Sadly, the default values for the read-noise and gain are not always right in the headers. In particular the Kitt Peak data obtained Sept-Oct 2000 had much higher than usual readnoise. We can either fix these in the headers now, or we can simply hardwire average values into flatcombine in section II. 2. How do we measure it? Here's a relatively painless way:
!cp dflat014.fits gainA.fits !cp dflat015.fits gainB.fits !cp zero107.fits zeroA.fits !cp zero108.fits zeroB.fits ccdproc gain*.fits,zeroA.fits,zeroB.fits bpmasks="" ccdtype="" xtalkcor+ overscan+ trim+ imcopy gainA.fits[1] gainA1.fits imcopy gainA.fits[2] gainA2.fits imcopy gainA.fits[3] gainA3.fits imcopy gainA.fits[4] gainA4.fits imcopy gainA.fits[5] gainA5.fits imcopy gainA.fits[6] gainA6.fits imcopy gainA.fits[7] gainA7.fits imcopy gainA.fits[8] gainA8.fits imcopy gainB.fits[1] gainB1.fits imcopy gainB.fits[2] gainB2.fits imcopy gainB.fits[3] gainB3.fits imcopy gainB.fits[4] gainB4.fits imcopy gainB.fits[5] gainB5.fits imcopy gainB.fits[6] gainB6.fits imcopy gainB.fits[7] gainB7.fits imcopy gainB.fits[8] gainB8.fits imcopy zeroA.fits[1] zeroA1.fits imcopy zeroA.fits[2] zeroA2.fits imcopy zeroA.fits[3] zeroA3.fits imcopy zeroA.fits[4] zeroA4.fits imcopy zeroA.fits[5] zeroA5.fits imcopy zeroA.fits[6] zeroA6.fits imcopy zeroA.fits[7] zeroA7.fits imcopy zeroA.fits[8] zeroA8.fits imcopy zeroB.fits[1] zeroB1.fits imcopy zeroB.fits[2] zeroB2.fits imcopy zeroB.fits[3] zeroB3.fits imcopy zeroB.fits[4] zeroB4.fits imcopy zeroB.fits[5] zeroB5.fits imcopy zeroB.fits[6] zeroB6.fits imcopy zeroB.fits[7] zeroB7.fits imcopy zeroB.fits[8] zeroB8.fits (whew!) epar findgain PACKAGE = nproto TASK = findgain flat1 = First flat frame flat2 = Second flat frame bias1 = First bias frame bias2 = Second bias frame (section= [1000:1200,1000:1200]) Selected image section (center = mode) Central statistical measure (binwidt= 0.1) Bin width of histogram in sigma (verbose= yes) Verbose output? (list = ) (mode = ql) then: findgain gainA1.fits gainB1.fits zeroA1.fits zeroB1.fits findgain gainA2.fits gainB2.fits zeroA2.fits zeroB2.fits findgain gainA3.fits gainB3.fits zeroA3.fits zeroB3.fits findgain gainA4.fits gainB4.fits zeroA4.fits zeroB4.fits findgain gainA5.fits gainB5.fits zeroA5.fits zeroB5.fits findgain gainA6.fits gainB6.fits zeroA6.fits zeroB6.fits findgain gainA7.fits gainB7.fits zeroA7.fits zeroB7.fits findgain gainA8.fits gainB8.fits zeroA8.fits zeroB8.fitsFor the record, this is what we've measured:
KPNO ---Sept 2002
hedit *.fits[1] RDNOISE 6.5 add+ ver- show+ hedit *.fits[2] RDNOISE 4.6 add+ ver- show+ hedit *.fits[3] RDNOISE 6.0 add+ ver- show+ hedit *.fits[4] RDNOISE 6.6 add+ ver- show+ hedit *.fits[5] RDNOISE 10. add+ ver- show+ hedit *.fits[6] RDNOISE 7.3 add+ ver- show+ hedit *.fits[7] RDNOISE 7.3 add+ ver- show+ hedit *.fits[8] RDNOISE 6.7 add+ ver- show+KPNO Aug---Sept 2000
hedit *.fits[1] RDNOISE 14 add+ ver- show+ hedit *.fits[2] RDNOISE 15 add+ ver- show+ hedit *.fits[3] RDNOISE 20 add+ ver- show+` hedit *.fits[4] RDNOISE 8 add+ ver- show+ hedit *.fits[5] RDNOISE 12 add+ ver- show+ hedit *.fits[6] RDNOISE 8 add+ ver- show+ hedit *.fits[7] RDNOISE 20 add+ ver- show+ hedit *.fits[8] RDNOISE 23 add+ ver- show+ KPNO >Jan 2001 hedit *.fits[1] RDNOISE 7.5 add+ ver- show+ hedit *.fits[2] RDNOISE 6.7 add+ ver- show+ hedit *.fits[3] RDNOISE 5.2 add+ ver- show+ hedit *.fits[4] RDNOISE 5.4 add+ ver- show+ hedit *.fits[5] RDNOISE 9.5 add+ ver- show+ hedit *.fits[6] RDNOISE 5.7 add+ ver- show+ hedit *.fits[7] RDNOISE 6.1 add+ ver- show+ hedit *.fits[8] RDNOISE 5.2 add+ ver- show+ The gain values have been measured as follows: hedit *.fits[1] GAIN 3.2 add+ ver- show+ hedit *.fits[2] GAIN 2.5 add+ ver- show+ hedit *.fits[3] GAIN 3.0 add+ ver- show+` hedit *.fits[4] GAIN 2.5 add+ ver- show+ hedit *.fits[5] GAIN 2.3 add+ ver- show+ hedit *.fits[6] GAIN 2.7 add+ ver- show+ hedit *.fits[7] GAIN 3.2 add+ ver- show+ hedit *.fits[8] GAIN 2.8 add+ ver- show+
Stay tuned...