[ ESO ]
FORS2 Quality Control:
Bias

CAL | HC | refs | QC
QUALITY CONTROL
    HOME
FORS2 QC
Trending & QC1
   bias
History
Pipeline
Data Packages
Data Management
QC links:
median level | read-out noise | fixed pattern | structure | pattern
 
QC PLOTS
  CURRENT HISTORY
median level (master)
read-out noise (raw)
read-out noise (master)
fixed pattern (raw)
structure (raw)
pattern (master)
median level (master, before 2002-04)
read-out noise (raw, before 2002-04)
read-out noise (master, before 2002-04)
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.


In April 2002 FORS received two new CCDs, which are still in operation. Parameters are measured and trended for each CCD and the 2 standard CCD modes:
- high gain, 100Kp/s, binning 2x2
- low gain, 200Kp/s, binning 2x2
There are also some measurements for unbinned data (binning 1x1) in the QC1 Database.

For the old CCD the trended read-out modes were
- high gain, 4port (ABCD) and 1port (A) read-out, binning 1x1
- low gain, 4port (ABCD) and 1port (A) read-out, binning 1x1

top Median Level

QC1 parameters

  • median level of master bias (QC1 database table fors2_bias, column median_master)
    This is the median level of all pixels in the master bias.
  • median level of first raw bias (QC1 database table fors2_bias, column median_raw)
    This is the median level of all pixels in the first raw bias.

History

Until 2008-03-31 the old pipeline was used, with which the medain levels were derived as follows:

  • median level of master bias (QC1 database table fors2_bias, column median_master)
    This is the median level of the master bias measured in a 1600x400 pixel window. The window is chosen to cover in both CCDs the regions that can be exposed. The short y dimension is due to the large vignetting of Chip 2.
  • median level of first raw bias (QC1 database table fors2_bias, column median_raw)
    This is the median level of the first raw bias measured in a 1600x400 pixel window. The window is chosen to cover in both CCDs the regions that can be exposed. The short y dimension is due to the large vignetting of Chip 2.


For the old CCD (until 2002-03) the median was calculated for the whole CCD.

Since mid-August 2014 the bias median level of C1 has been increasing. Finally, it crossed the threshold on 2014-10-05. The issue was investigated and fixed on 2014-10-09 - PPRS-058387.


top Read-Out Noise

The read-out noise is measured in the master (noise_master) and in the raw (RON_r) frames. The difference between values in raw files and master files controls the efficiency of the master creation process.

QC1 parameters

  • read-out noise of raw bias (QC1 database table fors2_bias, column RON_r )
    The difference between the first and second raw bias is computed. The standard deviation of the pixel values in the difference image, divided by √2, is taken as the read-out noise of the raw bias.
  • read-out noise of master bias (QC1 database table fors2_bias, column noise_master )
    The deviation of pixel values from the median level is determined, excluding deviations by more than 3·RON_exp to avoid deviations not due to the read-out noise. This value is then taken as the read-out noise of the master bias.


History

RON_exp depends on the number of frames and the averaging method. The pipeline used until 2008-03-31 used a simple average, whereas the new pipeline does a min-max rejection of the highest and lowest pixel. Therefore it uses 3 instead of 5 frames for the averaging and therefore the noise of the master bias is higher than with the old pipeline.

For the old CCD (until 2002-03) the rms scatter of the whole CCD was used as read-out noise for both raw (sigma_raw) and master (sigma_master) bias.

top Fixed Pattern

QC1 parameters

  • fixed pattern of raw bias (QC1 database table fors2_bias, column FP_r)
    The difference between the first raw bias and the second raw bias (in ADU) shifted by 10x10 pixels, is computed. This is the combination of fixed pattern and read-out noise (scaled by √2). The population standard deviation of this difference is computed and divided by √2. The read-out noise is then quadratically subtracted from the total noise and the result is taken as fixed pattern.
top Structure

QC1 parameters

  • structure of first raw bias (QC1 database table fors2_bias, column struct_r)
    The population standard deviation of the pixel values of the first input raw bias is computed. This is the combination of structure, fixed pattern, and read-out noise. The read-out noise and the fixed pattern contributions are then quadratically subtracted, leaving the large scale variations.
top Pattern

QC1 parameters

  • amplitude of possible pattern in master bias (QC1 database table fors2_bias, column pattern)
    The master bias is averaged along the y-axis and the rms of the resulting one--dimensional frame is determined. This allow to detect vertically striped patterns like the one shown below.


 
[ESO][Index][Search][Help][News]