StPxlDbMaker has been updated for y2013 Geometry and seems to be working properly. A comparison of the transformation matrices produced using root geometry and geometry tables shows agreement up to 5th decimal place:

A full list of the transformation matrices for the 3 sectors (1, 5 and 10) can be found in my personal folder, following the path "/star/u/mlomnitz/HFT/Mustafa/GMatrices.txt".

On another note, I have run StPxlFastSim.cxx assuming there is no smearing of the pixel hits to prove that we recover the same positions after transforming from the global positions (created using Montecarlo sim) to the local positions and back to the global coordinate system.

When comparing the 5th and 6th line, we see that the x and y coordinates are recovered without any errors. The discrepancy in the z component arises from an a mistake in StPxlFastSim.cxx:

Since there is no defined smearing resolution for the detector in the z direction, the code takes the local position to be equal to the global position yielding the observed error.

A full list of the obtained Pixel hits can be found in:
-Assuming no smearing"/star/u/mlomnitz/HFT/Mustafa/GMatrixCompareNoSmearing.txt"
-Using the proper x and y resolution for the detectors "/star/u/mlomnitz/HFT/Mustafa/GMatrixCompareSmearing.txt"

Next I will try to clean up the code in StPxlDbMaker.cxx and automate LoopOverPxlInner.C so that it automatically places zeroes in the positions for missing sectors and no longer requires manually inputing such in the tables.