optimize sat spot centroiding defaults for Y/J band

[dsavransky]

I still believe that the parameters can be slightly more optimized for Y/J band (mainly adjusting the "mask_radius" parameter so that it is wavelength dependent).

[semaphoreP]

Note that this is for pol mode centering. Adding Pauline since she's worked on this too and I think fixed it for HR 4796

[mperrin]

Legacy issue migrated from redmine issue tracker. Originally filed by @kfollette on 2016-03-21

[dsavransky]

Recovering rm journals for issue:

kfollette: See bug #521 (closed) for more info on this issue

jasonwang: We should revisit this bug because this is a pol star centering issue, unlike bug #521, which is a spec mode issue. We had issues in the last LLP run not being able to find the star center in J band, because the default 'mask_radius' value needs to change as a function of wavelength.

Should be push this issue to 1.5+?

jasonwang: Oh woops, this bug wasn't closed. Need to read more closely..