AB magnitudes, and a test

A calibrated spectrum generated in Jansky units (and others, eventually) can be converted to AB magnitudes by `abconv'. Our `calobs' spectrum may be converted by

   ICL> abconv spectrum=calobs output=abobs

(`abconv' works out the units of the input spectrum by looking at a .UNITS data object in the input file. In some cases, it may not recognise the units-this can happen with spectra that have come from some other system via a translation program. If that happens, and you know that the units are, say, Janskys, you can set the units by hand using

   ICL> setobj object=calobs.UNITS value=Janskys

`abconv' will recognise `mJy', or anything that contains `Jansky' and the words `milli' or `micro'. The same remarks apply to `flconv'.)

An interesting test of the system is to calibrate an object using itself, convert the result into AB magnitudes, and then compare the result with a spiketrum generated from the published AB magnitude tables. For example, remembering that our original spectrum of HD 84937 was `standobs', and given that there is a table called `hd84937a.tab' in the main Figaro directory that has the values in AB magnitudes, we can try

   ICL> spflux spectrum=standobs calspect=calib output=hdcal
   ICL> abconv spectrum=hdcal output=abhdcal
   ICL> gspike spectrum=abhdcal table=hd84937a output=abhdspike
   ICL> splot  spectrum=abhdcal reset accept
   ICL> splot  spectrum=abhdspike noaxes noerase accept

The result should be the spectrum of HD84937 in AB magnitudes, with a series of spikes just touching the spectrum, indicating that (at least at the tabulated wavelengths!) the spectrum has been calibrated to the correct value.

A much tougher test would be to use one standard to calibrate another, and then compare the result with the tabulated values. This does of course require that you really do have spectrophotometric data, with no filter changes, with all of the object in the slit in both cases, and with no clouds in the way.