Re: IBVS reject notice

[Tom Droege <tdroege2@earthlink.net>]

Brian Skiff writes:

>      From what I've seen/read, the MkIV data-reduction seems to be very
>shaky, and I myself am not ready to use any of it yet.  The dense
>time-series are great, but could you program your machine to slip off
>to a standard field say two or three times per hour and take pairs of
>frames with somewhat different exposure times?  That would help a lot
>in tracking transparency drifts.

I have commented on this before, but working in public tends to get you a 
reputation for sloppy work.  If you discuss problems as they come up, 
people focus on the problems and not the good work that led you to identify 
the problems.  When I go to meetings, I want to hear about things that went 
wrong, not about things that are right.  I want to be able to profit from 
others mistakes.  Rarely do people publish their mistakes.  A pity!

>    Any more I'm not sure that "preliminary" data of the sort that's
>coming from the MkIV cameras has all that much value.  Instead it will

No Mark IV data has been published.  We are doing the equivalent of talking 
about our data in the hall.  So you focus on the problems as above.  I 
would more appreciate if you concentrated on prodding us to work toward a 
better result, as in your suggestion to periodically pick up standard star 
fields.  All you have to do with tass is join in the mail list and ask the 
right questions and push for the right techniques to get better data 
taken.  Possibly you will even use some of it some day.  I think users of 
data are responsible to encourage takers of data to take the right data.  I 
have been reading the other survey papers.  They are mostly putting up big 
lists of variable stars, without too much discussion of the accuracy of the 
results.  Yes, eventually we want to get everything on the standard system 
to some error limit.  There may be other things more important than the 
accuracy of a particular photometric measurement.  More accurate photometry 
will require better pointing than the present Mark IVs will handle.  I am 
designing a Mark V mount which will solve some of these problems.   At the 
moment, I cannot slip off a few times an hour to study transparency 
drifts.  But I have recently discussed with Michael R. how we might 
accomplish this.  I am thinking of pointing TOM2 at the equator and TOM3 at 
+4 degrees.  This gives some overlap between TOM1 at +7, TOM3 at +4, and 
TOM2 at say +1.  We can then switch TOM2 and TOM3 from time to time to 
cross calibrate them.  This will give us overlapping images between the 
three systems to transfer the Landolt standards from one field to the 
others.  This might tie everything together.  Brian, what do you think?

I have no intention of attempting millimag photometry.  This makes no sense 
at my suburban location.  What I can do is relative photometry to some 
error limit.  I can see things change.  With all the telescopes I am 
building I can cover a large area of the sky frequently.  I can do the 
whole northern sky with a uniform search.  I may also be able to do it 
longer than a professional collaboration that depends on new and unique 
data for continued funding.  So I can answer questions about the mix of 
variable types over the sky and distribution of periods for each 
type.  According to Bohden Paczynsky, such questions are presently poorly 
answered.

>to a standard field say two or three times per hour and take pairs of
>frames with somewhat different exposure times?  That would help a lot
>in tracking transparency drifts.

Could you explain what you are after here?  What does the different 
exposure time do for the calibration?

Coming from a different discipline I note that photometry is a really 
arcane art.

>really be helpful to be able to produce "real" calibrated stuff on the
>standard system.  The time is past for announcing one new variable at a

To me the "standard system" seems just awful.  To me it seems more like a 
procedure than a science of measurement with error limits attached.  As a 
naive outsider to this arcane art, it would seem that if I compare a star 
that I measure to say a Tycho star that has some known error limit with 
some particular filter and some particular filter calibration than I can 
quote an error limit that I can tie to the "standard" system.  So to me, if 
I can quote an error related to the standards, then I have a measurement 
calibrated to the "standard system".  OK, I sort of understand what is 
behind all this due to a nice note from Michael Richmond.

I was really pointed at this work by Bohden Paczynski.  In his papers I do 
not see much mention of the need for calibrated photometry.  He emphasizes 
completeness of the survey to give a statistical significance to the 
various types of variable stars.  I designed the Mark IVs to meet what I 
read in Bohden's papers.  I have been operating the engineering run in a 
way to answer some of these questions.  I note that the other surveys have 
picked several fields in no particular position.  (Apparently mostly dense 
star fields to get a large number of results.)  I have been taking uniform 
coverage of a large area.  I.e. the 360 degree band at +5 to +9 
North.  Seems to me that this data would be useful to answer some of the 
questions Bohden discussed in his papers.

I feel that I am being a victim of paradigm shift.  ;^)

Tom Droege

At 02:11 PM 9/16/02 -0700, you wrote:
>      Any more I'm not sure that "preliminary" data of the sort that's
>coming from the MkIV cameras has all that much value.  Instead it will
>really be helpful to be able to produce "real" calibrated stuff on the
>standard system.  The time is past for announcing one new variable at a
>time with sketchy data, but instead to work wholesale from calibrated
>data.  It would also be good to see results for known variables, since
>most of these are essentially unstudied, and almost certainly nothing
>on standard photometric systems are simply recent.  New variables are
>now very much ho-hum---the unsubtle subtext of the IBVS editor's
>complaint.  It's like finding a single new main-belt asteroid.  After
>the first couple, even for the TASS amateurs this should be quite
>mundane business, and the interest will lie in having lots of them well
>observed on standard systems.
>
>      Another thought is to contact Taichi Kato of 'vsnet' to ask if
>there might be some way for them to archive reduced results and make
>them available through their GUI/SQL interface.  Again, it would be
>easier at the start at least to provide results for known variables
>which are already in their system.
>
>\Brian