Problem(s) Encountered:

When trying to subarray 4K with the scripts (roi1k, roi512), we couldn''t take data with prospero.    An exposure would start, but it would seem to hang, and then (after a long time) it wold report that the exposure had completed (nothing was written to disk) and then it would report that the observation had been aborted (which we hadn''t tried to do).  Things worked ok in the larger formats.  But as a result, we had longer than desired readout times during our frequent focus checks.

Solutions:

This appears to be a bug in the code.  Currently it is unclear what causes issues when subarraying to 1K or less.  Be advised that 1K ROIs can be performed when using OSMOS.