momentum conservation, daggering at source

[kostrzewa]

init_global_data performs enforces momentum conservation by requiring

p_si = -p_so

Does this mean that the operators at source are only "implicitly" daggered?

[martin-ueding]

This would assume that all the operators are real. Is that indeed the case?

[kostrzewa]

Well, no, just that the Dirac structures are self-adjoint. The ordering of quark fields (and resulting direction of propagators) must be fixed in the input file. While the gamma matrices are of course individually Hermitian, their products are not...

If there is no daggering, then one must specify the operators correctly. E.g., for the

< \bar\psi gamma_0 gamma_5 \psi  (\bar\psi gamma_0 gamma_5 \psi)^\dagger >

correlation function, one would have to define two operators. One for gamma_5 gamma_0 and one for gamma_0 gamma_5