I have collectd some sky flat and globular cluster data from the night of 20160409 (one our first VIRUS nights). The basic data are summarized below:
bs exp comments --- --- ----------------------------------- 200 01 closed-dome flourescent 201 03 twilight sky flat 201 04 twilight sky flat 300 02 NGC5204 (glob cluster) 301 01 NGC5272 (bright glob cluster) Available IFUSLOT values: 074 075 084 085 094 095 103 104 105
Initially I realized that my fiber position determinations using twilight skyflats were sometimes flawed by missing or undetected fiberd (denoted as "m" below) or by a wavy pattern along the slit dimesion (denoted as W below). If the flat treatment was good (all 112 fibers found) I place a "-" in the table below. I summarize two flats (both in obs=201): exp=04 at 20000adu, exp=03 at 40000adu.
exp=03 exp=04
IFUSLOT LL LU RL RU LL LU RL RU
074 m - - - m - - -
075 - - - - - - - -
084 - W - - - W - -
085 - - - - - - - -
094 - - - - - - - -
095 W W - - W W - -
103 - - m - - - m -
104 - - - - - - - -
105 W W W W W W W W
Hence, the list of "well-behaved" IFUSLOT positions:
N5204 (300,02) N5272 (301,01)
--------------- ------------------
075 1 star 2 bright stars
085 very faint 1 star on edge (?)
094 2 stars (?) 2 stars (one edge)
104 2 or 3 stars 3 or 4 stars
To look at just the fiber assembly stuff I need to play with:
% cat fibs #!/bin/bash if [ -z $1 ] then printf "Usage: fibs 094 \n" exit fi ifuslot="$1" xcol="532" z1="1" z2="1000" vir_signal_list.sh $ifuslot $z1 $z2 > SIGS virus_signals_fits.sh SIGS SIGNALS_$ifuslot.fits % ls Signals.LL Signals.LU Signals.RL Signals.RU % pwd /home/sco/VIRUS_Sample/view/NGC5272_01/ifu_104/T1 To run this: % ../../fibs 104
Recall that in addition to making a FITS image for an IFUSLOT, I can use the python code: vir_signals_map.py -v SIGS 8 $z1 $z2
vir_signals_map.py -v SIGS 8 40 100I am finding that, although the fibers are clearly labeled, this is not as useful as the FITS image approach. To match up my IFUSLOT=104 field I needed to do some image flipping, an easy thing to do in ds9.
At this stage I see a need to add more detailed information to the SIGS file above. In addition to a fiber number, it would be valuable to have the coordinates (X,Y) of where the fiber signal was measure on the original image (i.e. on o_LL_104.fits). Also, as I am assembling this list, I could generate ds9 region files that can be used to overplot fiber positions in the spectral image AND the spatial image. These things should be fairly straight-forward to implement now. However, as I have a list of working IFUs I might as well review each of these in the NGC5272 and NGC5204 fields to see if I can find any other recognizable fields.
Hence, the list of "well-behaved" IFUSLOT positions:
N5204 (300,02)
---------------
075 YES (?), 1 star, invert Y only (ngc5204_075.png)
085 NO, big difference between CCD (L and R)
094 NO
104 YES, 4 stars, invert Y (NOTE-A) (ngc5204_104.png)
rot NGC5204 13:13:19.8390 +18:07:40.265 257.586
N5272 (301,01)
------------------
075 YES, must invert Y axis, 4 stars identified
085 YES, 1 star, maybe a string of others (invert Y) (ngc5272_085.png)
094 YES (?), had flip in BOTH X,Y (ngc5272_094.png)
104 NO, not even very bright star near center.
rot NGC5272 13:42:36.6216 +28:19:16.840 276.683
NOTE-A: There may still be an orientation error in
treating the RL,RU amplifiers.
There appear to be other IFUs in both the N5204 and N5272 fields
where a matched pattern (to DSS) is likely. However, a brighter
set of stars would be desirable. Pointed VIRUS observations
where the dense part of an open cluster is placed on each IFUSLOT
at a time would be extremely helpful.
I had some semi-useful results with the 4 "good" IFU sets discussed above. Can I use the other 5 and/or improve the results for the good 4? There are two things I can look at:
1) Have fibermapX fill in missing fibers (or at least report mean fiber spearations. 2) Have fibermapX write out the normalized Peak signal levels.