Due to characteristics of the EB039D observations (many passes, multiple
modes, short scans) the normal standard plots weren't really sufficient to
provide an a an adequate review of the entire scope of the experiment.  We
therefore made a different set of plots manually, as described below.  These
are in general more exhaustive than the usual standard plots, but also more
complicated, since each pass was handled in a slightly different way. For
specific data-quality comments that can be inferred from the plots, see the
EB039D cover letter.

The names of the plots files and a brief synopsis of their contents for 
the three different passes are:

GEODETIC PASS:
    eb039d.geod-weight-[1,2].ps          station weights vs. time
    eb039d.geod-auto-[1,2].ps            station autocorrelation vs. freq
    eb039d.geod-cross-[1,2][a,p].ps      baseline amp & phase vs. freq
    eb039d.geod-ampphase-[1,2][a,p].ps   baseline amp & phase vs. time


  The two versions of the time-based plots (weight, ampphase) correspond
to the two ~hour-long geodetic blocks at the beginning and end of the
observations. These plots show all 8 IFs & both polarizations per station
on a single plot, with the below color-coding.

 IF1/RR = red    IF1/LL = blue      IF2/RR = cyan        IF2/LL = green
 IF3/RR = pink   IF3/LL = med.grey  IF4/RR = brown       IF4/LL = purple 
 IF5/RR = teal   IF5/LL = yellow    IF6/RR = steel blue  IF6/LL = lt.grey
 IF7/RR = black  IF7/LL = orange    IF8/RR = lt.green    IF8/LL = lt.purple

Each 2s integration is plotted as a separate point.  Each of the ampphase
plots shows baselines Ef-* and Tr-*, and is made by vector averaging over
the middle 80% of the band (no fringing applied), and have separate plots
for amp and phase.

  The two versions of the frequency-based plots (auto, cross) show different
scans from each block:  [1] = scan 9 (1153+805), [2] = scan 417 (0555+394).
The baseline plots show all baselines, and plot amp [a] and phase [p]
separately. The color coding for these plots is:  Red=RR, Blue=LL. These
plots are vector averaged over the scan.  The y-axes for the real and amp
plots self-scale per baseline, not necessarily going down to zero. The
phase scale is always +/- 200 degrees.  

  Da has signal in only IF5-6 in the geodetic mode.  As in EB039C (but unlike
in EB039B), we've flagged the Da IFs without signal prior to plotting to
avoid plot clutter ("flagged" == weights set to -|weights|).  


CONTINUUM PASSES (each parallel-hand polarization run separately):
  The Japanese stations Ym & Vm had signal only in LCP; thus
  the RCP plots show only the 9 EVN stations.

    eb039d.cont[R,L]-weight.ps            station weights vs. time
    eb039d.cont[R,L]-auto-[1,2].ps        station autocorrelation vs. freq
    eb039d.cont[R,L]-cross-[1,2][a,p].ps  baseline amp & phase vs. freq
    eb039d.cont[R,L]-ampphase-[a,p].ps    baseline amp & phase vs. time

  The polarization is given by the R or L following the "cont". The number
in [] for the auto and cross plots refers to which of two fringe-finder 
scans is shown: [1] = scan 13 (3C345), [2] = scan 401 for contR and
scan 271 for contL (both J2038+5119).

  The weight plots show all 8 IFs for the polarization per station on a
single plot, with the color coding as below:

 IF1 = red    IF2 = blue      IF3 = cyan        IF4 = green
 IF5 = pink   IF6 = med.grey  IF7 = brown       IF8 = purple 

In the continuum mode, the Da signal is in IFs 4-6 (covers IF5, most of IF4,
only a small likely not significant portion of IF6); other Da IFs have been
flagged prior to making the plots.  Each 1s integration is plotted as a
separate point.  The ampphase plots show only Ef-* baselines.  The phase(t)
plot omits all target sources (including W75), plots only IF4-5 (where the bulk
of the Da signal is; IF 4= red, IF5 = blue), and plots integration number
within this subset of data for each baseline along the x-axis rather than time
(with time used as the abscissa, all the details are hopelessly crowded
together into little vertical stripes).  One consequence of this is that gaps
in specific stations appear to get "collected" at the end; thus the final
stretch of well-determined phase for Ef-* baselines corresponds to scan 401
[J2038+5119]).  The amp(t) plot further omits the brighter fringe- finders to
avoid those dominating the y-axis scale, plotting only 2048+4310 and
J2029+4636.  All IFs are plotted (color-coding as for weight plot).

  The plots vs frequency are directly comparable to the corresponding plots
in the geodetic block, except for the scan used.  The autocorrelation
plots omit the first frequency point, to avoid the scale being thrown off
by Nt IF8/RCP, which had a real value of ~-5 throughout (and the resultant
ringing in the rest of the autocorrelation passband can be seen).


LINE PASSES (each parallel-hand polarization run separately):
    eb039d.line[R,L]-weight.ps             station weights vs. time
    eb039d.line[R,L]-auto-[1-3].ps         station autocorrelation vs. freq
    eb039d.line[R,L]-cross-[1,2][a,p].ps   baseline amp & phase vs. freq
    eb039d.line[R,L]-cross-3a.ps           baseline amp vs. freq
    eb039d.line[R,L]-ampphase-[a,p].ps     baseline phase vs. time

  The weight and ampphase plots are directly comparable to those from the
continuum pass, except here there is only one IF per pol.

  There are three scans shown in the auto plots.  The first two use the same
1-minute intervals as in the continuum-pass plots.  The autocorrelation plots
drop the last four frequency points to avoid DC spikes in some stations.  The
last one shows a maser-source scan (#260, W75) to demonstrate that the proper
IF was correlated. The cross plots show the same three scans, with the
fringe-finder scans plotting amp and phase (separately), but the maser-source
scan plotting only amp.  The plots of the maser line show the sub-set of
frequency points 440-660 in the 1024-point IF, selected to provide more detail
about the spectral shape of the emission.