obstime and date-obs: too smart for its own good?

[azangari]

I'm having an issue passing header parameters to mag files doing photometry.

Using either apphot or dataphot I enter datapars and set obstime to have the value "date-obs" (which a lot of cameras seem to have). Date-obs is of the formula YYYY-MM-DDTHH:MM:SS.SS, but only the HH:MM:SS.SS gets piped into the MAG file as otime.

The mag file says it has space for a 23 character string under otime. The date-obs string in its format above is exactly 23 characters. Coincidently the filter keyword will put that same string into ifilter (which also asked for a 23 character string) unchanged, so clearly there is space here for what I want.

The datapars help file says that phot doesn't use obstime, it's just there to be passed on for later calculations (along the filter and airmass). Is there a way to pass the full string on (aside from sticking the date as the filter and the filter as the date)?

I'd love it to have both the time AND date-- then I can write a program to convert the whole string to JD to compare multiple nights without an image name kluge?

Thanks!

[fitz]

The task is working as described, i.e. OBSTIME is used internally for informational purposes (e.g. you can use a ":obstime" colon command to print it out) but it is the OTIME value that matters in the output table). If the OTIME keyword points to a value containing the date field then the time is extracted when the field is parsed.

Other than changing the code there's no easy way to get the OBSTIME value printed out in the OTIME column. However, the OTIME can be in whatever units you like, so wouldn't setting 'otime' to point to a JD keyword field do what you want anyway? See the SETJD task if you need to compute a JD keyword. Hope this helps.