To minimize the noise of a PtSi IR camera we examined all the sources of radiation reaching the detector. In particular, we suspected that the cold shield which is supposed to be (nearly) 100% black -- was not. In long wavelength FLIRs where the scene radiation is the main source of noise this was never an issue. In MWIR cameras this could be a source of added noise. The problem was to evaluate the noise contribution of the reflected radiation from the cold shield by sensor measurements. To this end we replaced the lens by an equivalent black (painted) tube which was temperature controlled (heated) and was outside the FOV of the cold shield. The mean charge in the center of the FPA was evaluated as a function of the temperature of the tube. The analysis of the test data showed that the cold shield reflectance is one percent, a low value. As utilized in our IR camera, this commercially available cold shield was determined to be effective in keeping the noise low.