The Fringing Pattern

Images taken with EFOSC2 in 'red' filters (R, i, z) show considerable fringing. This is due to the slight variations of the thickness of the CCD. Note that the Gunn r (r#786) does not show noticeable fringing.

Fringe maps are produced by combining a number (as many as possible) of images with different pointings, allowing the removal of stars and other sources. The frames are first bias subtracted and flat fielded in the normal fashion, then averaged with rejection of the highest values at each pixel position to leave only the fringe pattern.

This page contains fringing pattern for EFOSC2, which can be used by visitors. They are combinations (following the procedure described above) of either 'blank' sky frames taken in technical time, or of data during visitor programmes. We thank those visitors who have shared their fringe maps with us for use by others; any users who wish to contribute a more recent fringe map than those listed below are encouraged to contact us.

Note that these frames have not been trimmed, and are the full 2060x2060 pixels of the CCD (or 1030x1030 for binned mode) for recent maps, and 2048x2048 for maps taken at the 3.6m. Note that the difference is due to a change in the default size of files; there is no difference in the CCD.

For very accurate removal of fringing users should construct their own fringe map from their own data frames, where possible, as the fringe pattern can vary slightly. Files are gzipped fits files (.fits.gz) and are ~3MB for 2x2 binning, ~13MB for 1x1.

The pattern is quite constant - tests in late 2008 with data taken at the NTT showed that the 2003 2x2 i-band fringe map gives good results. The intensity varies a lot, but the shape seems to be quite stable.

Filter ESO No. 1x1 2x2
Bessel R #642 TBD Aug 2008
Gunn i #705 Nov 2008 Sep 2003
Gunn z #623 May 2008, Nov 2008 TBD

How to correct for fringing?

Fringing is an additive pattern, hence it has to be subtracted from the images after the pattern has been scaled to the actual value of the fringing in the image.
  • Measure the difference between high and low level in some well defined parts of the image (preferably at positions with large fringing).
  • Measure the difference between high and low level in the same positions in the fringing pattern.
  • Compute the scaling factor by dividing the two values and scale the fringing pattern accordingly.
  • Subtract the scaled fringing pattern from the image ss.
You're on EPU
View live page