Camera animation state isn't saving

ParaView Support
1

I’m finding that the keyframes that I’m inserting using my macro method (Animation KeyTime integer error) don’t survive a save, close and reload of the pvsm file. Any ideas about how to make these animation keyframes stick?

Here’s a link to a screen cap with the issue shown.

2

HI @Devin_Richard_Bayly

So I see two issues there,

  1. Python state file do not save keyframes

This is definitely a bug, you can work around it by using the “Python Trace” feature that correctly save it. Also please open an issue on our bug tracker: https://gitlab.kitware.com/paraview/paraview/-/issues

  1. XML State file incorrectly load keyframes

While I see the issue in your video, I’m not able to reproduce it, here is a state file containing keyframes that loads correctly in 5.11-RC1:
state.pvsm (198.1 KB)

Finally I want to mention that your macro makes so much sense ! Being able to do that in proper ParaView would be a net benefice ! You could open an issue about this if you want.

@nicolas.vuaille you definitely wants to check this one out :slight_smile:

3

Hi @mwestphal, thanks for all your support these last two weeks!

That’s too bad the keyframes aren’t stored in state. I’ll submit an issue on the bug tracker.

In general do python traces pick up on things that happen as a result of script/macro executions? When I use my little macro during a trace I see a blank space between the camera updates, and animation time change lines.

While I see the issue in your video, I’m not able to reproduce it, here is a state file containing keyframes that loads correctly in 5.11-RC1:

and you were using my little macro to try to reproduce this? I find that the pvsm files retain their keyframes if I set them manually using the widget.

Finally I want to mention that your macro makes so much sense ! Being able to do that in proper ParaView would be a net benefice ! You could open an issue about this if you want.

Oh thank you!! I agree that this would be a really important feature since many 3d programs do their keyframing this way. Sure I’ll open a feature issue about this too.

4

I did not use the macros indeed as I did not expect it to have an effect on the resulting state.

I see some way forward:

  1. Compare the manual.pvsm and macro.pvsm to see the diff
  2. Comparing your macro with what the python trace would generate when generating manually

In any case that does sound like a bug.

5

oh, good idea! I forgot the .pvsm wasn’t binary so I can do some manual snooping. Thanks for the suggestion!