next up previous 78
Next: Second step-for Filippenko & Greenstein data
Up: Flux calibration
Previous: First step-Turning the table into a spiketrum


Second step-for Oke & Gunn data

The command `cspike' takes this generated spiketrum (hdspike) and combines it with the observation of the standard (standobs) to generate a new spiketrum, whose values are now the instrumental response sampled at the points of the original spiketrum.

   ICL> cspike spiketrum=hdspike spectrum=standobs output=calspike

generates this new spiketrum and calls it `calspike'. The values in `calspike' are calculated for each point by summing the counts in the observed spectrum over the appropriate wavelength range, dividing them by the wavelength range and the exposure time, and dividing the result into the flux density value given in `hdspike'. The units of `calspike' are therefore `units per (count per Angstrom per second)'.

`calspike' can be turned into the calibration spectrum required by `spflux' by interpolation. The most direct way to do this is just to use the command `interp'.

   ICL> interp spiketrum=calspike spectrum=calib

will produce a new spectrum, `calib', which will be the required calibration spectrum. However, you may find that you are not happy with `calib'. It may not be smooth enough. It may be that if there are regions of absorption in the spectrum, poor alignment of the wavelength ranges will result in some spurious values. It may seem to be cheating, but the most direct way to get round this is to edit the spiketrum `calspike' using the `spied' command.

   ICL> spied spiketrum=calspike output=modspike

will display the spiketrum and allow you to delete points, insert new points, and see the results of spline interpolation or global polynomial fitting to the modified points. When you are happy, you can run `interp' on the modified spiketrum to produce a less honest, but more satisfactory, result.

Global polynomial fitting may seem a trifle crude as a way of interpolating between spiketrum points, but it may be that this gives a better result in some circumstances.

   ICL> spifit spiketrum=calspike order=5 spectrum=calib

would be an alternative to the use of `interp' to generate the calibration spectrum.



next up previous 78
Next: Second step-for Filippenko & Greenstein data
Up: Flux calibration
Previous: First step-Turning the table into a spiketrum

FIGARO A general data reduction system
Starlink User Note 86
Keith Shortridge, Horst Meyerdierks,
Malcolm Currie, Martin Clayton, Jon Lockley,
Anne Charles, Clive Davenhall,
Mark Taylor, Tim Ash, Tim Wilkins, Dave Axon,
John Palmer, Anthony Holloway and
Vito Graffagnino
2004 February 17
E-mail:ussc@star.rl.ac.uk

Copyright © 2008 Science and Technology Facilities Council