[REMI][CORRECT] Add support for quad anodes

Description

SQS has recently began commissioning of one of its remaining components in the form of a soft x-ray spectrometer (which I will call AQSX for simplicity). Originally a multi-delay line detector from Surface Concepts was meant to be used here (the detector commonly called MCP or DLD in DET), but this development has not materialized yet and depending on whom you ask, may not in the short to mid term. As a stop gap solution, the company has provided SQS with a conventional quad DLD detector.

As commercial detectors these come with closed system readout electronics to provide the user with fully pre-processed and corrected data, but for several reasons integrating this did not work out well. During the commissioning some weeks ago I hence adapted the DLD online analysis for REMI for this purpose (reminder: REMI employs two hexagonal DLDs) which worked out quite well. Rather than using the vendor readout chain and as with REMI, here we directly record the raw signals via digitizers, leaving all signal interpretation to software.

We decided to stick with this solution for the upcoming community proposal next week, but naturally also extend this to offline analysis. This MR contains the changes necessary to the REMI notebook itself, which are fairly minimal. This of course hides some accompanying changes necessary in EXtra-remi, but these are also quite small with the majority coming from almost copy-and-pasted plotting code. As I already mentioned in the past REMI MRs, I want to reduce EXtra-remi to the bindings to RoentDek's libResort anyway. This can be coupled with renaming the REMI detector to DLD in general given its broader use. There is actually nothing specific to REMI in it, except possibly using terms such as "fish" in the report.

I have yet to confirm with RoentDek whether libResort can handle calibration of quad anodes (not calibration as in corrections, but think detector characterization) or we'll have to implement that part ourselves. As can be seen on the screenshots below, the wire dispersion (how skewed the lines in sum-over-pos are) is fairly minimal and we can run without it for now. Also given there are two layers instead of three, the scale factors can easily be determined manually (they translate time to spatial distance).

Given the beamtime's next week, this needs deployment until Friday the latest. Sorry!

How Has This Been Tested?

run = 12
in_folder = '/gpfs/exfel/exp/SQS/202222/p003490/raw'
calib_config_path = '/gpfs/exfel/exp/SQS/202222/p003490/usr/dld_proc_config.yaml'
quad_anode = True
ignore_ppl = True

Relevant Documents (optional)

Types of changes

• New feature (non-breaking change which adds functionality)

Reviewers

@ahmedk @baumannt @kluyvert