On Tue, 30 Nov 2004, David Berry wrote:
> > seems you have come to much the same conclusions that I did -- except I
> > was thinking about if I could make some use a MathMap with flux and
> > spectral coordinates as input and modified flux as output -- so this must
> > be more-or-less what's required for spectra.
> >
> > (Thinking ahead for a general system I guess you'd need to allow for other
> > Frames or SkyFrames as well as SpecFrames, the thought I had was that flux
> > can depend on azimuthal position, but that's not strictly a units
> > question).
>
> Flux depends on azimuth? Not heard of this one. Are you referring to the
> change in pixel size across an image caused by the projection? This is an
> issue which I'd like to shelve until we have a bit more experience in the
> spectral case.
Sorry should have been clearer, I actually meant airmass, certainly the
most significant flux change if you're interested in photometry. I did
have half a thought about pixel-level changes (in the guise of PSF
variation), but that maybe another problem.
> A case more closer to what I have in mind may be the use of TimeFrames
> rather than SpecFrames. To cater for the more general case, it may
> be better to use a name such as "DensityFrame" in place of "FluxFrame".
> That is, a Frame in which one axis represents an independent variable X,
> represented by a TimeFrame, SpecFrame, etc, and the other axis represents
> a density value, "Y per unit X", for some quantity Y.
Yes, that sounds more immediately relevent, time is part of the spectral
model.
Peter.
|