Topic 8: Calibrating your system Arne Henden Director, AAVSO
Steps Inspect for cosmetic quality Dark current measurements Gain and readnoise Linearity Scattered light Residual Bulk Image Best to do in warmth of house!
Cosmetic quality Don’t trust spec sheet, look for yourself Bias/dark will tell you about low-level traps/blocked columns Flats will tell you about weak column, point defects
ND5 bias
ND5 flat
Typical Flatfield APASS U16m ASA N8 astrograph 25% vignetting Corners outside corrected image circle Cosmetically beautiful
Dark Current Primarily to look for particularly hot pixels You CAN create a bad pixel map, interpolate across them for sky backgrounds Some hot pixels can be particularly bad (LED defect) Often can improve with lower temp
APASS-South LED camera defect Grew over several nights Required replacement of sensor (under warranty) Kodak took 4 months to deliver sensor to Apogee
Gain and Readnoise 1 Use Janesick’s method Take two light-box flats Form mean_flat1, mean_flat2 (average of central pixels) Take two bias frames Form mean_bias1, mean_bias2 Form flatdif = flat1 – flat2, sigma_flatdif Form biasdif = bias1 – bias2, sigma_biasdif
Gain and Readnoise 2 Gain = [(mean_flat1 + mean_flat2) – (mean_bias1 + mean_bias2)] / [sigma_flatdif**2 – sigma_biasdif**2] Readnoise = gain * sigma_biasdif/sqrt(2) Gain in e-/adu, readnoise e- Iraf obsutil/findgain Example: sonoita
Linearity and full well Can be done with lightbox, or with LED blinker Find exposure that does not saturate sensor; perhaps 30sec Take ramp of exposures from 0sec to 30sec Plot mean of central region
SRO STL-1001e linear to 65K ADU (actually doesn’t sample full well)
APASS KAF16803 linear up to 65K ADU; only samples approx ½ full well
SXV-H9
Nofs tek1k
Scattered Light Need to do on-telescope First, look through telescope WITHOUT camera – can you see sky, reflections? Next, try examining what the camera sees
Pinhole Camera Read Grundahl & Sorensen A&A Sup Ser 116, 367, 1996 Edmund pinholes 20-50micron Smaller pinholes = better resolution Greater distance twixt pinhole and sensor = bigger magnification
Pinhole images
Pre/post baffling to remove scattered light
Pinhole arrangement
Another pinhole image
temp
Raster scans Photometric (clear) night Take 27 images: one with bright star centered, 25 in a 5x5 pattern, last one with bright star centered Compare photometry at each position wrt. Center Example: Sonoita
Raster scan with problems
Raster scan with no flatfielding
Raster scan after problem fixed
Residual Bulk Image Common with front-illuminated sensors Particularly bad with KAF Worse for red light (watch out for autofocus programs!) Examples from R. D. Crisp
M19 300sec Z
Field after M19
temp
Photon mean free path
temp
RBI retention
Traps last a long time
RBI elimination
RBI flush results
Shutter errors Two types – non-photometric shutters (like blade), and exposure timing errors Photometric shutters usually remove shutter vignetting (SBIG is an example)
Linear/curtain photometric shutter
ST-402 shutter using the filter as shutter
SBIG 8300 shutter
Blade shutters
Shutter error elimination Take long-exposure flat Ratio with short-exposure flat to create shutter flat After regular flat-fielding, divide by shutter flat to further correct illumination