Flame crashes due to "CER_Core"

Flame Flame Questions
,
1

Hi!
I’m having some problems with crashes and can’t really find any clues so I thought I’d post here.

I’m using flame 2025.1 on mac Ventura 13.5.2 (22G91)

When rendering/hard comitting specific sequences in my project flame crashes at the same “place” in the progress bar. It seems to be at the very end of the render so only a few pixels left on the progress bar before completion.
The only noteworthy info I get in the terminal is the following:

Assertion failed: ((startElementIdx + nbElements - 1) < _elementList->getNb()), function testFlagInRange, file media.C, line 3243.
Application: flame 2025.1 x86_64
Pid: 5588
Error: abnormal termination, signal = 6
SIGABRT - Abort
Unknown user initiated signal. si_code: 0
libc++abi: terminating due to uncaught exception of type CER_Core::CER_Exception: Access to file denied: /Users/REDACTED/Library/Application Support/Autodesk/CER
[VREF]: checks canceled (200 queued)
[VREF]: shutdown completed
Stopping managed threads.

I can’t find the folder “Application Support/Autodesk/CER” on my system so I don’t know what’s going on. The setup diagnostics gives all disk permissions a checkmark.

I found a workaround that kind of fixes the issue temporarily. I just cut the footage that caused the problem ( I think it was a PNG graphic) right before it should end and rendered the first part. After that I removed the “cut off” end and extended the first part again and it would manage to complete the render without crashing.

Any ideas on how to fix this permanently?

Cheers!

2

It sounds like there is a frame on a bad portion of the drives. I suspect somewhere in the original source there is a checkerboard frame. Not sure of the details of the fix (was the png virtual or stretch and rendered to time???) or what sort of drive the source was coming from, but it obviously bypassed the bad spot.

1 Like
3

Hi!
Thanks for taking the time and sorry for the late reply, It’s been a busy week.

There are no checkerboard frames as far as I can see, the exports looked correct in the end.

“was the png virtual or stretch and rendered to time???”
Not sure what you mean but the png’s are one “still frame” that I cached to framestore and “dragged out” to correct length on the timeline.

Atm I suspect that a Segment Connection on the png graphics might be the culprit (I could hard commit once I removed all instances and re-added them without connecting) but I ended up making a support ticket so I’ll report back with any findings - in case it will help anyone in the future.

Thanks again for your help!