Flame 2021.1

Hello, i have problem with Mocha plugin which I would like to address. I use Mocha OFX as OFX node under Autodesk Flame 2021.1. When I use version 2021 8.0.3 some customization (like preferences or keyboard setting or color management) is not possible and when tried the mocha crashes back to “Launch mocha UI” selection. In flame console i see “ClientNode: --request-frame=75,1 returned no image” and “remote process has died”. When we use version 2022 9.0.1-49 it doesn’t work at all. The clip is not visible in Mocha Pro Plugin and the same problem repeats as in previous installed version (with similar output in Flame console as in previous version). I tried this setup on more Linux machines with Centos 7.6.1810 with no luck. All hardware is supported by the product (Quadro RTX 6000 with HP Z8 G4 configuration). Could you give me a hint which version is compatible or what can be done to use the Mocha? Thank you.

HI there, we are going to have to run some tests on this and let you know if we are able to repeat the issue. I will also check with @martinb and see if this is a known issue.

Can you go to help>view log inside of Mocha and see if you can save and send that to me as well? We may be able to use the logs to get to the bottom of this issue.

Let me know,
Mary

Older testing machines, the logs seems different then flame log.
Newer Mocha log:

Jan 11 19:39:20 (mochaui) [RepriseLicenseManager]: initialised with licence dir /usr/genarts/rlm
Jan 11 19:39:20 (mochaui) [RepriseLicenseManager]: successfully checked out ‘mpp-multihost-so’ version 20210915
Jan 11 19:39:20 (mochaui) [RepriseLicenseManager]: license expires 2022-04-28
Jan 11 19:39:20 (mochaui) Locale: default
Jan 11 19:39:21 (mochaui) Setting up 243 actions for usage logging
Jan 11 19:39:21 (mochaui) Setting up GL widget WITHOUT vertical sync
Jan 11 19:39:21 (mochaui) **************** GPU Information ****************
Jan 11 19:39:21 (mochaui) Platform:
Jan 11 19:39:21 (mochaui) Profile: FULL_PROFILE
Jan 11 19:39:21 (mochaui) Version: OpenCL 1.2 CUDA 11.2.162
Jan 11 19:39:21 (mochaui) Name: NVIDIA CUDA
Jan 11 19:39:21 (mochaui) Vendor: NVIDIA Corporation
Jan 11 19:39:21 (mochaui) Extensions: cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_fp64 cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_nv_copy_opts cl_khr_gl_event cl_nv_create_buffer cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_nv_kernel_attribute cl_khr_device_uuid
Jan 11 19:39:21 (mochaui)
Jan 11 19:39:21 (mochaui) Device:
Jan 11 19:39:21 (mochaui) Name: Quadro M6000 24GB
Jan 11 19:39:21 (mochaui) Vendor: NVIDIA Corporation
Jan 11 19:39:21 (mochaui) Version: OpenCL 1.2 CUDA
Jan 11 19:39:21 (mochaui) Driver version: 460.73.01
Jan 11 19:39:21 (mochaui) Global memory size: 23.9 GiB
Jan 11 19:39:21 (mochaui) Local memory size: 48.0 KiB
Jan 11 19:39:21 (mochaui) Number of computing units: 24
Jan 11 19:39:21 (mochaui) Maximum work group size: 1024
Jan 11 19:39:21 (mochaui) Maximum work item dimension: 3
Jan 11 19:39:21 (mochaui) Maximum work item sizes: (1024, 1024, 64)
Jan 11 19:39:21 (mochaui) Maximum clock frequency: 1114 MHz
Jan 11 19:39:21 (mochaui) Available for tracking: yes
Jan 11 19:39:21 (mochaui) Available for remove: yes
Jan 11 19:39:21 (mochaui) *************************************************
Jan 11 19:39:21 (mochaui) TaskManager: checking internal task status (0)
Jan 11 19:39:22 (mochaui) STL: Color space ‘matte_paint’ could not be found.

Older Mocha:
Jan 11 19:43:22 (mochaui) **************** GPU Information ****************
Jan 11 19:43:22 (mochaui) Platform:
Jan 11 19:43:22 (mochaui) Profile: FULL_PROFILE
Jan 11 19:43:22 (mochaui) Version: OpenCL 1.2 CUDA 11.0.228
Jan 11 19:43:22 (mochaui) Name: NVIDIA CUDA
Jan 11 19:43:22 (mochaui) Vendor: NVIDIA Corporation
Jan 11 19:43:22 (mochaui) Extensions: cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_fp64 cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_nv_copy_opts cl_khr_gl_event cl_nv_create_buffer cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_kernel_attribute_nv
Jan 11 19:43:22 (mochaui)
Jan 11 19:43:22 (mochaui) Device:
Jan 11 19:43:22 (mochaui) Name: Quadro M6000 24GB
Jan 11 19:43:22 (mochaui) Vendor: NVIDIA Corporation
Jan 11 19:43:22 (mochaui) Version: OpenCL 1.2 CUDA
Jan 11 19:43:22 (mochaui) Driver version: 450.102.04
Jan 11 19:43:22 (mochaui) Global memory size: 23.9 GiB
Jan 11 19:43:22 (mochaui) Local memory size: 48.0 KiB
Jan 11 19:43:22 (mochaui) Number of computing units: 24
Jan 11 19:43:22 (mochaui) Maximum work group size: 1024
Jan 11 19:43:22 (mochaui) Maximum work item dimension: 3
Jan 11 19:43:22 (mochaui) Maximum work item sizes: (1024, 1024, 64)
Jan 11 19:43:22 (mochaui) Maximum clock frequency: 1114 MHz
Jan 11 19:43:22 (mochaui) Available for tracking: yes
Jan 11 19:43:22 (mochaui) Available for remove: yes
Jan 11 19:43:22 (mochaui) *************************************************
Jan 11 19:43:22 (mochaui) TaskManager: checking internal task status (0)
Jan 11 19:43:22 (mochaui) Using at most 80 percent of system memory

Are you using an OCIO environment variable to set your OCIO config path?
Do you have any other OCIO configs on the systems?
Are you feeding an insert into the Mocha node or just the regular input?

Can I see your batch setup please?

Yes, there was set up OCIO environment variable for SapphireOFX, after removing it works with new version, testing with older version now. So the proper way how to use it is to setup OCIO just for SapphireOFX in s_config.text and remove it from environment variables (as described in Boris FX | Sapphire Plug-ins for OFX)?
Thank you

Hello, ok so it seems the problem was related with the old ocio.config where was not specified matte_paint role and mocha failed because of it. Now it’s fixed.
Thank you.

Fantastic, great to hear. I think we have a permanent fix for this coming in the next software update so you don’t have to do that next time.