Re: Calculation of mag errors

[Andrew Bennett <andrew.bennett@ns.sympatico.ca>]

On Thu, 3 Mar 2005 17:27:30 -0500, Stupendous Man
<richmond@stupendous.cis.rit.edu> wrote:

> ...
>  Hmmm.  Before this week, I would have told someone to ignore
>the uncertainty values for _most_ purposes.  Now, I would tell
>someone to ignore them for _all_ purposes.

Now I can dig out the Ensemble photometry
code I put away in disgust a year ago. It was
trying to do a proper error budget on the assumption
that the photon noise was calculated right in the
Database. All the unwanted variance got thrown
into a catchall "image noise" term that got applied to
all stars on an image. Underestimation of the photon
noise causes the other noise terms to be systematically
and hugely overestimated. The resulting large "errors" 
led to the  bright stars being weighted down by huge 
factors  compared to what they should have been, doing
the fit no good at all.

Another reason for keeping raw data ...

Andrew Bennett, Avondale Vineyard, NS, Canada