Details
-
Type:
Story
-
Status: Done
-
Resolution: Done
-
Fix Version/s: None
-
Component/s: obs_lsst
-
Labels:None
-
Team:DM Science
-
Urgent?:No
Description
The camGeom changes to handle guider and wavefront chips were done blind, without any real camera data from the corner rafts. A number of problems showed up when we started getting BOT data:
- SG0 and SG1 were swapped
- C00 is flipped top-to-bottom
- The WF chips think that they have 16, not 8, channels
- The rotations are incorrect
Please fix these.
Attachments
Activity
Taking this back as it turns out that there were problems at SLAC that lead to an apparent switch of SG0 and SG1. I'm hoping to be able to revert to the original geometry (the amp problems were still there, I think, so this ticket is probably not moot)
There was also LCA-13381 so that the WF chips are now known to be the top, not the bottom half of regular ITL devices. Some of the changes in this ticket were due to this, in particular the rotations of the SW0/1 devices.
I've reverted the SG0/1 changes and created a new PR. I have confirmed visually using pinhole images that the geometry is correct.
One small comment in the PR
Robert Lupton This ticket appears to be complete and merged. If so, can you close it please.
Tested on real BOT data of pinhole data.
I think I've fixed up the geometry, certainly it does the right thing with the data in /scratch/jchiang/BOT_data/pinhole/repo_patched on lsst-dev. Note: this data was patched due to bugs in the camera, and this branch should not be merged until tested on a new dataset that should be taken this week.
I had to make significant changes to make this work, so could you check that I haven't broken things that you did?