[Halld-offline] bug in EM bg generation in hdgeant

[Richard Jones]

Dear colleagues,

After chatting with Sean about the apparent mismatch between hdgeant output
with EM background turned on and accidental backgrounds seen in the various
detectors for real data runs at high intensity, I went back and had a
closer look at what might be wrong. There I found a very old formula that I
put in back in 2006 to approximate the background rates over a wide range
of beam energy cutoff values. That formula is really not valid any more, as
it is specific to the 3.4mm collimator, 12.0 GeV endpoint, 9.0GeV edge
position, 20 micron diamond, and a handful of other assumptions that were
considered the standard GlueX configuration in those days. Nowadays things
have changed. As a result, I have updated a bunch of things in the
simulation to make it adaptable to our present configuration, and give the
user more control over the details. In the process, I also found a bug that
made running with bg enabled extremely slow -- and also give incorrect
results at high beam current and/or low beamEmin settings.

The changes that fix everything and make it run efficiently with bg turned
on are now in the master branches of sim-recon (hdgeant) and HDGeant4 (push
request has been posted, not acted on yet). I have updated the control.in
file with new features related the bg generation, and updated comments that
explain how to use them. Please update your working copies of control.in
and merge in these changes. They are very important, even if your only use
for the bg generation is to give an accurate simulation of the accidentals
in the tagger.

If there are any questions, please contact me.

-Richard Jones
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20170305/1333a142/attachment.html>