The dE/dx calibration for Run XVI dAu 20, 39, and 62 GeV data samples is almost finished.
The bad run list contains: 17152012,17153058,17155032,17159050,17159056. For these run <dE/dx> is divated more than 2.5% from average. 
"Almost" means that calibration should be put in MySQL (waiting for new table) and be checked with dev version of library (with new corrections).
The check in dev has been done. See plot from RunXVI316.
The resolution:

The resolution obtained with dev.

The track length dE/dx dependence for dN/dx. Tail for length > 120 cm is due to lack of statistics.

Pad correction (Florian's dE/dx dependence for &phi; for sector 9, 8, 6, and 5)
Before pad correction (y = pad (normolzed from [1,Npads] => [-1,1], x is normolized pad row (Inner sector : x < 0, Outer sector : x > 0)

After pad correction

dE/dx versus &phi& BEFORE correction

dE/dx versus &phi; AFTER correction

Full list of plots dor dE/dx &phi; dependence can be found at http://www.usatlas.bnl.gov/~fisyak/star/Tpc/dEdxPhi/index.php

IFit 

dN/dx