[Halld-offline] too many photons reconstructed in b1pi

[Beni Zihlmann]

Hi All,
I noticed in the b1pi analysis code that get run Mondays and Thursdays that
the number of reconstructed photons jumped by almost 20% between
November 14th and November 21st. November 17th failed in both OSes.
see second last row middle plot:
https://halldweb1.jlab.org/b1pi/2011-11-14/Linux_CentOS5-x86_64-gcc4.1.2/

and compare to:
https://halldweb1.jlab.org/b1pi/2011-11-21/Linux_CentOS5-x86_64-gcc4.1.2/

I can not see what might have caused this. Any ideas?

cheers,
Beni