Re: genr8, HDgeant and photons

[Matthew Shepherd <mashephe@indiana.edu>]

Hi Matt,

Your charged particles are going to splatter in the calorimeter and  
also be reconstructed by the clusterizing algorithm.  Unless you are  
doing anything to specifically veto those requiring exactly two  
photons is a bad idea.  In the studies we did the only charged  
particle was the recoil proton and we were able to suppress the shower  
it generated by looking at reconstructed vs. generated polar angle.   
My suggestion would be to loosen your cuts and look for the two best  
photons to make a pi^0.  I believe the pi^0 factory should provide  
with all fitted combinations of two photons.  In the longer term we  
need to develop an algorithm to extrapolate tracks and "tag" showers  
that come from hadronic interactions.

-Matt

On Mar 14, 2008, at 11:14 AM, Matt wrote:

> Hi guys,
>
>    Curtis and I are trying to run a study where we generate an omega  
> of 0 width, allow it to decay to pi+, pi- and pi0, while letting the  
> pi0 decay to two photons. We're then going to fit the full event for  
> 5-constraints (4-E/p conservation, 1 from constraining the gammas to  
> come from a pi0) and look at the width of the reconstructed omega to  
> see how much the fit improves the resolution. The charged tracks  
> will be smeared by the resolution parametrization and the photons  
> will be reconstructed by the BCAL/FCAL factories and
>
>    However, I'm running into issues reconstructing the pi0 from the  
> photons. I used genr8 and allowed it to decay the pi0 so I actually  
> feed the final state photons into HDGeant. The input file for genr8  
> and the output from HDGeant can be found at
>
> http://www-meg.phys.cmu.edu/~bellis/gluex/
>
>    When I was running my fits treating the pi0 as a missing  
> particle, things looked good. As soon as I required two (and only  
> two) final state photons to be reconstructed, the fits looked bad.  
> Using hd_dump I looked a little further into the output file.
>
> hd_dump -DDMCThrown -DDPhoton -DDPi0 p_omega.OUTPUT.hddm
>
>    The photons always look like they have very low energy, more like  
> noise. And they never seem to look like the photons I pass in. The  
> DPi0 factory also never seems to reconstruct a  pi0.
>
>    I've updated all my libraries and HDGeant and done a clean build  
> so I'm not sure if it's in my initial generation or in the  
> reconstruction. I've also placed a file in that directory for genr8  
> where I pass in the pi0 to HDGeant and let it decay it, but it  
> doesn't seem to help matters when I process it that way either.
>
>     Any thoughts anyone?
>
> -- 
> -- 
> ----------------------------
> Matt Bellis
> Carnegie Mellon University
> (office) 412-268-6949
> (cell) 412-310-4586
> ----------------------------