oop_offset2

A note about the OOP offset 03/16/06 (rev. 03/21/06)

Previously, it was thought that my parameterization of dssxptar reduced to the usual ssxptar offset of 0.0032 rad (w.r.t. the HMS) at small P_SOS. The plot below shows my parameterization of the sos oop offset after making events coplanar with the HMS and with an hms oop central offset of 1.1 mrad.

My results are of similar magnitude as the usual ssxptar offset, but has the wrong sign. The usual ssxptar offset is positive and is applied in c_physics.f and s_physics.f

ssxp_tar_temp = ssxp_tar + s_oopcentral_offset

My parameterization causes an offset in the opposite direction!

ssxp_tar_temp = ssxp_tar_temp
& - (0.00561+0.0000331*exp(abs(spcentral)**2.959)) !3/16/06

I can not see where I went wrong in the Hee'p analysis. I checked my parameterization by changing s_physics.f

ssxp_tar = ssxp_tar
& - (0.00561+0.0000331*exp(abs(spcentral)**2.959)) !3/16/06

By simply removing the "temp", the corrected ssxptar is written to the ntuple. One can then look at a plot of 'ssp*ssxptar+hsp*hsxptar' to make sure the peak is centered on zero. This was seen with my choice of the oop offset.
plots (bottom right plot in each panel)
This could also be checked very simply by plotting 'ssp*ssxptar+hsp*hsxptar' when no out-of-plane offsets are written to the ntuple. The sos momentum, ssp, was corrected for saturation.

As can be seen, ssxtar without any offset (LHS) produces a peak at 0.00295 GeV. When (ssxptar-0.0035) is used instead (RHS), the peak is centered at -0.00007 GeV. This just shows that I included the sign of the offset correctly in the analyzer. I actually use the reconstructed phi angles to determine the ssxptar offset, which is independent of ssp and hsp.

The plot on the right shows that the SOS is sitting above the target. Consider an elastic event that is completely horizontal. If the SOS is sitting below the target, then the event will be reconstructed at negative ssxptar (the x axis points down). To make the event look coplanar with the particle in the HMS, a positive offset will need to be added to the reconstructed ssxptar. What I see is that a negative offset needs to be added to the reconstructed ssxptar to make the event look coplanar.