Re: instrumental magnitudes vs. calibrated magnitudes

[Chris Albertson <calbertson@logicon.com>]

jbush@pucp.edu.pe wrote:
> 
> michael
> 
> i have to agree and disagree with you
> 
> of course, the latter is more my nature :]]]]
> 

Mine too.  So read on....

Don't think of it as "missing stars in V band" think instead of "extra
stars in I band".

It could happen that we will never get a V band measurement of some
stars, but 1) there will be more then one Mk IV camera site.  A match
could be observed at an other site  2) If
you are looking for variable stars I band only is good enough.
3) Many astronomers seem to do just fine with only one telescope
on their mount.  They never get simultaneous matching observations.

Assuming we want to do unequal exposures, the time
over which the telescope must track the sky is determined by the camera
channel with the longest exposure.  You can't move on to the next target
until the last shutter closes.  We may as well adopt the convention of
using the longest exposure
as the "base line" and say we are closing the other shutter early.
This is clearly equivalent to giving "extra exposure" to V.

So now, by our convention, what you are asking for is to close the I band
shutter early.  Now you have what you want, roughly equal number of stars
in both frames.

But do you really have what you want?
Yes, roughly equal numbers but _not_ the same stars in each frame.
Remember, stars have color.  Many Blue stars will detected by the V camera
that will not appear at all in the I camera and many red stars will appear
on the I camera but not the V camera.  Some bright stars will appear in
both but it actually turns out that close to half the stars will be near
the detection limit.  In fact if you look at Mark III data where two of the
cameras are nearly identical you will find many stars seen in only one of
the two identical cameras.

One other point is that both cameras are read out at once.  So while you
could close the I band shutter, the CCD would just set there in the dark
collecting dark current while the V band CCD is integrating more image photons.
You'd be giving up S/N just to equalize the star counts.  Would it not be
better to keep the I band shutter open and increase the SAN then later if
you have too many stars toss then out. Not that I would do that, but the
end result be better then if you had closed the shutter early. 

I think the best place to thin out data is in post processing, after the images
are converted to lists of objects.  You can do this very early, e.g. keep only
the intersection of the I and V lists from each pair or later e.g. Michael's
tenxcat that is computed late in the database building process.  I'd suggest
a halfway point.  I'd suggest building something like a "twoXcat" and sending
that on the Michael so as not to flood him with too much "seen only once"
junk.

-- 
  Chris Albertson

  calbertson@logicon.com                  Voice: 626-351-0089  x17
  Logicon, Pasadena California            Fax:   626-351-0699