixed in last version and now renders the correct amount of hairs, children and vertices-per-hair presented by Blender in F12 mode.