Page MenuHome
Create Task
Maniphest T88971

Cycles Rendering Animation Stops (similar as T85276 is for EEVEE)
Closed, ArchivedPublic
Actions

Description

System Information
Operating system: Windows-10-10.0.19041-SP0 64 Bits
Graphics card: NVIDIA GeForce RTX 2080 Ti/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 466.27

Blender Version
Broken: version: 2.93.0, branch: master, commit date: 2021-06-02 11:21, hash: rB84da05a8b806
Worked: (newest version of Blender that worked as expected)

Short description of error
Rendering single images is OK, and restarting the rendering from the frame that was not rendered anymore is OK (until a few frames further the animation render stops again)
At that moment the CPU is at 100% or close, and both GPU's are idle
It seems to happen right after the save of the rendered frame and then the new frame rendering is not started. As we had the random crashes related to saves (T87958) there might be a relation

Exact steps for others to reproduce the error
The error seems random but frequent (frame 11, 23, 32, 72, 93, 142 ....) just by rendering the animation{F10162854}

blender_debug_output.txt6 MBDownload

System.txt546 BDownload

dracul_slaap1.blend112 MBDownload

[Based on the default startup or an attached .blend file (as simple as possible)]

Related Objects

Event Timeline

Luc Servranckx (GreyWarden) created this task.Jun 8 2021, 8:47 PM
Richard Antalik (ISS) changed the task status from Needs Triage to Needs Information from User.Jun 9 2021, 2:01 PM
Richard Antalik (ISS) added a subscriber: Richard Antalik (ISS).

Can you render provided file on CPU?
Des this happen when you click on File > Defaults > Load Factory Settings, open .blend file and render?

Luc Servranckx (GreyWarden) added a comment.Jun 9 2021, 11:33 PM

Thanks for the quick response. Something strange happened: without me changing any parameter that I am aware of the rendering suddenly changed its behaviour and I cannot reproduce the error anymore (note that before I had 18 freezes while rendering this animation. What I noticed is that before the render screen was not "cleared" when starting a new frame (the render blocks were overwriting when a new frame started). Now the render screen completely clears (checkerbox gray) before each frame and I have no problem in rendering the animation in one go.

I do not know how this is possible, but as I can no longer reproduce the error I suggest you park this bug for now. Could it have been an interference from my multicore Ryzen CPU (I noticed 2 blender processes in the task manager while only 1 blender was started) ?

Richard Antalik (ISS) closed this task as Archived.Jun 10 2021, 12:21 PM

Thank you for update, will close for now, and if you manage to reproduce this issue please open new report.

Luc Servranckx (GreyWarden) added a comment.Jun 13 2021, 8:09 PM

For your information:
I had again same problem with another file. When I switched off the post-processing "Compositing" (which was a straight input from render to output without other nodes) the problem dissapeared.
No need to reopen the problem but I hope that this solution is known and valid for others who have the same problem.

Richard Antalik (ISS) reopened this task as Needs Information from User.Jun 13 2021, 11:50 PM

I will re-check this, but such random bugs are hard to track down, especially when they depend on hardware. It would be best if you have case that can be reproducible by following set of exact steps.

Can you check if this also happens in latest alpha from https://builder.blender.org/download/daily/ ?

Richard Antalik (ISS) added a comment.Jul 6 2021, 4:31 AM

Sorry for delay, I have only now checked provided file and unfortunately it is too complex to investigate.

Please simplify file as much as possible, while it still crashes. This usually reveals one or two objects that cause problems. Also check if this issue could be caused by denoising or similar
render features.

Richard Antalik (ISS) closed this task as Archived.Aug 9 2021, 3:53 AM

No activity for more than a week. As per the tracker policy we assume the issue is gone and can be closed.

Thanks again for the report. If the problem persists please open a new report with the required information.