It may be a silly point, but in split_IMA the size of the reference pixels regions and the "science" regions are hard-coded for a full-frame case. It might be that in some cases users use a subarray. E.g program 15146 (PI Riess) does exactly that, even though for few reads and in a rapide sequence, thus within their exp time there is not much drift, even when going under gyro...but yet, it might be worth generalizing the code?
It may be a silly point, but in split_IMA the size of the reference pixels regions and the "science" regions are hard-coded for a full-frame case. It might be that in some cases users use a subarray. E.g program 15146 (PI Riess) does exactly that, even though for few reads and in a rapide sequence, thus within their exp time there is not much drift, even when going under gyro...but yet, it might be worth generalizing the code?