From jwf@ipac.caltech.eduWed Sep 10 14:43:55 1997
Date: Thu, 7 Aug 1997 16:44:22 -0700 (PDT)
From: jwf@ipac.caltech.edu
To: 2mass@ipac.caltech.edu
Cc: chas@ipac.caltech.edu, sstrom@donald.phast.umass.edu,
    stiening@ipac.caltech.edu
Subject: WG Mtg #129 Minutes

           IPAC 2MASS Working Group Meeting #129 Minutes
                             8/05/97

Attendees: R. Beck, R. Cutri, T. Evans, J. Fowler, V. Frey,
           T. Jarrett, J. Mazzarella, H. McCallon, B. Nelson,
           B. Wheaton, S. Wheelock, J. White


AGENDA

1.) Data Compression
2.) Read1 Solos
3.) August Shutdown Activities
4.) M67 Photometric Study
5.) J Bias Fluctuations
6.) Saturation Analysis
7.) Southern Hemisphere Scans


DISCUSSION

1.) Data Compression

    J. White reported that the reason why the hardware compression on the DLT
7000 is not effective is that the sustainable data rate into the drive is not
high enough. Apparently the drive compresses as much data as it has when the
time comes to write to the tape, even though more data could have been used
and higher compression achieved if the additional data had arrived in time.
According to Sun, IPAC is the first customer to report this problem. Since
there is no reasonable expectation of speeding up the data rate into the
drive, software compression will have to be used. John reported that DLTLOAD
will be capable of reading data that have been either software-compressed or
hardware-compressed or both, with or without differential-subtraction data
conditioning.


2.) Read1 Solos

    T. Evans and R. Cutri reported that the H-band 1-out-of-1 Read1 sources
(see last week's minutes) resulted from noise fluctuations. Here, "1-out-of-1"
refers to one apparition out of one frame spanned by the group of sightings as
organized by PFPREP; the single frame could have been anywhere in the scan,
not just near the beginning or end. When such detections occur too close to a
scan endpoint for 6-deep coverage to be possible, they are in fact removed by
MAPCOR after being used for persistence/glint processing. Such sources from
within 6-deep coverage are not removed and pass through the system to
the data base, where they will be removed during catalog preparation. Tuning
of FREXAS is expected to reduce the occurrence of these.


3.) August Shutdown Activities

    R. Cutri reported that the observatory is now in "shutdown" condition. No
data have been taken in August, and resumption is expected near the end of
August or early September, at which time M. Skrutskie will spend a week or two
at IPAC to work on tuning DARKS. Meanwhile, the focus at IPAC will be on
parameter tuning, quality analysis, and rerunning of certain nights.
Suggestions on which nights would be most useful to reprocess are welcome.
Lists of tunable parameters are due by Friday, August 8 (email to Roc).


4.) M67 Photometric Study

    V. Frey reported on a photometric study of M67 that she has been doing
with the protocamera data. 36 scans of M67 were used to obtain stellar
photometric data of very high internal consistency. Probability of variability
was computed, and several stars were found to have statistically significant
variation, but the largest amplitudes were still only about 0.1 magnitude.


5.) J Bias Fluctuations

    R. Cutri reported that he, B. Wheaton, and G. Kopan are working on
developing diagnostics for the J bias fluctuations. The indications are that
these are intermittent and not a problem on the majority of scans, so that
rejection of scans where they occur may be a viable approach to eliminating
their effects on the products. Rejected scans would have to be repeated later
at the observatory, and since the 24-hour quality asessment uses calibration
scans only, the turnaround time for rejected science scans could be large, but
it is hoped that not too many scans will be rejected.


6.) Saturation Analysis

    B. Wheaton reported that the saturation analysis has proceeded to the
point where single values for individual bands will be available soon. More
flat sequences need to be processed before the best values can be determined,
however. The generation of threshold images for each band, in which each pixel
has its own soft-saturation threshold value, will take significantly longer,
partly because this work has a much lower priority than the J bias fluctuation
analysis.


7.) Southern Hemisphere Scans

    R. Cutri and H. McCallon reported that scans in the southern hemisphere
exhibit a direction-dependent error in starting position. They have found the
error to be about a quarter of a degree, which means that opposite-direction
scans lose half a degree of overlap. So far, only calibration scans have been
done in the southern hemisphere, and calibration stars at the center of the
tile have been observed. Calibration fields with multiple objects will suffer
from the loss of half the north-south scan overlap, however, and it seems
likely that science scans may have a similar starting-point problem. This
issue is presently under investigation.