[Aavso-photometry] When to submit "Fainter Than" versus actualnumbers.

[Jeff Hopkins]

Hello Gary,

Since the procedures are different for CCD photometry from single 
channel work I am not sure exactly what's going on.

First, what do you mean by "time series?" From my experience with 
single channel variable star photometry, they are basically two 
conditions. One where the star is changing brightness on the order of 
minutes (e.g., Algol or a new supernova) and long term variable where 
there are no significant changes in brightness over a period of hours 
to days. In the first case for fast variables I take just one set of 
3-10 second reading in each color for the program star and two sets 
for the comparison star bracketing the program star measurements. 
Then I reduce the data and determine a magnitude of the program star. 
Since there is only one magnitude for each color determined, there is 
no standard deviation.

For slow variable stars I do essentially the same thing, but do it 
three times and determine 3 magnitudes for the program star which are 
then averaged and a standard deviation determined.

The reason for the bracketing the program star measurements with 
comparison star measurements is to try to cancel or even out any 
changes in the sky during the measurements. With CCD photometry where 
the comparison and program stars are in the same frame, that should 
not be a problem as the sky is the same for both at the time of the 
exposure.

Now again I don't know your procedure, but it would seem to me that 
one exposure would produce one program star magnitude (after 
processing). Multiple exposures would produce multiple program star 
magnitudes which then could be averaged and a standard deviation 
determined.

For your BZ UMa data over 4 hours, I think you would have been better 
off grouping sets of 3 measurements and figuring a standard deviation 
and average for those. A standard deviation of 0.192 is not good. The 
star may actually have changed over that time so the SD is rather 
meaningless.

As I have mentioned before, it does not makes sense to me to worry 
about SD of comparison and check stars. The sky usually does change 
even over small time periods. When capturing both the comparison and 
program stars at the same time (could be done with dual channel 
photometry), you don't care about the sky condition, because the 
important part is the reference to the comparison star. That is the 
whole basis of differential photometry.

I am not familiar with the software you mentioned, but if it can 
produce a program star magnitude for each exposure, that's what I 
would use and forget about any other data from the program.

Jeff

At 10:52 -0700 1/26/05, BailyHill at aol.com wrote:
Hello Jeff and All;

I started thinking about this, and had a couple of questions and comments. 

First, I use Maxim for software.  The way I understand it works, is 
that one designates a reference star in the image, and tells Maxim 
what value to assign to it.  So in the results, the Reference star 
always is defined with this value.

One then tells Maxim which star in the image is the "Object".  One 
also may designate several "Check Stars".  My understanding is that 
the "Object" and "Check" stars are measured, relative to the 
Reference Star.  So the "Object" and "Check" stars are not measured 
for absolute brighness, but are measured relative to the "Reference 
Star".

So when I do a time series, I have several check stars with their 
values measured relative to the "Ref star".  It seems to me that if 
these check stars are close in magnitude and color to the Object, 
that the Std Dev of the Check Stars gives a good representation of 
the total error of the Photometry Process. 

In fact, a recent run on BZ Uma at minimum, the 15.6, 16.3 and 17.0 
Check Stars had Std Devs of .045, .056 and .066 respctively, for 95 
measurements, over a 4 hour period.  BZ Uma had an average of 15.830 
with a Std Dev .192.  It seems clear that BZ Uma was active that 
night. 

In that same run, the brighter check stars 13.1, 13.7 and 14.8 had 
Std Devs of .018, .024 and .017 respectively, for these 95 
measurements of 2 minutes each. 

So if Maxim works as it is described in the manual,  I don't 
understand why the Std Dev is not the best representation of the 
process--assuming of course that the Check stars are not variable.

It would be interesting to hear how other software packages handle 
the time series measurement.


Clear Skies
Gary


-- 
Jeff Hopkins
HPO SOFT
http://www.hposoft.com/Astro/astro.html

*********************************************************
  Small minds speak about people * Average minds speak of events
************   Great minds speak of ideas!    ****************
*********************************************************
		     Hopkins Phoenix Observatory
  		       7812 West Clayton Drive
		   Phoenix, Arizona 85033-2439 U.S.A.
        			www.hposoft.com