2My interests Cataclysmic Variables Exoplanet transits Oph Arcturus SpectroscopyMostly into scientific imagingAlmost finished with Masters degree at SAOPartner Alnitak AstrosystemsPhysics Teacher at Watershed School
12CCD calibration math Very simple equation! raw frame - dark frame Remove dark current noiseraw frame - dark frameFinal frame =Brightens weak pixels, dims strong pixels( flat frame - dark frame)NormalizedVery simple equation!
13Normalization Done automatically in your image processing software! (Flat frame-dark frame)pixel valuesAssumes that flat light source is even!100110Average value = 10010.91.1pixel valueaverage valueNormalized value =90100Done automatically in your image processing software!
14Applying the normalized flat to your image frame (Raw frame - dark frame) pixel values300302275305331300Calibrated frame÷By normalized flat10.91.1298305
15Importance of dark subtraction Assume 10 ADU of dark noise in the flat frameFlat frame-dark frame11010090Average: 100Raw flat frameSubtract darkAverage: 110110120Normalized values too low!10.911.0910.91.1100110Overcorrected images!
16Importance of staying in linear regime If non-linear, pixel values will read less than actual valueNormalized flat pixel value too smallFlatted image pixel value too large: Overcorrected images!linearADU valuesNon-linear# of photons arriving at detector
17Characterizing linearity Aim at 6-9th mag star near the zenithExpose series of images with increasing exposure lengthMeasure flux inside apertureDivide flux by exposure time to get flux/secWill be similar at each exposure length in the linear regime
18Figure 3. Detector linearity test Figure 3. Detector linearity test. The normalized flux rate is linear to 1% up to maximum pixel values of around 23 kADU.
19Noise considerationsMake master bias/dark (s/n improves as the square root of the # of frames combined)Dark OR bias correct flatsMillion photon flats106/avg ADU = # of frames = 40 frames!No matter what, flats add some noise to final calibrated image
21Acquisition methods All sky flats Light box Twilight flats Dome flat Electroluminescent panel
22What makes a good flat? Evenness of illumination ADU values at upper range of linear regime of CCD detectorLonger than 2 seconds to eliminate shutter effectMany dark subtracted sub framesRepeatable filter wheel positioning
23The rotation method for evaluating flats Expose / rotate 90˚/ exposeDark subtract and use second set as flats - “flatted flat”Look at histogramAnalyze standard deviation ()
24Basic statisticsPoisson distribution of ADU values centered on a mean valueWidth of distribution measured by standard deviation, 99.7% of all values lie within 3 of the mean3
25Statistics IIFor a given light source, range of values is constant regardless of mean value!1000030000100100/300000.3% uniformityHistogram of flat (mean 30k ADU)100/100001% uniformity3 x better!100Histogram of flat (mean 10k ADU)Standard deviation is a measure of evenness of illumination!
26How many ADU is enough?Maximum value of any pixel must be in the linear regime of the chip.Anti-blooming chips go non-linear somewhere mid-rangeNon-linear pixels in flat will result in incorrect normalizationFunny artifacts in flatted imagesGood StatisticsNon-linear pixels
27All sky flatsSum lots of images dithered to get enough ADU’s for good stats.Can be important for photometry or back illuminated chips because spectral response matches raw imagesStar artifacts difficult to remove completelyTough with wide field images/big non-linear stretches
28Light boxes Needs proper baffling and reflective illumination Careful attention to corner shadowsBulky and difficult to use robotically
29Twilight flats Racing against the clock Neutral point in sky is not fixedVirtually guaranteed to have gradients in wide field imagesPossible star artifactsCan you get all filters covered in one twilight?Quality is not repeatable!
30Twilight flat case study AprilAverage transparency (clear sky clock)No visual signs of cirrus12.5” RCOS with ST2000 (identical setup)Moon below horizon
31Twilight flatted flat at Galaxy Quest Standard deviation = 171 ADU
32Dome flats Painted section of dome illuminated by light source Difficult to eliminate gradientsRequires careful set up and testing
33Dome flatted flat at SSRO Standard deviation = 187 ADUData courtesy Jacob Gerritsen, SSRO
34Electroluminescent panels Proper design ensures excellent flatnessEasy to diffuseCompactNot all panels are broad spectrumVariation in manufacturingStability of power supplyAlnitak Astrosystems!
41Ha FlatsCourtesy Doug Baum (Flip-Flat owner) at Nightvision Astronomy
42DOs DON’Ts Use even illumination Master dark/bias subtract individual flat framesSigma combine lots of calibrated flat framesCheck your flat quality with the rotation methodOverexpose into non-linear regimeApply noise reduction or smoothingStretch histogram
43$100 off on XL’s today and tomorrow Flat -Man XLI hope I’ve convinced you of the quality and repeatability of electroluminescent panels. Ours are broad spectrum, fully adjustable, and very even!$100 off on XL’s today and tomorrow