OTOY Forums

Posted: **Tue Jun 10, 2014 12:10 pm**

This is the recommended thread for downloading, and posting any questions, suggestions or support queries for OctaneRender for Poser STABLE RELEASE.

Pls read http://poserphysics.blogspot.com.au/2016/03/before-you-submit-octane-support.html prior to posting support issues.

Poser Version
Windows: Poser9/PoserPro2012, Poser10/PoserPro2014 and Poser 11/11 Pro, on Windows XP, Vista, 7 and 8.
OSX : PoserPro2012, PoserPro2014 and Poser 11/11 Pro 64bit, on MAC OS X 10.9.4 onwards.

Nvidia Driver Version
Windows : Nvidia drivers 344.x or later. Avoid Nvidia driver 361.x.
OSX : CUDA Driver Version: 6.5 or later

Network Rendering
For Network Rendering, you must use the Octane slave release 2.25 available from the Commercial Product News & Releases subforum.

If you have a support issue, pls provide the following information:
- Operating System (ie. Win XP, Vista, 7, 8)
- 32 or 64 bit
- Amount of RAM
- Graphics Card(s) - model (ie. GTX 580, 3GB)
- Nvidia driver version
- Poser version (Poser -> Help -> About Poser) - list the build number (eg. 9.0.3.23027)
- OctaneRender for Plugin version (ie. 2.23.1.56)

Also have ready a copy of
- octaneplugin.log and cuda.log file from your Poser temp direction (Poser -> Edit -> General Preferences -> Misc tab to find the temp directory), and
- config.xml, which is in C:\Users\[username]\AppData\Roaming\Poser Pro\9 for Poser 9/2012 and C:\Users\[username]\AppData\Roaming\Poser Pro\10 for Poser10/2014.

Still TODO
- Implement subd processing for props (only implemented for figures at the moment) - will require a Poser SR patch
- Specular map image power defaults to 1 at the moment for skin materials with specular maps. Probably should be about 0.2.

If you have questions....
- Check the specific Octane 2.0 features in this plugin http://render.otoy.com/manuals/Poser/?page_id=298
- Read the latest up-to-date on-line manual
- Read the PDF manual (pdf manual is installed into C:\Program Files\Smith Micro\[poser version]\Documentation)
- Read the FAQ
- Watch the Features Video
- Read the Features Blog Posting
- Watch the Introductory Tutorial
- Watch the Instancing Video
- Watch the HDR Light Studio Video
- Check the Facebook Page
- Post your question on this forum thread

Limitations
Refer to the on-line manual

Download Links
Windows
- OctaneRender for Poser 2.25.0.68 (64 bit)

The previous plugin release can be downloaded from:
- OctaneRender for Poser 2.24.2.62 (64 bit)

The last version of the Windows plugin that was built with 32bit support can be downloaded from:
- OctaneRender for Poser 2.24.1.61 (32 bit)

OSX
- OctaneRender for Poser 2.25.0.68 Poser Pro 64bit (Mac OS X)

The previous plugin release can be downloaded from:
- OctaneRender for Poser 2.24.2.63 Poser Pro 64bit (Mac OS X)

Paul

Posted: **Tue Jun 10, 2014 1:06 pm**

working fine at this end first of 2 test renders
with version 2.0

Posted: **Tue Jun 10, 2014 3:40 pm**

A couple of minor issues I've come across:

1. In the Render Setup window, if you select a mesh and change one of the parameters and then press Enter, cursor focus returns to Poser even though it appears the cursor is still in the Render Setup window. Clicking on the field in the mesh settings still doesn't return focus (focus stays in Poser). If you click a different field (for the same mesh), you can no longer return to the original one by mouse-clicking. Selecting another mesh in the Material List tree and then re-selecting the first mesh seems to fix it. Maybe a picture would come in handy:

This is different from the behavior when using the keyboard to enter material/texture settings (e.g., rotation/scale values) - those seem to work like you'd expect.

2. Applying subdivision (1) to a particular mesh causes some errors, and then Octane gets stuck here:

Then the plugin freezes and I have to force close Poser.
Here's what the "mainWacros" window says:

Code: Select all: Traceback (most recent call last): File "C:\Program Files\Smith Micro\Poser Pro 2014\Runtime\Python\addons\OctaneRender for Poser\SetupForm.py", line 594, in EventOpenViewport File "C:\Program Files\Smith Micro\Poser Pro 2014\Runtime\Python\addons\OctaneRender for Poser\UiManager.py", line 5139, in ShowViewportForm File "C:\Program Files\Smith Micro\Poser Pro 2014\Runtime\Python\addons\OctaneRender for Poser\RenderManager.py", line 336, in StartRender WindowsError: exception: access violation reading 0x00000000000000BC

Here's the CUDA error log:

