SD from simanalyze and OTF differ too much

[teuben]

For the skymodel simulation at 115 GHz, the TP beam is nominally 57.06".
Simanalyze produces a gmc_120L.sd.image file, but this differs from the imsmooth produced OTF map in total flux by about 7% and more disturbingly the difference map has a 3 banded horizontal structure.

In skymodel-b.fits the region1 has 6644 Jy. (see also skymodel.md)
The OTF has around 6402 Jy (slighty dependent on the exact beam)
The SD map from simanalyze has 6035 Jy.

[teuben]

there is sample code in data/issue14.py that will show, sorry, it's also using QAC.