[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: "Flat Fields Are Not The Problem"
- To: firstname.lastname@example.org
- Subject: RE: "Flat Fields Are Not The Problem"
- From: Glenn Gombert <email@example.com>
- Date: Sun, 24 Aug 1997 14:23:22 -0400
- Old-Return-Path: <firstname.lastname@example.org>
- Resent-Date: Sun, 24 Aug 1997 14:18:48 -0400 (EDT)
- Resent-From: email@example.com
- Resent-Message-ID: <"iiZHn.A.cw.0rHA0"@kani.wwa.com>
- Resent-Sender: firstname.lastname@example.org
After "mulling-over" some of the things that Michael has mentioned
that he has found below several (more ) queestions came to mind.
(1). It seems to me that since the cameras are operated from a fixed
locatron that the VCO drift might be the single bigest variable in the
system. I have noticed that drift can be detected from say starting a run at
10:30-11:00 and seeing noticable trailing by 5:00 AM
The answer to this may be a simple parts subsitution (maybe Tom
could recommend some parts subsitution) that may make the VCO more stable by
a factor of 5-10. Parts of these type (crystal controlled oscillators) are
easy to obtain and may help to reduce some of these errors.
(2). Since a number of use have noticed the FWHM of the PSF vary
considerably over the course of an image Mike's suggestion of an estimate of
the error (and a correction applied to each frame). My be necessary to
compensate for these type of errors. (He's probably busy already mod'ing the
Star program to accomplish this).
(3). Some sort of a 2-D PSF fit (maybe using DAOPHOT) may be the
only other way to compensate for some of the errors that exits across the frame.
At 11:16 AM 8/22/97 -0400, you wrote:
>There has been a lot of discussion lately about getting proper flat
>fields for TASS cameras. I now think our photometric problems are NOT
>flat field related and will not be cured with better flats. I have come
>to this conclusion based on the following.
>To date both Tom and I have been using sky backgrounds to generate flat
>vectors for use in the star program. These flat vectors are typically
>slowly undulating and look just like what you would expect if a two
>dimension flat with dust doughnuts and some vignetting were squashed to
>one dimension. What you may not know is that these flats typically show
>a 5 percent total range, peak to trough. If these flats were total
>fiction, and I don't believe they are, they could contribute at most a
>0.015 mag SIGMA error. Since we have a lower error limit of 0.050 mag
>sigma the problem MUST lie elsewhere. Last night I believe I found the
>source of the problem.
>I was reprocessing several of my earlier images against a catalog of
>good Tycho stars. Good in this case means sigvt and sigbt < 0.050 and
>no other Tycho star within 10 pixels. This catalog should then give
>good astrometric and photometric solutions. For the images I was
>processing this resulted in approximately 20 good Tycho stars per image.
>I then compared the magnitudes of the Tycho stars and the instrumental
>magnitudes and found they differed by about 0.140 mag sigma, much worse
>than I expected. I then graphed the mag difference as a function of RA,
>result - a random scatter plot. I then graphed the difference as a
>function of Dec, voila - a straight line! The error was a very strong
>function of declination! A straight line fit showed a slope of 0.160
>mag per degree. That's right, a difference of about 0.450 mags from one
>edge to the other! This same slope was repeated in subsequent images on
>the same night.
>I then used the computed slope to correct the instrumental magnitudes
>and achieved a difference between the Tycho magnitudes and the corrected
>magnitudes of 0.060 mag sigma. Since the Tycho stars in the comparison
>were good to only 0.050 mag sigma this is a good match.
>What could cause this error slope? Could it be a hidden flat fielding
>problem? In this case, again no and here's why. If the problem were
>flat related then the slope seen above would be the same no matter what
>size aperture was used to measure the magnitude. In fact this is not
>true. The slope for both the PSF based aperture and the 5x5 circular
>aperture were both about 0.160. For larger apertures the slope slowly
>decreased until it was 0.095 for the 11x11 circular aperture. This
>indicates that the PSF was quite different from one side of the chip to
>the other. The larger apertures were able to partially compensate but
>What does all this mean for our future processing? It means that if we
>want to break the 0.050 mag sigma limit we must adjust the instrumental
>magnitudes by a low order fit against a catalog like the good Tycho star
>list I used above. We probably don't have to worry about getting
>non-sky flats since the fit above would adjust for any low order errors
>in the flat itself.
Glenn Gombert <email@example.com>