Date: Tue, 23 Sep 1997 18:14:56 -0700 (PDT)
Subject: 2MASS WG Mtg #130 Minutes

           IPAC 2MASS Working Group Meeting #130 Minutes

Attendees: R. Beck, R. Cutri, D. Engler, T. Evans, J. Fowler,
           L. Fullmer, T. Jarrett, D. Kirkpatrick, G. Kopan,
           B. Light, J. Mazzarella, H. McCallon, B. Nelson,
           J. White


1.) 14-Bit Tests
2.) Data Compression
3.) QUALITY Subsystem Requirements
4.) Sample Data Base Loading
5.) Optical-Catalog Associations
6.) Calibration Strategy Change


1.) 14-Bit Tests

    R. Cutri reported that a test is underway in which the raw
data are simulated as 14-bit values by masking off the two least
significant bits in the pixel values of all frame data. The
purpose of this test is to judge the impact of going to 14-bit
A/D converters in the camera electronics. This option is being
considered as a solution to the problems encountered with the
current electronics in the cameras for both hemispheres (although
it is possible that just the A/D converter for J band in the
northern-hemisphere will be replaced, in this case with a 16-bit
A/D converter; this would be expected to solve the J-band
single-quadrant bias-jump problem).
    J. Fowler reported that the standard deviations about the
dark trimmed averages increased by about 0.12 DN for the Read1
darks and by about 0.33 DN for the Read2-Read1 darks. G. Kopan
reported that early checks of the flattened-frame noise showed
increases similar in size, possibly a bit smaller. R. Cutri
requested all team members to investigate differences in the
products once the test run finishes, which should be this evening
(9/23/97). The old results are in o2/TEST/970608n, and the new
results are in /o1/TEST/970608n.

2.) Data Compression

    R. Beck reported that the newly installed compression scheme
was getting five nights' data on a single tape. These were summer
nights, typically 14 GB each. It was noted that the previous
estimate of three nights per tape was based on longer nights.
This should tend to offset the recent $10 price increase of DLT
    R. Cutri reported that the combination of the IPAC
precompensation and DLT hardware compression has been tested at
UMASS and will likely be installed at Observatory soon.

3.) QUALITY Subsystem Requirements

    The requirements for the QUALITY Subsystem have been
continuing to accumulate as new failure modes are found. The
current system is working well, but additional capabilities are
being designed and implemented to aid in interpreting the numbers
displayed, and interface discrepancies are being worked out. C.
Beichman has requested an updated requirements document and
schedule for version 1.0 of this capability. R. Cutri, L.
Fullmer, and D. Kirkpatrick will collaborate on this task.

4.) Sample Data Base Loading

    T. Evans reported that problems have arisen in the database
loading activity that is being performed to support the sample
database composed of the results from eleven selected observation
nights. Some fields appear to overflow occasionally in some input
files, and other files appear to be missing. The former problem
has been encountered with some extended-source files, and the
latter with some calibration files. T. Evans will work with T.
Jarrett and S. Wheelock to resolve these discrepancies. The error
information provided by Informix leaves much to be desired, as it
does not indicate which field it finds unacceptable, and not
always which line of input data.

5.) Optical-Catalog Associations

    R. Beck requested information on the schedule for the
upcoming coordinated redeliveries of POSPTS, GALWORKS, and CALMON
in support of the optical-catalog associations. Adding the
association fields to the output of POSPTS has caused changes in
several software interfaces, necessitating the coordination in
redeliveries (DBMAN is also involved, but will not attempt to
operate on the new data until some future time after the
eleven-night sample database has been successfully set up).
    Since the 14-bit test is still running, and since one
additional diagnostic capability has been added to the POSPTS
design, it was decided to wait until Thursday to perform the
redeliveries. It should be noted that this version of POSPTS will
include the final position-refinement capability, which H.
McCallon was able to code and test sooner than expected. Tests
showed that the final positions differed from the POSFRM solution
by about 0.1 arcsec in the scan direction and by about 0.05
arcsec in the cross-scan direction.

6.) Calibration Strategy Change

    R. Cutri reported that the calibration strategy has been
changed from requiring two calibration observations every two
hours to one every hour. A calibration observation consists of
six calibration scans. This will permit better time resolution of
calibration changes, resulting in smaller blocks of data having
to be rejected due to failure of the requirement to be bracketed
by three acceptable calibration scans.