The tests were done comparing the momentum resolution of two venus runs produced for mock data challenge 2. Mdc 2 data is not very suitable for FTPC testing as it was calculated at extremely high geant resolution, so that each track caused several hits in each padrow. However, all the available data at half field is from mdc 2.
One event of each run was processed through the fast simulation chain, assuming that the quality of ExB corrections will not change with the field strength. Fast simulation makes it possible to compare reconstructed tracks to geant tracks by simply comparing the constituent points.
The plots show the reconstructed momentum divided by the geant momentum and are in good agreement with earlier studies done with realistic simulation parameters and the measured magnet field. (Earlier simulations done by Michael Konrad assuming a perfectly uniform field looked somewhat better.) Only perfectly reconstructed tracks with 10 hits that actually belonged to the same geant track are used in the plot.
The first plot shows the resolution at full magnetic field, the second at half field. Both peaks are nicely centered around one, showing essentially correct momentum resolution, but the RMS of the distribution increases from 15 to 20 percent when going to half the magnetic field. This is in contradiction with the obvious assumption of a linear increase of the errors, which, however, is not really to be expected at closer inspection. Also, the number of properly reconstructed tracks is smaller in the second plot, but it is yet unclear if this is due to a larger range of delta spirals in the smaller field, to some other effect or just statistics.