I simulated prompt gamma when a neutron is captured in Nitrogen. I used the same code under condition of two different GEANT versions.(geant 4.9.1p02, geant4.9.4p02) I compared normalized prompt gamma yields of two simulation results with experimental result(NIM 56 (1967)325). (<-- I read gamma kinetic energy when processName is "nCapture".) It shows quite different result at 10.83 MeV. I used Neutron high-precision models.
Created attachment 145 [details] prompt gamma ratio(Nitrogen)
This problem is caused by the final state adjustments for better conservation. I added capability to disable the adjustment for reproducing old result. You will get the old result by setting an environment variable of “G4NEUTRONHP_DO_NOT_ADJUST_FINAL_STATE”. However the environment variable significant may give impacts to result, for example you may lost residual nucleus in “Inelastic” reactions.