Relationship of "roll" angle to platform encoder values

[Matthew Hasselfield]

The "roll" is the 3rd angle describing the position and orientation of the focal plane center on the sky, along with az and el in horizon coordinates. Positive roll is clockwise, as projected onto the sky (right-hand rotation about the vector from observer to sky).

In the SATP, +ve boresight rotation causes the focal plane to rotate CCW, as projected on the sky. Similarly, in the LAT, +ve corotator movement causes the focal plane to rotate CCW, as projected on the sky (see Confluence "Focal Plane" notes).

So, the appropriately signed "roll" of the focal plane should be computed as:

  roll = -boresight_encoder_value  # SATP
  roll = el_encoder - 60 - corotator_encoder # LAT

This leads to some changes needed in a few places in sotodlib.

• Book loading, for SATP.
• In the docs PR, where roll_* are incorrect (SATP) or vague (LATP).
• In update-obsdb (@AERAdler), "roll_center" and "roll_throw" should be computed as described above. (Current code has "boresight_center", "boresight_throw" -- ok to have that, for SATP; need corotator_center/throw for LAT; but please add the roll values as defined above.
• Possibly in the instrument model and sims? @keskitalo, @smsimon Instrument model doesn't care about "roll", I know -- but the question is whether positive boresight / corotator angle corresponds to CCW motion of the focal plane as projected on the sky.

[Alexandre Adler]

Thanks for doing the math Matthew. Do you know the keyword denoting the corotator angles for the LAT? Would it be corot_enc or something to that effect?

[Matthew Hasselfield]

Thanks for doing the math Matthew. Do you know the keyword denoting the corotator angles for the LAT? Would it be corot_enc or something to that effect?

A very timely question... please prepare to use "corotator_enc". However, that's not currently what the bookbinder / agent are set up for. See new issue, #578