• From previous eye-scans it is known that the cluster-finding efficiency is close to 100%. We lose or have too much tracks because of problems located mainly on the tracking level.
• We do not have a really credible simulation for low level tracking effects so far. That is why it was decided not to have correction for tracking efficiency but only for geometrical acceptance.
• It means that we must not lose any good primary tracks but reject as much as possible background ones. Here with the word 'background' I mean the following cases:
  • tracks of secondary particles if they can be distingushed from the primary ones, e.g. using the actual/potential point ratios
  • split tracks have to be handled carefully, after locating them they should enter the analysis only once, e.g. we should use the longest piece of their segments or combine them into one
  • muons coming parallel with the beam, can be rejected based upon their big impact parameters
  • etc.
• While the above problems may be on the percent level, you could naively think that they will not really harm the analysis. However they enter in the later stage when using dE/dx, where at present we need an almost perfect description of the shape of the energy loss distribution, with small systematics errors. The most sensitive particles are the less abundant ones: antiprotons, but kaons are affected as well.

In order to fulfill the above requirements a very careful
study is needed on the single track level.