The weird square thingy...

Right. So finally I asked somebody about the weird square thingy. We discussed it, and agreed it couldn't be a problem with clean. So he suggested I go back to the visibilities.

So I went back to the U-V plane, to look at the data. Yup. There it is. A couple of U-V visibilities needed to be flagged. No more weird square in two channels.

But all of this just points out that we darn well better invent autoflagging before the new correlator comes completely on line. It's one thing to look at data from 256 channels. It's a completely other thing to look at data from 100,000 channels. No one will use the thing if they have to find 'bad' data by hand.

So that's my next task. To figure out... wait for it... where the group is with the autoflagging. Because no one can remember.