_h1. Bad pixel mask

We have studied the bad columns observed in the SN reduced images. See the talk given at the SV meeting on March 4th in the link below. The bad columns can be traced back to hot pixels in the bias images. The column containing the bad pixel shows the following behaviour:
  • the rows preceding the bad pixel during readout shows a 0.1% value of the bad pixel every 60th row. (Juan Estrada suspects this is caused by the serial register during readout)
  • the rows succeeding the bad pixel during readout shows a constant 0.1% value of the bad pixel value.

Two examples are shown for CCD 1 and 45 taken on Jan 22nd 2013. CCD 1 has the readout going from right to left, while CCD 45 has it going from left to right. The upper panels show the bias values while the lower panels are normalised to 1 by dividing the pixel values by the hot pixel value.

See also report by Sein and John M. at the DESDM meeting on Feb 27th 2013: