You are here: Home / Surveys & Projects / WFCAM / Technical / Crosstalk

Crosstalk

How crosstalk is measured and removed
All the detectors show similar cross-talk patterns. The cross-talk artefacts are essentially time derivatives of saturated stars with either a doughnut appearance from heavily saturated regions (±128 × m pixels either side, where m is the interleave factor m×m from microstepping eg. 1×1 2×2 3×3, plus progressively weaker secondaries further multiples on), or half-moon-like plus/minus images from only weakly saturated stars. Adjacent cross-talk images have features at ~1% of the differential flux of the source, dropping to ~0.2% and ~0.05% further out. Beyond three channels the effect is negligible. The following figure illustrates where cross-talk images are expected in detector X-Y space relative to the position of saturated objects for each quadrant. The cross-talk pattern rotates from quadrant to quadrant because the readout amplifier for each quadrant is on a different edge of the detector (denoted by the red lines).

crossdiag.gif

Fortunately these artefacts are non-astronomical in appearance and do not "talk" across the detector quadrant boundaries, or between detectors. However, they do not generally stack out for a jitter sequence.
An example of cross-talk is shown below for a heavily saturated image. The cross-talk image is to first order the time derivative of the signal in the saturated image channel along the readout direction. Significant cross-talk only appears to be generated by saturated images.

crosstalk1.jpg