Transitioning from Channel Access (CA) to PV Access (PVA)

[fredericpoitevin]

We currently rely on Channel Access (CA) to our YAG cameras, which sends full arrays through sometimes low-bandwitch network channels (in particular XRT cameras), which we then use for simple analysis like centroid finding.

@silkenelson and @patoppermann mentioned last week that our cameras offer a more modern PV Access interface, which should be faster but also enable to do simple analysis upstream - namely allowing to read e.g. centroids directly as PVs.

Use cases I can think of:

• transfocator scan: plot beam position and size on the fly as transfocator stage moves.
• beam pointing: use beam position relative to cross-hairs for alignment (currently using lcls_tools and CA).

@lg345 do you think of more use cases?

[silkenelson]

FWIW, the calculation PVs are available now and do not need switching to PVA.

[fredericpoitevin]

Thanks Silke, I should have clarified that I was referring specifically to the hutch python code, e.g. https://github.com/pcdshub/mfx/blob/222-prepare-for-exafs/mfx/optimize/beamline_hw.py#L60