Comparison of the acceptance of FMS (black) and the proposed preshower (red).
The bottom of the vertical scintillators is shorter than the rest and will reduce the acceptance in this area (10 cm in front of the FMS, 17.4 cm shorter). Here, we also exclude four scintillators due to the beampipe support.
The dashed line excludes one full row of towers in the FMS (outer and inner edge). This is conservative, probably half a tower is good enough for the clustering.
For transverse asymmetries, the acceptance for cos(phi)>0.5 is most important. In this range, the additional loss of acceptance of the preshower is very small (phi < -1.05 || phi > 4.19).