CD16 vs CD15

[Andrew Bennett <andrew.bennett@ns.sympatico.ca>]

CD16 has the new tweaked value of VVH to give improved
saturation performance. CD15 provides a comparison.

Saturation values.
Saturation values were extracted manually from 2 images
each in V and I on CD16 and from 1 image in V and I on CD15.
Quadratic fits were performed (X1, X2, X1^2, X1X1, X2^2).
The only significantly non-zero coefficients were those of
X1 and X1^2 in both V and I. Contrary to my previous investigation 
of CD3/CD5, there was no significant variation with X2.
The significant coefficients for CD16 were around 6 times 
smaller than for CD16 i.e. around +-3% instead of +-20%
So Tom's tweaking is a great help.

Star magnitudes
Star amplitudes were extracted using my PSF-fitting method but 
without Dark/Flat correction (still not fully automated - anyway
one does not really want Flat corrections here!)
A set of stars was extracted which:
1) Matched Tycho2
2) Fitted the mean PSF. This gets rid of close pairs and other
nastiness such as cosmic rays.
A quadratic fit was performed as above on the magnitude differences.
No coefficients of the CD16 V-data were significantly 
different from zero.
The X2 coefficient of the CD15 V-data came out "just significant"
at 5%.
CD15 V-data with Dark/Flat correction gave an insignificant X2
coefficient but X2^2 barely "significant" at 5%.
I wouldn't put much weight on this "significance": 2 out of 15
coefficients tested came out with values one would expect 1 time
in 20. 
Anyway, the new CD16 data look good.

There is still quite a lot of scatter. The trend of +-20% seen in
the CD15 saturation values would have been easily seen if it
happened in the sensitivity too; it isn't there. (I think that
disagrees with my CD3/CD5 results too.) But +-3% could
hide in the scatter of around +-2% (+-0.02 mag) in the X1 and X2
trends and the quadratic terms are even less well defined . So 
somebody should analyze a lot more data to get better
statistics ...

Somebody should do some I-images too. I thought I had but
when I tried to read them in to the spreadsheet, they were
not there. But we did manage to deliver a record crop of
grapes this year ...

Andrew Bennett, Avondale Vineyard, Nova Scotia, Canada.