S. Corder
Holo report, 2007

10m antennas:

3mm: The slight improvement in resolution afforded by the brightness
     of mars allows us to easily see the feed legs in a number of 10m
     antennas (these are 21 by 21 patterns).  I have NOT been able to
     average all the baselines and bands (my pipeline seems happy to
     do the averaging on the 6m antennas for some reason I can't find
     yet).  I am SURE that when averaging the signal from the 9 reference
     antennas and six bands (that's a sn improvement of 7).  We also 
     need to use a new feature Mel is developing in MIRIAD to improve
     the amplitude calibration on long baselines.  I will re-run the 
     pipeline when I have Mel's new improvements and if the averaging
     doesn't work then I will make more of an effort to figure out
     why.  A "typical" image of each dish are posted on the web at:
     cedarflat.mmarray.org/holography/1mm or /3mm

C1: Large, edge of dish surface deviations exist and are repeatable on
    difference reference antennas and different runs.  There is also a
    ~50um deviation very near the peak of the illumination pattern but
    for the large part things seem okay here. Surface rms is 25-30um
    (weighted by amplitude) after the removal of focus and pointing
    errors.

C2: Doing a little less okay with surface rms (post-fit, weighted by
    amplitude) in the ball park of 35-40um.  There are repeatible and
    rather ugly surface deviations here, three in high power points of
    the dish.  This antenna is rather well centered.

C3: I feel like there should be a drumroll.  This antenna had so little
    available information from before that I was scared to look at the
    holograms.  As it turns out, the illumination patterns are pretty
    well centered and the surface errors are not incredibly large.  
    The surface errors are on the order of 25-30um.  There are a few
    regions of deviant surface.

C4: Illumination pattern is not all that well centered but Dave has
    adjusted this antenna TWICE so I don't think it is getting any
    better (and it IS better than it was April of last year).  There
    are minor surface deviations reported with the post-fit, weighted
    rms ~30um...but in general the deviations aren't repeatable day
    to day so I think this might be our best antenna of the 10m aside
    from the illumination offset which, like I said, I don't think will
    improve much.

C5: The surface rms here is on the order of 35um.  The illumination
    pattern is well centered.  There are some repeatible surface
    deviations but the two sessions do not show the same size of
    deviation (they agree in direction which is good).

C6: Not available for the big 3mm runs.  I have highly oversampled
    (in the sky plane) images but these don't go out far enough to
    talk about much more than "is it centered in illumination".

6m antennas:

C7: This antenna has been having trouble with phase stability so 
    I'm not sure how believable the hologram is.  The illumination
    pattern, for example, seems pretty offset for a 6m antenna. 
    I'm not sure how real this is.  The surface error reports
    anywhere from 28-49um.  I don't know what to believe.  The
    average image is shown here because the averaging works for
    the 6m..but if the results are so discrepant antenna to antenna
    averaging here (for c7 at least) may not be meaningful.

C8: The surface rms (post-fit, amplitude weighted) is in the range 20
    to 30um and I believe the averaging is representative.  The average
    is shown.  Nothing too serious.  This antenna is pretty well
    behaved.  There may be maybe two sections that could use a little
    adjustment but frankly it couldn't much better.

C9: Again, I see variation from 25um to 40um in terms of antenna by
    antenna or band to band.  This is NOT a shadowing effect as I have
    flagged all the shadowed baselines.  There are a few regions that
    could be adjusted.  We did NOT adjust this antenna last year as
    it was on the borderline of needing it.  The images share a number
    of characteristics and it will be interesting to see if this similarity
    persists to an even greater degree when I can correctly account for the
    planet model.  

C10: Looks like about 30-40um.  For the most part, there is a residual
     focus term at the center that doesn't average out well...this may 
     just be an issue with Mars though the "after correction" map of
     last year looks remarkably similar to the Mars data from this year
     indicating that I am doing things marginally consistent.  There
     are two regions that could be adjusted with some expectation of
     improvement.

C11: Looks like about 25-35um.  There are two regions that would VERY likely
     improve the surface accuracy.  Most of this gain would be in the 
     inner region where we were not allowed to make adjustments before.
     There is also a negative region which would be nice to fix.

C12: Looks like 35-50um....i.e. not particularly great.  This antenna
     was our big problem child from the fall.  Comparing the post-correction
     data to our new data there aren't that many similarities; however, this
     antenna was changed ALOT before so it may have suffered from more 
     relaxation.  It needs some work, most notably in the middle of the
     dish but there are residual variations around the entire dish.  The
     residuals at the edge of the dish are not surprising as we did not
     correct for the cos(pitch angle) when calculating our corrections,
     i.e. we largely under-corrected relative to our model.  

C13: Looks like 25-35um.  Pretty good.  There are some edge deviations that
     might be real but I want to compare more closely with the old data
     and the new data with the better amplitude corrections/mars model.

C14: I can't agree with Misty.  I see surface errors of 20-30um on
     this antenna.  There are a couple of little regions that might
     benefit from a small adjustment...however, it isn't a big deal.
     One this I DID notice on this antenna was that I would point at
     the beginning and then the "on source" integrations would drop
     pretty dramatically by 20-50% depending on the cross
     correlation...it looks like a massive pointing drift.  Now, to be
     sure, this sort of rapid pointing offset change is not taken into
     account by my code.  However, I saw this drift on EVERY dataset,
     i.e. both mars datasets I have (one only 10m) show this behavior.
     It doesn't look SO MUCH like a pointing drift because it drops
     but it doesn't continue to drop to zero...it levels off...almost
     looks more like decay of lock or something.

C15: 25-35um surface error.  The residuals largely agree with
     residuals after correction (from last year) with a little
     amplification that might have something to do relaxation or with
     the mars model not being accounted for here.  There is still
     a large residual error in the heart of the dish that the focus
     subtraction doesn't take out.  This was present both now and
     before.

----------------------------------------------------------------------
1mm: Okay, so you said not to but I did it anyway.  We have FANTASTIC
     weather and no science to run so I took 1mm grid like beam patterns.
     These are over-sampled in the sky plane but the data quality is
     sufficient to really give an "okay" aperture plane image.  The 
     results are rather interesting so I provide them here.  I'll provide
     a more detailed report folding in the "cross" data later but my
     pipeline, while calibrating the cross data FINE, doesn't do much in
     terms of imagining and gauss fits.
     
10m:

C1+3+4+6:  All pretty off-centered and all seemingly astigmatic but the
     aperture plane is poor.  See images. 
C2+5 were not available.

C7: astig
C8: fine
C9: a little oddly off centered and some higher order junk going on..
    but not in the middle of the beam.
C10: Some stigmatism but not exactly astig present.
C11: offset centered a bit more than expected but not aweful...marginally
     astig.
C12: Not good.  Centered but quasi-serious astig visible.  
C13: Astig.
C14: More offset than other bima antennas.  Surface, from what I can tell
     with this limited resolution, looks "okay".  Consistent with higher
     resultion 3mm data.  Not so sure why it appears more off-center.
C15: Largely okay.  Reasonably well centered, some junk in the beam (worse
     than 14) but nothing in the core...but again, resolution can mask
     alot.

that's it.  More to come later but wanted to give you a report of some
kind before Monday.  Check the images.  The color bar indicates the
amplitude sacle and the contour intervals are given at the bottom of
the image.  I have checked and the information is all visible on screen.
(None of this "the type is too small business!)  ALL images are aperture
plane with "off dish" points masked.

Stuartt