Dear PI,

We have examined the data for project RD000 on 15 Apr. 1998. Here's a summary:

1. Tape weights vs. time plots have been generated for the entire time range
   of your experiment. These are a measure of tape record/playback quality,
   representing the fraction of valid data samples. Data with weights below
   70-75% should be flagged.

2. Delay, rate, phase, and amplitude plots were made for the observation.

3. Autocorrelation bandpass plots were generated for all antennas for scans
   on OJ287 4C39.25 3C345 OQ208 VR422201 0553+398.

4. Cross-correlation bandpass plots were generated for all baselines to HN for
   scans on OJ287 4C39.25 3C345 OQ208 VR422201 0553+398.

5. Weather and Tsys plots for the VLBA antennas for your project will be sent
   with the data tape(s).

NOTES:
BR - Something very strange is going on with the sniff of BR.  The sniffer
fringe fit gives a delay offset of over 500ns for the period of the first
correlator job, then reverts to about 100 ns for the rest.  The difference
should show up as several turns across the 8 MHz basebands.  But the
crosscorrelation plots from the sniffer and from AIPS don't show any
such effect.  The phase slopes look reasonably constant.  I think this
is a sniffer problem, not a problem with the data.  I fringe fitted the
first job in AIPS and got delays a bit over 100 ns for BR.  John Benson
now has the data to debug the sniffer.

GG - Between something like 06:36 and 08:50, there are various log
data gaps and problems with the tape drive.  I'm not sure why, but the
job generator didn't include the station in jobs during this time, even
though the data seemed to be there for some.  I suspect that missing
log pieces confused it.  I don't see that the amount of missing data
justifies any effort to recover it, especially since such effort might
not work anyway and this is not a critical station.

NY - There were intermittent problems with poor playback on some
tracks and a period of almost 2 hours where playback was poor on
almost all tracks.  Basically, the recording is not robust and at
least some channels were lost for something like a total of 10 hours.
It looks like X band channel 4 was much more severely affected than
others.

ON - Playback was mostly ok, although the station was in search for
two periods of about an hour and one of about half an hour.  Also
there was poor performance on a couple of tracks for a total
of about in hour (in two hunks) that wiped out one or more channels.
The rest of the time, things look mostly ok, although slightly 
ragged weights suggest playback that was not as robust as most stations.

KK - Has somewhat ratty weights for the first tape.  It looks good
after that, other than a 10 minute or so period late in the experiment
where it looks like a track was clogged (playback?) and channels were
protected by barrel roll (I think).

MC - lost about an hour near the end of the experiment.  It looks like
there were tape problems.  Better notes in the logs would be useful.

MK - Had a tape drive problem at observe time that caused it to not
run for about the first hour.  Then it was in stall on correlation for
another half hour after it should have started.

NL and, to a lesser extent, LA and PT - Had a few more low weight
points than other VLBA sites.  The amount is too small to affect the
experiment, but I presume that the recordings are not up to quite the
same quality as the others.

GN lost about an hour after the break for the another project.  It
seems that the correlator was in stall.  I had this recorrelated and
it was fine.  The recorrelation was done only with the VLBA stations
(a compromise for somewhat reduced effort).  That job has been
distributed separately from the rest of the experiment.  In the
future, we probably should start a new job when GN comes back after
these breaks.

BR, FD, GC, GN, KP, MC, MK, OV, SC, and WF have nearly perfect weights
indicating good recordings (ignoring the lost data at the start at MK,
right after the break at GN, and very near the end at MC).

NL and LA have their new S band filters.  They seem to have eliminated
the RFI problem at those sites.

MC has a very strong spike near DC in the 8420.99 channel.  It also
has strong RFI at S band.  In this experiment, MC was the only site
with significant RFI problems.

PT has some sort of RFI near the edge of the 4th S band channel.
It is not strong enough to be a big problem.  Oops - I speak
too soon.  It is very strong at around 4:21.  It is not always
there.

MC at 9:17 has a very very low amplitudes in S band channel
3 compared to the other S band channels. This was not true
at 0:04.  There are too few cross correlation spectra plots
to let me investigate this.

GN has no fringes from about 20 - 21 hr.  Don't know why.

Several sites show periods of an hour or so where the sniffer
fringe fitter show no data while the weights plots show no problem.
I presume that this is a sniffer effect, but am not sur