[ ESO ]
NACO Quality Control:
LAMP FLAT frames

CAL | HC | refs | QC
Trending & QC1
Data Management
QC links:
gain | FPN | lampflux
QC1 database (advanced users): browse | plot
   Click on CURRENT to see the current trending (Health Check).
   Click on HISTORY to see the historical evolution of the trending.

The LAMP FLAT frames are obtained within the NACO daytime calibrations every night the science data are taken. Each LAMP FLAT consists of 6 raw frames (3 pairs lampON/lampOFF) obtained with the instrumental setup matching the one used for science observations. Only the short wavelength filters are supported (J, H K regime). For the long wavelength (L-band) the "flats on sky" are taken.

The LAMP FLATs are supplied for both imaging and spectroscopy but only the imaging lamp flats are processed by the NACO pipeline. Normalized master lamp flats contain the pixel-to-pixel gain variations, the gain map, or fixed pattern noise as an instrumental signature which is removed from the science observation by dividing the dark subtracted science frame by the master lamp flat.

Example master flats are from 2016-11-26

Ks LAMP FLAT frame - H, S13, Double_Read This is an example of the master calibration LAMP FLAT frame taken with filter H and camera S13. The Double_RdRstRd read-out and HighDensity detector modes were used. It is a pipeline product of the "naco_img_lampflat" recipe. The bright spot in the upper left quadrant of the frame is a feature characteristic to the camera S13. H LAMP FLAT frame - Ks, S27 This is a master calibration LAMP FLAT frame taken with filter Ks, camera S27. The Fowler sampling read-out detector modes were used. It is a pipeline product of the "naco_img_lampflat" recipe.

top FPN

QC1 parameters

parameter QC1 database: table, name procedure
"fpn": fixed pattern noise naco_lampflat, fpn fixed pattern noise calculated by the pipeline recipe naco_img_lampflat and normalized to the number of counts.

The trending plot shows the fixed pattern noise value calculated for the lampflats by the naco_img_lampflat recipe. It is the largest noise component in the raw frames. This QC parameter is mostly independent from the average counts in the flats.

top lampflux and illumination gradient

QC1 parameters

parameter QC1 database: table, name procedure
"lamp_flux": the flux of the lamp normalized to DIT (ADU/s) naco_lampflat, lamp_flux median flux of the difference between the first lamp ON and the first lamp OFF raw frames of the lampflat, divided by DIT; calculated by the naco_img_lampflat recipe.
"flat_med_max": naco_lampflat, flat_med_max median of the highest flux values of the raw frames
"flat_med_min": naco_lampflat, flat_med_min median of the minimum flux values of the raw frames
xgradient, ygradient naco_lampflat, xgradient, ygradient The xgradient and ygradient parameters have been introduced 2015 to monitor the illumination gradient over the FOV. I might indicate a problem with the calibration unit.

These QC parameters are used to monitor the quality of the master lamp flat calibration product. The lower thresholds of the mean number of counts checks that the singal-to-noise ratio of the master flat is large enough to minimize the statistical noise. The trending plot is used to monitor the performance of the internal lamp. The data points correspond to the median flux of the difference between the first lamp-on and the first lamp-off raw frames, divided by DIT.

To measure the illumination gradient over the FOV, four pixel-windows have been defined on the master lamp flat product, each one 100x100 large, and with a distance of about 100 pixel from the edge of the chip with centers at

  • x=150,y=512 (left)
  • x=850,y=512 (right)
  • x=512,y=150 (bottom)
  • x=512,y=850 (top)

Based on these four windows one flux gradient in x and one in y is calculated. The flux gradients are normalized to the size of the FOV (Delta x = 1024 pixel). E.g. xgradient = -0.2 means, that the counts are 20% less on the right rim than on the left rim of the detector. The gradient monitors a symptom, not a cause. Note that the lamp flat FOV illumination gradient is different between detector read modes since of a contamination issue.