BCAL hit/shower times shifted in simulation

[sdobbs]

In simulation, the reconstructed times for DBCALPoints and DBCALShowers are found to be shifted from their truth values by ~150 ps and ~300-400 ps, respectively.

The results of this shift are reflected in the beta-vs-p and delta T distributions used for PID, and are seen in many places, e.g. page 5 of [http://argus.phys.uregina.ca/cgi-bin/private/DocDB/ShowDocument?docid=2902](this presentation) and the attached figures.

new_BCAL_BetaVsP.pdf
new_sh_delta_t.pdf
hits_delta_t.pdf

Similar problems are not seen in data, e.g., https://halldweb.jlab.org/work/halld/data_monitoring/RunPeriod-2015-03/ver24/Run003185/HistMacro_BCALReconstruction_p3.png

[sdobbs]

Quoting an email from @beattite

Okay, I just ran a bggen sample, then the plugin on that sample. The
delta t distribution centered around 0.14 or so.

Then, I updated to the newest version of sim-recon (including Will's
changes to DBCALPoint) and re-ran a bggen sample, then ran the plugin
over the new bggen sample. The delta t distribution centered around
0.22 or so.

This was what I expected when we made the change a while back. The
simulation using Georgios' numbers and reconstruction not using those
numbers would offset the delta t negatively, so fixing the
reconstruction to also use Georgios' numbers makes the delta t
distribution problem worse.

[markdalton]

Thanks for binging this to our attention. We’ll figure out what’s going on.

Best,
Mark Macrae Dalton

On Mar 15, 2016, at 11:25 PM, Sean Dobbs notifications@github.com wrote:

Quoting an email from @beattite https://github.com/beattite
Okay, I just ran a bggen sample, then the plugin on that sample. The
delta t distribution centered around 0.14 or so.

Then, I updated to the newest version of sim-recon (including Will's
changes to DBCALPoint) and re-ran a bggen sample, then ran the plugin
over the new bggen sample. The delta t distribution centered around
0.22 or so.

This was what I expected when we made the change a while back. The
simulation using Georgios' numbers and reconstruction not using those
numbers would offset the delta t negatively, so fixing the
reconstruction to also use Georgios' numbers makes the delta t
distribution problem worse.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub #193 (comment)