On the night of Jul 25/26, 2023, under fair conditions, I acquired images of
in order to determine the color terms for our ASI camera and its filters. The results were ... so-so, but I've learned a lot from the attempt. I plan to observe M11 soon, which I hope will provide better measurements.
I chose as my targets the fields
Good sources for information on these fields are
The setup was
Notes from the night:
In general, I took sets of 30-second images for each passband (BVRI) in each field: twenty images each in V and R, and forty each in B and I, to which our camera is less sensitive. After cleaning them all in the usual manner, I tried extracting aperture photometry and applying ensemble photometry to derive mean magnitudes for each star. However, most of the stars had rather large uncertainties in individual 30-second images, and many of the fainter stars in each field were completely undetected.
So, I co-added all the images in each field to yield a single, deep image in B, V, R, I, then measured aperture photometry for that deep image. Some of the fainter stars did appear in the output of this method, thank goodness. I ran some tests using the individual 30-second images and found that the estimated uncertainty in each instrumental magnitude produced by the phot program was a pretty good match to the uncertainty estimated by the ensemble analysis; that meant I could trust the uncertainties in aperture magnitudes for the single, deep images in each passband.
The graphs below show the measurements for each field which served as input to the color-term analysis.
At first, I analyzed the measurements of each field individually. I iterated several times in each field, discarding a few stars in each which appeared to be outliers (almost always the very faintest stars in the set). Unfortunately, the number of "good" stars remaining in each field was relatively small; in addition, the PG1633 field had a blue star, but no really red ones; whereas the SA 110 (=Landolt 120) field had two red stars, but no blue ones.
Therefore, I combined the two sets of measurements in the following manner; I'll use the B-band in the explanation, but I carried out the operations for all four passbands: B, V, R, I.
(B - b1) = a + c*(b1 - v1)
The values of c in the equation above are the first-order color coefficients of for each passband. We can use them to convert instrumental magnitudes to the catalog system.
So, here's the combined dataset, before making these corrections ...
... and here are the results after applying the corrections.
I still consider the results preliminary, but here they are. Note that the only term which is really significant is the one for B-band.
# filter num_star color term stdev after
# ---------------------------------------------------------------------
b 14 0.1913 +/- 0.066 0.085
v 14 0.0154 +/- 0.033 0.046
r 14 -0.0573 +/- 0.144 0.128
i 14 0.0601 +/- 0.047 0.039
# ---------------------------------------------------------------------