[Primexd] [EXTERNAL] Re: Accidental background subtraction

[Shepherd, Matthew]

Hi Igal,

Can you explain how accidentals enter into your analysis (an how you remove them)?  You must be looking at complete reconstruction of a topology in order to be able to choose which beam photons are accidental?  Maybe you are using the term accidental in a way that is different from what people commonly call accidentals:  beam photons that are out of time with the primary event.

In the standard calibration the only thing we use is the event vertex.  The beam energy does not enter because we are using an inclusive pi0 sample.  So it is hard to see how the conventional notion of an accidental is relevant.  You can clearly pick a beam photon using the photons that make the pi0, but since the beam photon itself doesn't enter the analysis, one doesn't reduce combinatorics or background by being able to select one beam photon.  And you wouldn't necessarily know that just because two photons came from the same bunch it was the bunch that really produced the beam photon.  Maybe there is something I'm missing....

As we were discussing yesterday, it is helpful to try to ensure both photons that make the pi0 come from the same bunch.  Colin and I took a look at this, and our skim cut is rather loose:  | delta t | < 10 ns, but this still probably removes a fair bit of background given the readout window size.  We could probably tighten this further to say 4-5 if desirable, but the procedure seems to work well so far.

It is too bad you won't be able to make the calorimetry meeting -- it would be useful to discuss these technical calorimetry things in the calorimetry meeting and save writing email.  I'll try to look through the slides from the PrimEx meeting again as I did last time.

Matt


On Apr 15, 2020, at 3:26 PM, Igal Jaegle <ijaegle at jlab.org<mailto:ijaegle at jlab.org>> wrote:

All,

The subtraction of the accidental background has a dramatic effect and appears to be necessary for the proper FCAL energy calibration of the PRIMEX data. In short, all pi0's mean in the inner rings got shifted to lower mass after the accidental background subtraction. In short, if the accidental coincidence is too large, I believe it is almost impossible for a fit to capture the real pi0 mean particularly with bkg with large "slope".

Matt, I will not be able to attend the CAL meeting tomorrow but I will present my observation at the PRIMEX meeting this Friday.

tks ig.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/primexd/attachments/20200416/bc78b04c/attachment-0002.html>