RE: Problems with Flatcomp and Linux generated files

[mgutzwiller@lanvision.com]

By the time the files get to FlatComp it shouldn't matter how the original
images were generated.  FlatComp looks only at the star lists generated by
the Star program.  With its default settings FlatComp is however very
selective on the files it considers good enough to create compensation
vectors with.  Specifically the default value for the Var Limit is 0.007.
Most files won't make this cut.  And I have had nights where none of the
files passed the cut.  You might try raising this limit to 0.01 or 0.015 to
get some files to pass.  A limit greater than 0.015 shouldn't be used though
since the data may not be reliable enough to calculate the compensation
vector.


Mike G.


		-----Original Message-----
		From:	Nicholas Beser [mailto:beser@aplcomm.jhuapl.edu]
		Sent:	Tuesday, March 23, 1999 1:07 PM
		To:	tass@wwa.com
		Subject:	Problems with Flatcomp and Linux generated
files

		I was writing a tutorial on TASS data analysis for our web
site, and as
		an example I pushed a three sensor set through the five
programs (Dark,
		Flat, Star, Starpost and Flatcomp). Everything seemed to go
OK for Dark,
		Flat and Star. Starpost seemed to reject more than I though
it should
		(based on the fact that it seemed to accept data from a
noisy sensor and
		reject data from one with better characteristics). Flatcomp
however did
		not accept any of the data. There seems to be something
missing in the
		Linux collected data sets that Flatcomp normally sees. There
were many
		CCD0 files and some CCD1 files in the directory. I expected
it to
		process the CCD0 data.

		Has anyone else tried this? I can send a CD-ROM to Mike
Gutzwiller (I am
		assuming that he wrote the software) if he would like to try
it with my
		data.

		Nick Beser
		JHU/APL