Tom Droege took a set of images with the Mark IV at his home in Batavia, varying the focus setting. Here, I analyze the results of this focus test for the V-band camera. The bottom line: the center of the field looks best around a setting of 465 or 470, which also yields the sharpest cores in the corners of the field (but big wings).
Tom used the following settings for focus on the V-band camera:
I looked at three different portions of the frames, each sub-section a box 200 rows by 200 columns.
Below, I show pictures of each subsection at the different focus settings. I've used the same contrast settings in each image: full white is <= -24350 DN, and full black is >= -23650 DN.
First, the center of the image:
|
|
|
|
|
|
|
|
|
|
|
To my eyes, a setting of 470 looks best, but 465 is very close to it. See the numbers in table below, after all the pictures.
Now, the "upper-left" corner, which I believe to be the North-West corner.
|
|
|
|
|
|
|
|
There's a trade-off here between the width of the core, and the size of the outer halo.
Now, the "upper-right" corner, which I believe to be the South-West corner.
|
|
|
|
|
|
|
|
In all cases, I examined raw images. I did not perform any dark subtraction or flatfielding, because the shape of the core of a star is largely unaffected by such niceties.
I used two tools to evaluate the PSF in each of these sub-sections.
I didn't use the second method to evaluate the central images; the FWHM is the tool of choice there, since there is no halo that I can see.
For each subsection, I picked three stars and measured the quantity for each one (yes, I used the same three stars for each focus setting). I list below the values for each star, and a rough average.
FWHM values for stars in "center"
Star 1 Star 2 Star 3 Average
------------------------------------------------------------------
445 FWHM 7.1 6.8 7.1 7.0
450 4.7 4.8 4.9 4.8
455 3.2 3.1 3.2 3.2
460 2.86 2.81 2.87 2.85
465 1.85 1.51 1.82 1.7
470 1.81 1.82 1.75 1.8
475 2.44 2.59 2.58 2.5
-------------------------------------------------------------------
FWHM and f5/f10 for stars in "upper-left"
Star 1 Star 2 Star 3 Average
------------------------------------------------------------------
445 FWHM 5.2 4.6 4.6 4.8
f5/f10 0.90 1.00 0.84
450 4.7 4.3 4.4 4.5
0.91 1.01 0.81
455 4.1 3.4 3.4 3.6
0.87 0.96 0.89
460 3.6 3.2 3.3 3.4
0.86 0.96 0.83
465 3.2 2.7 2.6 2.9
0.83 0.90 0.83
470 3.3 2.9 2.6 2.9
0.82 0.92 0.75
475 2.8 2.7 2.1 2.5
0.77 0.84
--------------------------------------------------------------------
FWHM and f5/f10 for stars in "upper-right"
Star 1 Star 2 Star 3 Average
------------------------------------------------------------------
445 FWHM 5.5 5.5 5.6 5.5
f5/f10 0.81 0.87 0.90
450 5.0 4.8 4.8 4.9
0.85 0.94 0.93
455 4.4 4.1 4.2 4.2
0.85 0.92 0.92
460 4.2 4.1 4.1 4.1
0.85 0.93 0.92
465 3.2 3.2 3.3 3.2
0.81 0.86 0.89
470 3.2 3.1 3.2 3.2
0.80 0.86 0.89
475 2.7 2.6 2.7 2.7
0.76 0.83 0.85
--------------------------------------------------------------------
It looks to me like position 465-470 yields the best stellar PSF at the center of the image, and also does a fair job in the corners. It appears that one can get a sharper core in the corners by shifting the focus position upwards to 475 or so, but at the expense of losing more light to an extended halo.
Andrew Bennett found a similar best focus position for stars near the center of the chip, but his metric indicated that one could improve the PSF in the corners significantly by shifting to much lower focus positions. Andrew, could you comment?
I'll try to work on the I-band images from Disk 6, then images from Disk 7. Tom, if you have preferences on priorities, let us know.