[ ESO ]
VIMOS Quality Control:
detector linearity and gain

CAL | HC | refs | QC
QUALITY CONTROL
    HOME
VIMOS QC
Trending & QC1
   detlin
Pipeline
Data Package
Data Management
QC links:
non-linearity | gain
 
QC PLOTS
  CURRENT HISTORY
gain blue green
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.


top Non-linearity

The (non-)linearity of the VIMOS detectors is assessed by a series of flat exposures with increasing exposure time. After bias subtraction, the function

ADU = A + B * EXPTIME + C * EXPTIME^2

is fitted to the average level of the flats. The coefficients A, B, and C are saved into an output file. They are not used further for data reduction since the VIMOS detectors do not show significant non-linearity.

For monitoring, the QC parameter effective non-linearity is used. It is the difference between the flux obtained at the reference level 40000 ADU from the polynomial fit and the flux predicted by the linear term only, normalized by the polynomial flux. The mean level of the flats with the highest exposure time must, therefore, be well above 40000 ADU for a correct calculation of effective non-linearity.

QC1 parameters

parameter QC1 database: table, name procedure
mean effective non-linearity vimos_detmon, lineff difference between the flux obtained at reference level from the polynomial fit and the flux predicted by the linear term only, normalized by the polynomial flux
reference level for lineff vimos_detmon, reflevel none (user-defined)
maximum exposure level vimos_detmon, counts_max mean exposure level of flat frame with highest exposure time

Trending

The effective non-linearity is monitored for both read-out modes (low and high gain).

History

December 2010 until September 2012: flat-field flux was variable on short time scales which resulted in arbitrary values for the effective non-linearity
September 2012 until December 2013: highest flat-field flux for high-gain read-out was well below 40000 ADU so that the effective non-linearity was not well determined

since end of March 2014: highest flat-field flux for high-gain read-out again well below 40000 ADU; HC plot no longer supported

top Gain

There are two exposures for each exposure time value of the flat-field sequence. The following equation is calculated for each pair:

mean(flat1) + mean(flat2) - 2 * mean_bias = gain * [sigma^2(flat2-flat1)-sigma^2(bias2-bias1)] .

Following the Photon Transfer Curve (PTC) method, the final gain value is calculated from a linear fit to the above equation. Alternatively, the median of the individual gain values can be calculated (method MED). Please note that gain is expressed in electrons per ADU.

QC1 parameters

parameter QC1 database: table, name procedure
gain vimos_detmon, gain from MED method (see above) since this is less sensitive to flat-field instabilities; PTC method was used until 2015-01

Trending

Gain is trended for both read-out modes (low gain, 2148x2440; high gain, 2148x4096).

History

July 2010: new detectors
2011-03-01: video board of FIERA A exchanged; gain changed for Q2 and Q3
2011-04-18: gain values for all quadrants adjusted
2011-06-01: video board of FIERA A exchanged again

2015-02-02: gain computation restricted to flat fields with less than 5000 ADU, usage of MED method instead of PTC; this gives a more accurate estimate of the true gain


 
[ESO][Index][Search][Help][News]