STATUS OF RICH RECONSTRUCTION
      -----------------------------
          P.Schiavon - 01/12/13




      This study has been performed using the run 13127.
      Among ~1.600.000 events only ~550.000 were usable (correct
      geometrical distributions of tracks; something wrong after 1/3
      of the run; no records on logbook); 80.000 were with at least
      one track reconstructed with momentum;
      the sample after the cuts (see below) consisted of ~3000 events.

      As known, the RICH photon chambers suffered from electrical
      instabilities; the radiator gas was a mixture of C4F10 and N2
      of not exactly known percentage (=>refractive index);
      the chamber memory (for beam halo muons) is large (see MC simulation); 
      moreover, with reference to ring reconstruction, most of the tracks
      were at small angles (i.e. in the beam region).

   *  File cathode.ps shows the 'photon' (after clusterization) 
      distribution on the chambers (cathodes), integrated over all the
      events; well visible the distribution of the photons (essentially
      beam halo; beam tracks are not processed), with different intensities
      according to the chamber response; the white bands are dead zones.

      The aim of this analysis is to sort out some signal to show that the
      RICH is able to see Cherenkov photons; to this purpose, severe cuts
      are used on the track sample to select regions with less background:
        track momentum > 60. GeV/c (relativistic particles = all rings of 
        maximum radius);
        track crossings on RICH entrance window at a distance from centre 
        larger than 200. mm (tracks far from beam region);
        divergent tracks in the vertical plane;
        Saleve side lower chambers;
        number of 'photons' per reconstructed ring >= 4 and < 30.
   *  File tracks.ps contains three pictures: the first one shows, after cuts,
      the track impacts on the RICH entrance window; the second, the 
      particle 'impacts' (as if reflected by the mirrors) on the detectors;
      the third one, the photons (clusters).

      After the Rough transformation to the particle direction, each
      cluster on the detectors becames a 'photon', with polar angles theta 
   *  and phi; the file theta.ps (1st) shows the distribution of the theta
      angle (mrad) for all the selected photons; (2nd) the distribution of
      the reconstructed 'ring' theta angle; (3rd) the distribution of the
      difference of theta-photon and theta-ring for all rings; a signal
      peak is significantly visible in (2nd); the peak is around the value
      of 45. mrad, roughly expected from the estimated refractive index
      of the radiator gas mixture; the distribution in (3rd) exhibits a
      badly shaped gaussian, much broader than expected.
   *  To make the comparison, file mc.ps shows the corresponding distributions
      for the MC simulation, for the same setup; striking the much
      narrower distributions.
 
      File scram.ps shows the two distributions (1st) and (2nd) with 
      the background superimposed; the background has been evaluated 
      processing the RICH chamber info with the tracks of a previous event
      (random coupling); the background shape fits well the background
      under the peaks, strenghtening the hypothesis of a physical signal.

   *  File mom.ps shows the selected momentum distribution (1st) and the
      reconstructed ring photon multiplicity (2nd).
   *  File n-1.ps shows the reconstructed refractive index; the value
      obtained (n = 1.00103) is compatible with a mixture of 55% of
      C4F10 and 45% of Nitrogen, sligthly different from the measured 
      one (50:50), assuming n = 1.00153 for C4F10.

      File rings.ps shows three typical ring displays, both as a general view
      and a more detailed one; the crosses are the particle 'impacts';
      in the detailed view, grey dots are clusters and black dots are the
      clusters defining the ring; there is no pulse height presentation;
      the photon multiplicity and the particle momentum are indicated;
      the likelihood values are meaningless; the histogram in the upper
      part is that of theta-photon for the track, the basis of the ring 
      selection.