Code: Select all: Started logging on 10.06.14 11:21:11 OctaneRender version 2.00.0009 beta (2000009) Subdivider: A non-manifold edge incident to more than 2 faces was found Subdivider: An edge connecting two vertices was specified more than once. It's likely that an incident face was flipped Subdivider: A non-manifold edge incident to more than 2 faces was found Subdivider: An edge connecting two vertices was specified more than once. It's likely that an incident face was flipped Subdivider: Face 902 will be ignored Subdivider: A non-manifold edge incident to more than 2 faces was found Subdivider: An edge connecting two vertices was specified more than once. It's likely that an incident face was flipped Subdivider: Face 903 will be ignored Subdivider: A non-manifold edge incident to more than 2 faces was found Subdivider: An edge connecting two vertices was specified more than once. It's likely that an incident face was flipped Subdivider: Face 904 will be ignored Subdivider: The specified subdivmesh contains disconnected surface components.

Apparently there's something wrong with the mesh, but a freeze seems like a severe reaction. :lol:

[Edit] I just remembered that subdivision isn't supported for props (figures only), is that the problem? But still, a freeze seems a bit severe?

My specs are:
Win 7 Professional SP1
nVidia driver version 337.88
Poser Pro 2014 (10.0.3.26510)
Core i7 2600, 8GB RAM
Two GTX 770 4GBs

Thanks!

Posted: **Tue Jun 10, 2014 5:26 pm**

With regards to subdivision - figure only subdivsion is Posers subdivision. Octane subdivision works for both
The figure/prop needs to be welded for Octane subdivision (and rounded edges).
Did you run the Weld script in the Octane Plugin menu in Poser?

Posted: **Tue Jun 10, 2014 8:45 pm**

Ah ok, thanks for the info re: subdividing props.

I hadn't tried welding it, but I just did, and I still get the same errors. The prop renders fine with 0 subdivision, though.

Posted: **Tue Jun 10, 2014 10:06 pm**

I've been getting a lot of Cuda Error 702's, especially when messing with displacement. Sometimes causes Poser to crash as well, but not always.

Here's the log after one error (I thought I cleared it before):

CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device array
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to deallocate device memory
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Could not get memory info
CUDA error 702 on device 1: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Failed to unload module
CUDA error 702 on device 0: The device kernel took too long to execute. This can only occur if timeouts are enabled. The context cannot be used anymore and must be destroyed. All existing device memory allocations from this context are invalid and must be reconstructed.
-> Kernel execution failed (pt)
CUDA device 0: Path tracing failed

???

FYI, I have two GTX 670's (Though I typically only use one for rendering to keep from lagging the computer).

Posted: **Tue Jun 10, 2014 10:50 pm**

1. In the Render Setup window, if you select a mesh and change one of the parameters and then press Enter, cursor focus returns to Poser even though it appears the cursor is still in the Render Setup window. Clicking on the field in the mesh settings still doesn't return focus (focus stays in Poser). If you click a different field (for the same mesh), you can no longer return to the original one by mouse-clicking. Selecting another mesh in the Material List tree and then re-selecting the first mesh seems to fix it. Maybe a picture would come in handy:

Yes, I noticed this two, and spent many hours trying to fix it - by python conspired against me. I will take another look.

2. Applying subdivision (1) to a particular mesh causes some errors, and then Octane gets stuck here:

Can you send me the prop pls?

I've been getting a lot of Cuda Error 702's, especially when messing with displacement. Sometimes causes Poser to crash as well, but not always.

Potentially the displacement amount is too high. Pls refer to the "Displacement" section of the manual at http://render.otoy.com/manuals/Poser/?page_id=298.

Paul

Posted: **Tue Jun 10, 2014 11:31 pm**

Hi Paul,

Actually, it seems like it could be a problem with Octane itself, not the plugin. Loading the OBJ into Standalone and increasing the subdivision to 1 makes it crash instantly.

In case you still want to test it, though, here's the OBJ, which should be all you need (hope I'm not violating any copyright laws!).

Thanks!

[link removed]

Posted: **Tue Jun 10, 2014 11:39 pm**

Actually, it seems like it could be a problem with Octane itself, not the plugin. Loading the OBJ into Standalone and increasing the subdivision to 1 makes it crash instantly.

Thanks - I'll submit it to the Otoy devs for actioning.

Paul

Posted: **Wed Jun 11, 2014 7:42 am**

Actually, it seems like it could be a problem with Octane itself, not the plugin. Loading the OBJ into Standalone and increasing the subdivision to 1 makes it crash instantly.

The Otoy devs have investigated this issue and found the error to be happening inside the OpenSubiv library (which is external to the Octane library). So it will not be a simple fix, but they are looking into it in more detail.

So for figures which cause this error - use Poser Subdivision instead of Octane subdivision.

Paul

OTOY Forums

OctaneRender 2 for Poser (Windows/OSX) [STABLE]

OctaneRender 2 for Poser (Windows/OSX) [STABLE]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]

Re: OctaneRender® for Poser (Windows) 2.0 [TEST]