Hi Jianyang

I finally got the stardust stuff to the point where I think you can
make use of it.  I am sending a file that includes a list of all the
data (both Wild 2 and Anne Frank) along with the commanded exposure
time and the direction of the shutter.  I think that is the
information you need to start looking at the photometry/shutter issue.
(The long-exposure cruise data won't be of any use to you, but the file
is what I produced for the archive documentation.)

It looks like the Anne Frank data might provide a good test.  Long
strings of short but constant exposures.  If the images don't have
other issues involved, then these may be good enough to get the timing
offsets for both the forward and backward shutter directions.  Then we
can see if the Wild 2 images are consistent with that result.

Just remember that in the headers and labels of the Anne Frank data,
1.65 ms has already been subtracted from the exposure time in the even
numbered frames.  So you need to take that into account when you are
looking for your own offsets.  (The commanded exposure times in the
file I'm sending do not include the correction, so you can use them as
a starting place if you want.)

There are a few places where missing images make the shutter direction
ambiguous, because it's not known if the image was never taken (which
changes the parity) or if it was take and lost (which preserves the
parity).  I've included brief notes about missing files, and when its
unknown whether the parity changes, I've included a question mark next
to the shutter direction.  There are a couple spots where images are
missing from a sequence, but the documentation suggests that this was
intentional, and all images were actually obtained, but not kept.  We
should check this out to make sure that the data are consistent with
that.

Also, you should use the earliest available data that we have
(e.g. .img files), so that you don't end up unknowingly using
something that has already had a change applied to it.


One question that Ken had was why the Forward/backward parity seems to
be different for Wild 2 and Anne Frank.  It could be that the team got
Anne Frank backwards, like they did for Wild 2.  Or it could be that
one of these "unsaved" frames was never actually taken, switching the
parity.  Hopefully you can resolve that by comparing the two encounter
data sets.


As I mentioned last week, we'd like an independent measure of the
offset times of the shutter in both directions, if we can get it.  I
think a brute force method is perfectly acceptable for finding the
result.  (E.g. set up a nested loop for a forward offset and a
backward offset, each ranging from +5 to -5 ms, and see what
combination produces the most consistent result in the photometry.)
Feel free to do something more elegant or more in-depth if you want.

I'd also like to document everything that we have done to come to the
conclusions that we eventually settle on, so if you generate plots
that you think help to illustrate the results, make a copy so we can
include them in the archive.

Thanks again for looking at this.

Tony