Raw data FITS header issues

mgalloy commented 3 years ago

There several issues with the raw data as of the 20210222 data:

[x] FILTER keyword for dark data is sometimes "0.000"; it should be either not present or the actual value of the filter wheel position
[x] OCCLTR-Y seems to have a bug in the format code for its output, i.e., it shows a value of "0 3f".
[x] MODWAIT is the number modulation cycles, so it should be an integer, not a floating point value.
[x] SAVEALL and T_COMPS are true/false so values should be just T or F (no quotation marks).
[x] The comment for T_COMPS is cutoff.
[ ] A reminder that OBSSWID needs to be calculated according to the comment.
[ ] TELESCOP is not set and there is no comment.
[ ] LOCATION is still set to "BOULDER" and there is no comment.

bberkeyU commented 3 years ago

Looking in my notes I cant find desired comment values for TELESCOP or LOCATION

bberkeyU commented 3 years ago

We now poll the filter wheel position when the observing code starts up this should set FILTER properly every time.

The observing code now supports BOOLS as a header type and SAVEALL and T_COMPS look to be correct.

T_COMP comment changed hopefully for the better.

OCCLTR-Y is now a number 0.00, not the correct number I need to do more debugging but it is at least getting the right type/format in the header.

MODWAIT was a test variable to help me track and debug things. I am no longer using it so it shouldn't be in the header anymore.

OBSSWID TODO.

mgalloy commented 3 years ago

Let me know the next time you generate some data.

bberkeyU commented 3 years ago

There are a couple of "darks" out there on twilight now. But right after I submitted the last comment I noticed the data is identically zero, still trying to chase down the issue with the camreas.

bberkeyU commented 3 years ago

Looks like the cropico was at issue (I dont think I will have time to hunt it down this morning), but by turning it off I was able to get good darks 20210225.194454.50.ucomp.l0.fts and 20210225.194454.50.ucomp.l0.fts should reflect the best headers as of today.

mgalloy commented 3 years ago

OK, I checked off several items for this issue.

mgalloy commented 3 years ago

Maybe put a "TODO" for the TELESCOP and LOCATION comments so we don't forget?

bberkeyU commented 3 years ago

In 20210225.195420.29.ucomp.l0.fts We should see the T_COMPS Comment: [TF] Lyot turning temperature compensation on

And TODO's for TELESCOP and LOCATION

mgalloy commented 3 years ago

INSTRUME=              'UCoMP' / Upgraded Coronal Multichannel Polarimeter
TELESCOP=     '20 cm One Shot' / NSO One Shot telescope

mgalloy commented 3 years ago

FILTER is set to an empty string for darks in current data, e.g., in 20210312.204118.51.ucomp.l0.fts:

FILTER  = ''                   / [nm] Prefilter identifier

It should be missing if a dark, e.g.:

FILTER  =                      / [nm] Prefilter identifier

bberkeyU commented 1 year ago

Mike what is the current status on this issue?

Do we still need to change FILTER to missing if we are taking darks?

mgalloy commented 1 year ago

I would prefer that, but the current pipeline figures it out that it is a dark.

bberkeyU commented 1 year ago

There is a mismatch between my local checkout of the HAO labview fits libraries and my locally installed version. If the ucomp computer us using the version that is on my laptop, we can resolve the filter issue by making a change to the FITS HEADER ADD string.vi; to check if we have an empty string after the "String Subset" command, and then only surround the input string with single quotes if it is not empty.

If we are running the current labview library in subversion more work will be needed to eliminate the bug.

NCAR / ucomp-pipeline

Raw data FITS header issues #12