FORS1 Quality Control: QC1 from screen flat frames

FORS1:
QC1 from screen flat frames

QC1 PLOTS: Screen Flat CONAD FPN

QC1 ASCII DATA

Year:
Item:
Filter:

QC1 database (expert mode)

SCREEN_FLAT_IMG: structure parameters, contamination, conversion factor

This page refers to the old FORS1 CCD, which was in operation until March 25, 2007. Data on the new CCDs can be found here

Modes
Parameters are measured for the 5 standard filters UBVRI_BESSELL and CCD mode fast (4port), high gain.

Parameters

Efficiency: median port A signal divided by integration time. This is only a nominal 'efficiency' actually measuring the stability of exposure level (influenced by e.g. aging of lamps etc.).

Photon noise: measured as 100x100 pixels sigma in a difference frame made of two equally exposed raw files (hence fixed-pattern noise cancels out). Photon noise should scale as (signal)^1/2.

Fixed-pattern noise: A derivative frame of a master M is created (by shifting a copied version M' by 10 pixels in X and Y so that fixed-pattern noise is randomized, and calculating the difference frame M -M' ). It is divided by a background fit. A 100x100 pixels sigma is measured. The photon noise is subtracted geometrically from that number. Fixed-pattern noise should scale with signal.

Sigma AB: fractional difference between median levels in ports A and B (which show the largest difference).

Contamination: measured as a change of large-scale gradients across port A (by comparing to a reference frame), after removal of small-scale contributions. This parameters measures the introduction of additional structure in the flats (e.g. due to dust). To avoid overcrowding, only values for U_BESSELL are shown (which have the largest values).

Noise vs. flux: checks for scaling properties of small-scale noise contributions. Fits with typical values are plotted as red lines.

Conversion factor (1/gain): measured in a raw file (high gain mode) as 100x100 pixels sigma, corrected for fixed-pattern contribution. The sigma is compared to the square-root of the signal averaged across 4 ports. (To avoid crowding, only data from the B_BESSELL filter are marked by red circle, all others by dots). The number given is in e^-/ADU and is averaged across all 4 ports.

CCD contamination

CCD Contamination is present in FORS1. This time dependent effect is clearly visible when computing the ratio of two Imaging Screen Flats taken at two different epochs. In the following example, two U Master Screen Flats taken one just after a decontamination process (1999-11-02) and the other just before the next intervention (2000-01-27) have been used to compute the ratio.

When the contamination is judged to be too high, the detector is usually baked. Such operation is performed on a periodical basis and this is reflected into a saw-tooth behavior of the contamination amount. A detailed description of contamination measurements is given in contam_report_fors1.ps (400kB).

NOTE: from the beginning of FORS1 operations the contamination was growing very fast and this required very close in time calibrations in order to properly remove it from the science frames.

ADU conversion factor: CONAD

The values of the CONAD mesured over the fours ports (ABCD) are averaged and trended. For the time being, the trend is monitored only for one readout mode (ABCD,1x1,high).

QC home
QC and trending
Paranal autrep db