Hi,
trying to explore the problems in tpc reconstruction, I have explored the results from the tpc riemann before and after the kalman fit. In particular I have plotted, for eta_c events:
cbmsim.Draw("TrackPreFit.getMom().Mag()>>pre(100,0,2)");
cbmsim.Draw("TrackPostFit.getMom().Mag()>>post(100,0,2)");
The following is the result:

It is not clear to me why so many tracks are shifted below 0.1 GeV, and they are cut away from the PidCorrelator (which has a cut 0.1 < p < 10 GeV/c). This could cause the problems with low efficiency.
Could you please investigate about? They are for sure screwed. The final distribution with TrackPostFitComplete is quite strange, at the end.