Re: GSC 748-1618

[Tom Droege <tdroege2@earthlink.net>]

I agree with Michael R. that his process would produce a better agreement 
with Michael K's curve.  So I will take his side and share the cost of the 
dozen donuts should he prove wrong.  But this is for the moment a thought 
experiment.

At first I thought this would involve reprocessing all the data, but it 
might be easier than that.  We do have all the .cal files.  So one looks 
through all the .cal files for occurrences of the object and comparison 
star. These must be matched up in time.  We keep pairs of measurements 
taken at the same time and throw away all the others.  Now we can derive a 
delta measurement from the comparison star as Michael R suggests.

OK Michael S.  This looks like a job for S Michael man.  ( How did tass 
attract all these Michaels?)

Ask your data base the above question for GSC748-1618.  Should be a fun 
project.

However, thinking just a little more about this, the .cal files have had a 
color transformation applied.  To escape this possible bias, we have to go 
back a few steps.  This does involve reprocessing all the data.

Tom Droege


At 04:22 PM 2/17/03 -0500, you wrote:

>   Suggestion: Michael K's measurements are, I believe, differential
>with respect to a comparison star.  The TASS values are based on
>ensemble photometry, using many stars over a frame ... and since
>the field centers aren't always the same, the stars in the ensemble
>can change.  In addition, the variable and the comparison star
>are wandering around through the field from night to night in the
>TASS dataset.
>
>   One might make a fair comparison between the datasets in this way:
>
>        a) make a light curve using Michael K's differential measurements
>        b) select the TASS values for the variable (has already been done)
>        c) select the TASS values for the comparison star
>        d) calculate a differential value from the TASS data:
>
>                   TASS variable - TASS comparison
>
>        e) compare the two sets of differential measurements
>
>   I'll bet a dozen donuts that the scatter in the TASS measurements
>decreases significantly.
>
>   Of course, this only makes sense if the comparison star appears
>in a significant fraction of the TASS images.
>
>                                        Michael Richmond