This is the plan of changing pedestals for selected 8 APVs for disc #1.

For every set only APVs listed in given column should have 'low' peds, all remaining 32 APVs should peds 'high'.

The cartoon below table visualises which APVs wil be changed for set 1 (violet cicrcles) . For subsequent sets the the circle travel by 1 toward IP.

The method:
* I hacked fgrRawDaqReader and tapped directly in to FGT data block is unpacker, added 240 1D ADC histos for each APV,
labeled histos by  RDO_ARM_APV, read first 200 events from each file.
There was no DB mapping involved, the order of ADCs in the data  block defined the mapping (and RTS reader which is use to crawl through FGT data block - I did not touched that)

* from each run I got 240 1D ADC plots for each APV
* for each assembly (1AB, ...) there is 10 APVs x 7 data sets

Fig 2 shows example of ADC spectra for 3 APVs and 3 data sets.
The 1st row is for the set0, where all APVs had pedestal 'high' , I show only ADC range 0-1500.
The 2nd row shows the difference between set 1, for which ped in APV 0 was lowered - you see the large wiggle. This makes the change in ped well visible. For APV 1,2 the difference is almost zero. Y axis range is fixed to +/- 1500 ADC.
The 3rd row shows set 2, for which ped in APV 1 where lowered, and APV0 are back 'high'. The wiggle is now visible for APV1.
You get the pattern.

Fig 3 shows all 7 data sets, for assembly 1AB. I marked wiggles 'ok' if they show where they should.

----------
(most likely) conclusions.

For set 4  the APV were change as intended for set 5  - I assume operator error.
For set 5 no APVs were changed and it is identical to set 0 - I assume operator error.
Set 6,13017059, with all APVs 'low' did work, so we can address all APVs.

Fig 2. Example of base line (set0) and 2 other sets, from which baseline was subtracted. The wave pattern marks APVs for which pedestals has changed

Fig 3. Full scan of assembly 1AB, for set5 none of APVs had lowered pedestals.