Bad (dead/hot) towers.
In the CEM, CHA, WHA there are two PMTs. Towers marked bad in the relevant database tables can then be improved by ignoring the bad PMT. In the PEM, PHA this cannot be done (although from 4.11.2 hot towers are removed by looking for isolated high-energy spikes).
Before 4.11.2 nothing was done about dead/hot towers.
From 4.11.2 onwards and the XXXPMTDeadChannels databases (PASS 8 (CEM, WHA)/ PASS 9 (CHA)), the runs that were bad due to dead towers are made good. However, due to some confusion about what the tables contained, hot towers were made worse (basically they were treated as if the hot PMT was a dead PMT).
The runs containing hot towers have been marked bad by Kostas, from his webpage:
1. Dec. 4, 2003: Re-clasified 10 runs (~ 2.4 pb-1) with hot central towers (CEM, CHA, or WHA) as BAD for the DQM good run list v4 (used for data processed with calorimetry code versions up to 5.2.*). Details in the "Results" page.
In adition, the tables contained both 0/1 and 1/2 as definitions of dead/hot. The tables have been remade as
XXXPMTBadChannels and these have been put in the updated PASS 12.
The code in development treats both dead and hot towers correctly and will be tagged into the next release.
Summary:
Before 4.11.2: dead/hot towers not corrected.
4.11.2 onwards: dead OK, hot bad (labelled bad runs).
Next release & PASS 12: both corrected.