In my last post, I showed graphs of the input-referred read noise spectra at several ISO settings for the GFX 100 II. Spikes were visible which I attributed to the PDAF pixels. I ran a similar test with the Hasselblad CFV100C, and there are no such spikes.