My cuts were:

0.8 GeV < W < 1.15 GeV
|p_miss_par| < 0.25 GeV
p_miss_perp < 0.15 GeV
ps >500
|x|<0.5m
|vz|<0.2m
|(p_optics - p_sh)/p_mag| < 0.3

Vetos required projected position outside of 0.35m and projected time
difference (t_diff = veto_time - clus_time): -40ns < t_diff < 10ns for
uncharged particles.
Vetos required projected position inside 0.35m and projected time
difference -10ns < t_diff < 10ns for charged particles.
Clusters within 10ns of each other were combined into a single cluster.
All clusters that passed the above cuts were considered valid in both
the regular and background analysis.  This allowed for multiple clusters
to be found, but would remove extraneous clusters with the background
subtraction.