Eclipsing Cepheid paper

[Doug Welch <welch@physics.mcmaster.ca>]

Folks,

Tom has asked how one or a small number of bad points effect
the power spectrum. The answer is that they primarily serve
to raise the level of the "grass" across all frequencies.
(This is due to the fact that the Fourier Transform of a 
delta function is a constant level.) All that does is reduce
the signal-to-noise of true power spectrum peaks.

People on this list may be interested in a preprint that is
the result of a large survey - the MACHO Project, finding relatively
rare but really interesting variables:

http://arxiv.org/abs/astro-ph/0201481

The power spectrum of such data will show the alternating high and
low peaks at low frequency of an eclipsing binary and the fundamental
and harmonic peaks of the Cepheid's lightcurve (harmonics only if
it is sensibly non-sinusoidal) at much higher frequency. If there
is interest, I will place a file of times and magnitudes for one
of these systems where people can easily get at it. (It would be
a good test set for running DFT's and CLEAN on people's computers.)

Cheers,
Doug

PS. The analysis in this paper is largely the result of the work of 
my former M.Sc. student (and now Ph.D. student), Dave Lepischak.