Change page style: 

Bias Images (N and S)

Example bias images are available for the following instrument/detector combinations:

 



GMOS-S Hamamatsu 

A high signal-to-noise 4x4 binned bias image for GMOS-S with the Hamamatsu CCDs is given below (overscan subtracted). These data were taken in slow read, using all twelve amplifiers. A single line plot through the bias image is also presented. As it can be seen, the bias images have some structure, in a similar way than GMOS-N CCDs has. This structure is readily visible in this high signal-to-noise example but is still present in individual bias images. There is also one bright column associated with amplifier #5.

It is recommended therefore to always create and subtract bias images from science data in order to remove the structure. In addition, overscan subtraction is recommended particularly for faint target spectroscopy as the overscan level is expected to drift with time both in absolute value and with respect to each individual amplifier. In the data presented here the average overscan levels were as follows:

 

  • extension [1]: 3316
  • extension [2]: 3270
  • extension [3]: 3359
  • extension [4]: 3218
  • extension [5]: 2985
  • extension [6]: 2948
  • extension [7]: 2942
  • extension [8]: 3025
  • extension [9]: 3084
  • extension [10]: 3071
  • extension [11]: 3113
  • extension [12]: 3162

GMOS-S Hamamatsu bias image
GMOS-S bias image with the Hamamatsu detectors, binned 4x4, slow readout, using all twelve amps.

Line cut through GMOS-S Hamamatsu bias image
Line cut through the GMOS-S bias image above.

 

 


GMOS-N e2v DD

A high signal-to-noise 4x4 binned bias image for GMOS-N with the interim upgrade e2vDD CCDs is given below. This image is composed from 1000 individual bias images, and has been overscan subtracted. These data were taken in slow read, using all six amps. A single line plot through the bias image is also presented. There are several notable features in this bias image:

 

  • The level is not zero, as might be expected from an overscan subtracted bias image. The average bias level after overscan subtraction is on the order of 0.6 ADU.
  • There is one bright column associated with every amplifier. If the amplifier is on the right side of the detector then the bright column is the furthest one to the left of the image (the last image pixel read out before the overscan). For six amp data this leads to the unfortunate result of two bright columns in the middle of each detector. The level in each column is between 1-3 ADUs. The accelerated schedule for the installation of these detectors in GMOS-N left no time to optimize the operating voltages which may have reduced or eliminated this effect.
  • In addition to an overall bias level and the bright columns, the bias images have additional structure. This structure is readily visible in this high signal-to-noise example but is still present in individual bias images.
  • There is evidence of high-frequency noise in four of the amps, particularly in CCD1 but at a lower level also in CCD3. The bias image for CCD2 show no evidence of this effect.

 

It is recommended therefore to always create and subtract bias images from science data in order to remove the features listed above. In addition, overscan subtraction is recommended particularly for faint target spectroscopy as the overscan level is expected to drift with time both in absolute value and with respect to each individual amplifier. In the data presented here the average overscan levels were as follows:

 

  • extension [1]: 1180
  • extension [2]: 1059
  • extension [3]: 915
  • extension [4]: 1106
  • extension [5]: 1187
  • extension [6]: 1105

 

Please remember that with this detector controller the science extensions do not correspond to the pixel data physical location in a straightforward manner. Refer to the diagram on the GMOS data format and reduction webpage for the mapping of image extension to pixel location.

GMOS-N e2VDD bias image
GMOS-N bias image with the interim upgrade e2vDD detectors, binned 4x4, slow readout, using all six amps.

Line cut through GMOS-N e2vDD bias image
Line cut through the GMOS-N bias image above. This is a high signal-to-noise image derived from 1000 individual bias frames, the read noise on a single bias image varies by amplifier between 1.35-1.45 ADUs. Please see the GMOS-N Array (e2v DD) webpages for gain and read noise information from individual amplifiers and for the different readout modes.




GMOS-S EEV blue

An example bias image for GMOS-S with the E2V CCDs is shown below. In slow read out and using the best amplifiers the bias levels during the commissioning observations were 599 ADU/pixel, 579 ADU/pixel and 627 ADU/pixel for detector 1, 2 and 3, respectively. The greyscale used blow is chosen to show the spatial structure in the bias frame, rather than the relative levels in the three detectors. The data are from the commissioning and have been released as part of dataset GS-CAL20030524.


Bias image: Slow read and best amplifiers.



GMOS-N EEV red

An example bias image for GMOS-N with the original red EEV CCDs is shown below. In slow read out and using the best amplifiers the bias levels during the commissioning observations were 749 ADU/pixel, 633 ADU/pixel and 387 ADU/pixel for detector 1, 2 and 3, respectively. The greyscale used below is chosen to show the spatial structure in the bias frame, rather than the relative levels in the three detectors. The data are from the commissioning and have been released as part of dataset GN-CAL20010910.

Bias image: Slow read and best amplifiers.