Flux of medium resolution arm in simulated pfsObjects looks overestimated

Looking into the problem of redshift determination of the weekly end-to-end processing at IPMU, it turns out to be that the problematic objects are in pfsObjects including the medium resolution arm. The flux (in nJy) in MR looks significantly higher than that in LR for the same object (see attached plots). The problem happened after w.2022.03 release, so I guess this is due the Jacobian related stuffs. I hope you check the issue independently and fix if this is a real problem.