GPItv sat spots finder seems less robust than it should be

[dsavransky]

There doesn't seem to be a way to invoke the 'constrain' option in GPI? It's getting quite confused on sat spot finding for plenty of images. Admittedly not of the best contrast.

We should make the gpitv sat spot finding use the same options as the pipeline now does with the constrain flag.

Examples: check some files around S20151105S0045 or so.

[dsavransky]

Recovering rm journals for issue:

mperrin: OK, so there actually is an option for constraining the sat spots, but even with that turned on, does not work to find some of them in GPITV contrast display.

S20151105S0064 is an example of this.

On the other hand for S0090 on the same night, the pipeline contrast is clearly bogus (1e-5) but the gpitv display gives a more reasonable 1e-4.

kfollette: Setting target version to 1.5. Maybe a project for hack day?