Good Calibration of CES

La Silla - Science Operation Department

CES Calibration frames

HOME

INDEX

HELP

NEWS

HOW SHOULD MY CALIBRATIONS LOOK?

This is a quick overview to guide you in calibrating your data. It should give you an idea of how the calibration images must look to obtain optimal results.
You will find many effects that might seem to indicate an error in the observations. The images below are meant to show you that most of these features are not a sign of an error but some normal and expected effects, and they can easily be removed from the science image as explained in the CES manual

Bias

Normal calibration image obtained with 4x1 binning, normal read out.

The image shows a section of the CCD from Y-coordinate: 0-780 pix
covering the full width in the X coordinate :0-256 pix

The bias is generally uniform although there is some electronic noise on one of the edges of the spectrum (at smaller wavelenghts). It is a still a good calibration since it can be easily subtracted from the observed image.

The lines on the botton are ripples also produced by the electronic noise. It is an insignificant effect that can also easily be removed.

Dark

Normal calibration image obtained with 4x1 binning, normal read out.

The image shows a section of the CCD from Y-coordinate: 0-780 pix
covering the full width in the X coordinate :0-256 pix

In this one-hour dark there are lots of cosmic rays presents in the image, and also presumably a light leak in the lower left corner. The electronic noise from the bias frame is also present here.

Background

Normal calibration image obtained with 4x1 binning, normal read out.

The image shows a section of the CCD from Y-coordinate: 0-780 pix
covering the full width in the X coordinate :0-256 pix

Compare this image with the previous dark frame.

Now with the shutter open there seems to be more light coming from the room, appearing in the bottom of the spectrum, which is easy to remove.

The light leak present when the shutter was closed can still be seen.

Flat fields

This flat field was obtained using a quartz lamp

Normal calibration image obtained with 4x1 binning, normal read out.

The image shows a section of the CCD from Y-coordinate: 0-300 pix
covering the full width in the X coordinate :0-256 pix

In this is low cut image we can see a ghost around 100-170 pixels.

Ghosts are due to reflections on the echelle grating and they are an unavoidable effect; it cannot be removed, it is the way the flat field will always look.

If the image is cut at higher intensities then the apearence of the ghost changes considerably. Fortunately only few orders are affected by ghosts and it is possible to move the grating, so the spectral line of interest is shifted and does not fall on the top of the ghost

So different cuts on the intensity are a good way of showing the structure of the ghosts.

Thorium-Argon Lamp

Normal calibration image obtained with 4x1 binning, normal read out.

The image shows a section of the CCD from Y-coordinate: 0-300 pix
covering the full width in the X coordinate :0-256 pix

We can see normal spectral lines plus the ghost mentioned above with less intensity that the other lines and shifted vertically.

There is also a rectangle of light that is a very low resolution emision line. This feature is due to stray-light from the exit slit wich has been dispersed by the predisperser and imaged directly by the camera mirror.