...
The following calibrations steps are relevant:
A) ideal gains (ET=60 geV @ ADC=4067)
B) relative gains from slope ration within eta bin (Joe's blog)
C) Matt computed E/P for pp500 isolated electrons using A*B as input
D) bad towers masked as bad in DB
E) ~60 towers with cross fibers and other complications were isolated by Joe & Matt : list , Matt explained more in e-mail
...