From:                                         KENNETH H NORDSIECK

Sent:                                           Friday, August 05, 2016 5:05 PM

To:                                               'Janus Brink'

Cc:                                               Steve Crawford (crawford@saao.ac.za); Keith Browne (keith@salt.ac.za)

Subject:                                     RSS test #2

 

Hi All

 

Here is my first attempt to discuss how to look at the various RSS engineering tests during maintenance.  This material will be kept in my outbox, at

http://www.sal.wisc.edu/~khn/salt/Outgoing/Maintenance/

 

You will now see an ET2  subdirectory there, compiling stuff on Engineering Test #2, "Slitmask Focus".  It has results for the 20160725 on-telescope data for this test, and a comparison with the last time we did this, 20141212.  The analysis program, thrufoc_rsscartesian.py and a README describing how to use it are in

https://github.com/saltastro/SALTsandbox/,

subdirectory rss_thrufocus_analysis (or will be, after Steve has merged an update).

 

This test examines overall alignment and optical imaging performance. First, look at the .txt output of  thrufoc_rsscartesian in Maintenance/ET2. A detailed explanation of the program and its output are in the README in the sandbox.  The most interesting output for our purposes is the "Best focal plane" table, the "imaging" and "maskrim" rows (the "longslit" row is subject to outlier points, so ignore it for now).  Here is a table comparing the five tables we have (2 dates, 3 filters in one, 2 in the other):

 

20141212

20160725

micr

arcmin

arcmin

arcsec

micr

arcmin

arcmin

arcsec

4340

focus

tip

tilt

fwhm

spots

focus

tip

tilt

fwhm

spots

longslit

515.7

0.72

-9.93

0.43

39

imaging

487.2

-3.27

-8.12

0.47

329

maskrim

444.6

-5.69

-9.32

0.55

54

6645

6530

longslit

645.7

-1.13

-5.24

0.46

40

longslit

532.1

-4.50

-24.20

0.71

40

imaging

646.7

-2.31

-10.88

0.46

331

imaging

488.0

-3.98

-14.57

0.58

337

maskrim

617.4

-4.33

-11.6

0.54

55

maskrim

433.1

-9.36

-14.91

0.45

55

7840

7685

longslit

508.5

2.04

-11.82

0.38

39

longslit

873.1

0.04

-17.77

0.36

39

imaging

502.6

-0.3

-14.62

0.43

329

imaging

852.3

-3.83

-16.72

0.40

330

maskrim

468.3

-3.42

-15.43

0.53

54

maskrim

788.0

-11.38

-13.64

0.39

55

 

The “focus” numbers are different because the temperature and filters are different, not a concern.  The tip and tilt numbers describing how the focal plane is oriented relative to the detector are comparable, within 5 arcmin.  The fwhm are comparable.  So it looks good.  A more detailed comparison of the imaging can be had by looking at an optional output of the program (activated by specifying the fwhmfile= on the command line).  The fits files this produces are also in the Maintenance/ET2 subdirectory, along with a pdf showing a display of them in ds9.  Each file is a fits cube showing the fwhm (in unbinned pixels, which is 0.125 arcsec) of the  spots in the Cartesian  grid, stacked by focus position.  What I did for the pdf was to load them all up, step each of them through the cube to find the best focus, and plot them all on the same scale.  You see they are comparable.  The differences are due to the best focus in general being between the focus run positions, so the images are not actually quite best focus (try stepping though focus on each of them to see this).  At best focus, we expect a perfect pinhole would give a best fwhm of about 2 pixels (0.25 arcsec), which is the design performance of the optics.  These numbers are somewhat above that because the mask holes are not infinitely small.

 

Comments welcomed,

Ken