search

legacysurvey / legacypipe

Image reduction pipeline for the DESI Legacy Imaging Surveys, using the Tractor framework
BSD 3-Clause "New" or "Revised" License
37 stars 22 forks source link

Odd zero-points for MzLS night of 2016-05-17 #338

Closed djschlegel closed 4 years ago

djschlegel commented 5 years ago

There's one night of MzLS data with zero-point values (ZPT and CCDZPT) that are unusually large numbers, implying those data are deeper.

There's no indication in the observing logs that that one night in the middle of a run was different than any other night, so this must be something different in the reductions/gains/flats? Example file mosaicz/MZLS_CP/CP20160517v2/k4m_160518_052742_ooi_zd_v2.fits.fz

The attached figure shows the distribution of ZPT values for MzLS, with this one night in red.

The latest from Frank on this issue from Mar 29, 2019: "Each night is calibrated with the flats for that night. I checked the flats for the night and they were properly normalized to unity. I also checked the zeropoint vs sky level correction and it seemed fine. To do more I would have to take more drastic steps of removing calibrations and trending for several preceding nights and start from the earlier date."

This issue is moved over from the deprecated legacyzpts product https://github.com/legacysurvey/legacyzpts/issues/51

mosaic-zpts
moustakas commented 4 years ago

@djschlegel can you check on this otherwise punt back?

djschlegel commented 4 years ago

This is now being tracked in issue #490, and close to being closed. The problem was due to the flat-fields for that night not being very good.