Nsubjettiness often giving 0 for collimated photons

[Noah Steinberg]

Hello all, I'm interested in trying to reconstruct collimated photons. Without giving two much background I have a simulation which contains many events with pairs of photons from the hard scattering event which have a 0.025 < delta R < 0.4

I am trying to cluster these photons into jets, first by clustering them with the anti-Kt algorithm with the R parameter = 0.4. Then I re-cluster these jets using the Kt algorithm and compute the Nsubjettiness of these jets. My goal is to compute tau2/tau1. Here is the relevant part of my delphes card

module FastJetFinder JetFinder1 {

set InputArray EFlowFilter/eflow

set OutputArray jets

# algorithm: 1 CDFJetClu, 2 MidPoint, 3 SIScone, 4 kt, 5 Cambridge/Aachen, 6 antikt
set JetAlgorithm 6 #anti Kt algorithm
set ParameterR 0.16

set Recluster 1 #flag to recluster with kt algorithm and same R parameter

set ComputeNsubjettiness 1
set Beta 1.0
set AxisMode 4

set JetPTMin 20.0

}

What I am finding is that for many, many of the photon jets in this event that although tau[0] gives a non zero value, tau[1] is very often 0. This does not happen all the time, and obviously depends on the delta R between the photons, but for jets containing two photons I would expect tau[1], which is the 2-subjettiness, to be non zero. Is this something with delphes that I need to change? Or perhaps this is just physics. Any help would be appreciated.

[Noah Steinberg]

Any help would be appreciated