b113 Eos hang on Zoom to All Magic Sheet

Here's the logs on Google Drive for now: drive.google.com/.../1b-G4TXbm0lsRpAJzpprb3M9GwjQ2WD9H

Thanks for getting us these.  Unhelpfully, "Zoom to All" on a Magic Sheet isn't a logged command.  Do you remember when it happened relative to:

• changing the icon on Groups 27-31
• labeling Group 29, 30, and 31
• Bringing up Group 31 on the level wheel
• Shift+Updating 
• Disconnecting the Mac

My guess is between bringing up Group 31 and Shift+Updating....?  And did you perform Shift+Update on the Mac or the Ion?  

Thanks for getting us these.  Unhelpfully, "Zoom to All" on a Magic Sheet isn't a logged command.  Do you remember when it happened relative to:

• changing the icon on Groups 27-31
• labeling Group 29, 30, and 31
• Bringing up Group 31 on the level wheel
• Shift+Updating 
• Disconnecting the Mac

My guess is between bringing up Group 31 and Shift+Updating....?  And did you perform Shift+Update on the Mac or the Ion?  

Exactly, between bringing up the group and the save. The save was initiated as soon as I realized it was going to potentially crash. The save was initiated from the console. Also, if it helps, the zoom to all was initiated via 2 finger double tap on a touchscreen.

So on investigation, it seems like the Ion struggled with resizing the Magic Sheet as it was also handling the Augment3d model, which happens to be close to our recommended limit of 200,000 triangles.  To resize the MS, the software momentarily used a lot of memory to re-render SVGs.  It needed to swap memory with the harddrive, and that swap is what caused the hang.

I would suggest that you consider un-ticking the "High Quality" box on Magic Sheets to make them less memory-intensive to render (this is a new feature in v3.1.1)

You could consider reducing the size of your Augment3d model by replacing the drapes with flat objects. 

Ok, thanks! Will do. Does hiding objects in an A3D model improve performance or do they need to be deleted? Would the A3D model have less of an impact if it is kept the same but run on a capable tether instead of localy on the Ion?

In my latest show I have 400K triangles, I'm running off line. I found splitting the screen into four significantly reduces the GPU load. Also reducing the number of light sources on helps. Hiding objects doesn't appear to help. If nothing has changed in the cue it still renders a new frame.