>
> |
%META:TOPICINFO{author="ToddHunter" date="1201637820" format="1.0" version="1.1"}%
%META:TOPICPARENT{name="OOFWithMustang2008Jan23Meeting"}%
As promised, some results from the MUSTANG data:
The aim of this first processing was primarily to re-check that all of
the data formats are OK and that the fit converges. It turned out that
the results that came out are actually quite encouraging already, so I
am attaching some plots.
First some caveats:
- at this stage I am still assuming a tapered illumination
pattern, not a truncated one as we should. This will be simple
to implement.
- I am fitting at only one wavelength, ignoring bandwidth effects.
- I am fitting with data from the four pixels simultaneously.
- I've removed the first and last five seconds of each map since
the telescope is behaving wildly then.
The following are the three observed beam maps when processed through
the OOF software:
- obsbeam-1:
- obsbeam-2:
- obsbeam-3:
The following are the three best-fitting beam maps when fitting up to 5-th
order zernikes:
- bestfitbeams-0:
- bestfitbeams-1:
- bestfitbeams-2:
And the following is the corresponding phase map:
- aperture-phase:
One can already note some interesting points from these:
- The phase map has a familiar structure -- we have seen this sort of deformation at q-band before. Encouraging!
- Observed maps look good but stripes in the scan direction (i.e., baseline structure) are visible. We are probably limited by this effect at this stage.
- The best-fit maps are reproduce a lot of the structure in the observed maps. However, it is clear that from the best-fitting maps that we are fitting for some of the baseline structure present in the observation. For example, look at the long elongated structure in bestfitbeams-2.png
So I think one of our focuses should be optimising the pre-processing of the MUSTANG data to minimise baseline structure. Having a box scanning strategy where there are several different scan directions would also help I think (in a way analogous to the Emerson deconvolution
algorithm).
Best,
Bojan
%META:FILEATTACHMENT{name="obsbeam-1.png" attr="" comment="obsbeam-1" date="1201636926" path="obsbeam-1.png" size="9291" user="ToddHunter" version="1.1"}%
%META:FILEATTACHMENT{name="obsbeam-2.png" attr="" comment="obsbeam-2" date="1201636944" path="obsbeam-2.png" size="8758" user="ToddHunter" version="1.1"}%
%META:FILEATTACHMENT{name="obsbeam-3.png" attr="" comment="obsbeam-3" date="1201636957" path="obsbeam-3.png" size="8622" user="ToddHunter" version="1.1"}%
%META:FILEATTACHMENT{name="bestfitbeams-0.png" attr="" comment="bestfitbeams-0" date="1201636979" path="bestfitbeams-0.png" size="8175" user="ToddHunter" version="1.1"}%
%META:FILEATTACHMENT{name="bestfitbeams-1.png" attr="" comment="bestfitbeams-1" date="1201636993" path="bestfitbeams-1.png" size="7451" user="ToddHunter" version="1.1"}%
%META:FILEATTACHMENT{name="bestfitbeams-2.png" attr="" comment="bestfitbeams-2" date="1201637007" path="bestfitbeams-2.png" size="7560" user="ToddHunter" version="1.1"}%
%META:FILEATTACHMENT{name="aperture-phase.png" attr="" comment="aperture-phase" date="1201637029" path="aperture-phase.png" size="25071" user="ToddHunter" version="1.1"}% |