# RASTART/END headers are incorrect for LATISS images

XMLWordPrintable

#### Details

• Type: Story
• Status: Done
• Resolution: Done
• Fix Version/s: None
• Component/s:
• Labels:
None
• Story Points:
1
• Team:
Ops Middleware
• Urgent?:
No

#### Description

The values in the SAL messages from the pointing component that are used for the RASTART and RAEND headers are incorrect.

Prior to about 2021-02-11T1845Z, they were in decimal hours and were offset by -37 seconds from the correct value (due to TAI/UTC).

After about 2021-02-11T1845Z, they are in decimal degrees but are still offset by -37 seconds (of time, not of arc).

It is expected that there will be a further change in the future to remove the offset.

Please adjust obs_lsst to produce the correct values for this metadata.

Tiago Ribeiro Please correct if I got anything wrong.

#### Activity

Hide
Tim Jenness added a comment -

The translator currently is set up to

1. Ignore RASTART before May 2020 (it was reporting a value derived from the mount coordinates so was unusable)
2. Assumes that all Rubin instruments use degrees for RASTART

Current data therefore needs the RASTART header corrected by the time offset. If the time is out by 37 seconds I think you are asking for a correction of 15*37*cos(dec) arcseconds to be made. We should be able to see that in the difference between the RASTART and RA headers. I take it that we have some data somewhere at NCSA that has this style of RASTART onsky header?

Show
Tim Jenness added a comment - The translator currently is set up to Ignore RASTART before May 2020 (it was reporting a value derived from the mount coordinates so was unusable) Assumes that all Rubin instruments use degrees for RASTART Current data therefore needs the RASTART header corrected by the time offset. If the time is out by 37 seconds I think you are asking for a correction of 15*37*cos(dec) arcseconds to be made. We should be able to see that in the difference between the RASTART and RA headers. I take it that we have some data somewhere at NCSA that has this style of RASTART onsky header?
Hide
Tim Jenness added a comment -

I've had a quick look at a couple of files and after I correct for the hours->degrees problem and 37 seconds I have:

 $translate_header.py -p lsst.obs.lsst.translators -m fixed tests/headers/latiss-AT_O_20210212_000006.yaml | grep RA Analyzing tests/headers/latiss-AT_O_20210212_000006.yaml... RA: 260.181360944445 RASTART: 260.1785517385836 RAEND: 260.17851507256665   (lsst-scipipe-0.3.1) milo:obs_lsst_test (tickets/DM-28783)$ translate_header.py -p lsst.obs.lsst.translators -m fixed tests/headers/latiss-AT_O_20210210_000011.yaml | grep RA Analyzing tests/headers/latiss-AT_O_20210210_000011.yaml... RA: 355.38776875 RASTART: 355.41750341182313 RAEND: 355.3854958801916 

It looks like there was poor tracking on the observation on the 10th but it ended up in the right place. I think the observation are showing a 8 to 10 arcsecond error now.

Show
Tim Jenness added a comment - I've had a quick look at a couple of files and after I correct for the hours->degrees problem and 37 seconds I have: $translate_header.py -p lsst.obs.lsst.translators -m fixed tests/headers/latiss-AT_O_20210212_000006.yaml | grep RA Analyzing tests/headers/latiss-AT_O_20210212_000006.yaml... RA: 260.181360944445 RASTART: 260.1785517385836 RAEND: 260.17851507256665 (lsst-scipipe-0.3.1) milo:obs_lsst_test (tickets/DM-28783)$ translate_header.py -p lsst.obs.lsst.translators -m fixed tests/headers/latiss-AT_O_20210210_000011.yaml | grep RA Analyzing tests/headers/latiss-AT_O_20210210_000011.yaml... RA: 355.38776875 RASTART: 355.41750341182313 RAEND: 355.3854958801916 It looks like there was poor tracking on the observation on the 10th but it ended up in the right place. I think the observation are showing a 8 to 10 arcsecond error now.
Hide
Tim Jenness added a comment -

Tiago Ribeiro would you mind taking a quick look?

I wasn't sure about applying cos(dec) when correcting for the 37 seconds.

Show
Tim Jenness added a comment - Tiago Ribeiro would you mind taking a quick look? I wasn't sure about applying cos(dec) when correcting for the 37 seconds.
Hide
Tim Jenness added a comment -

Merlin Fisher-Levine thank you for agreeing to review.

Show
Tim Jenness added a comment - Merlin Fisher-Levine thank you for agreeing to review.
Hide
Robert Lupton added a comment -

Looks fine. I didn't check that the dates were correct

Show
Robert Lupton added a comment - Looks fine. I didn't check that the dates were correct

#### People

Assignee:
Tim Jenness
Reporter:
Kian-Tat Lim
Reviewers:
Robert Lupton
Watchers:
Kian-Tat Lim, Merlin Fisher-Levine, Robert Lupton, Tiago Ribeiro, Tim Jenness
Votes:
0 Vote for this issue
Watchers:
5 Start watching this issue

#### Dates

Created:
Updated:
Resolved:

#### Jenkins

No builds found.