Hi Paul,
I would say the DPReview article makes a lot of good points. I have not evaluated all they statements, but there is an awful lot of hype about MFD and understanding the differences may matter a lot.
Regarding DoF, you will get equivalent DoF and bluriness using equivalent focal lengths at equivalent apertures. That is a fact. If DPReview says it is so, it is still a fact. Large aperture 24x36 mm lenses used be not so good, but nowdays we have a lot of great options and with live view we can even focus them correctly!
Another statement often made is that MFD gives different perspective. But, there is only one way of changing perspective and that is to move the camera. With shifts we can have different projection, but that is something different from perspective.
Their DR comparison (house against bright sky) does not include the GFX, only the Nikon D800/D810, Canon 5DIII and Petax 645Z. I checked out the raw data and raw histograms on Nikon D810 and Pentax 645Z are very close. So, it is a well excuted test.
What is confusing to me is the difference in exposure 3s on the D810 and 1s on the Pentax 645Z. That needs some more raw digging. I see more noise in the raw data on the Pentax and that is not consistent with Bill Claff's data, for instance. More digging needed.
Essentially, I would expect the Pentax 645Z to have a 65% ISO advantage over Sony 24x36, like performing at 160 ISO like 24x36 mm at 100 ISO.
Regarding MFD and smaller formats I think we need to look at things like use cases:
- Need for high speed flash sync?
- Is moiré an issue?
- Need for tilt and shift?
- Portability/mobility
- Axial chroma a problem?
- Print sizes?
- Etc
Best regards
Erik
I can't disagree more with the Dpreview post, click bait. Many of the points they make just don't stand up, especially the BSI tech on the Sony and their comments on DOF MF/35mm. Even the comparison of the ISO 64 to 100 between the D810 and GFX has limited value again due the raw conversion of LR.
Also curious what raw converter was used for the high ISO testing of the GFX, if it was LR/ACR I am not totally surprised by the red cast. There are other raw converters that will do a better job. In fact the in camera jpgs may be better than the image if was from raw. If you shoot beyond 1600 ISO, you will need something else besides LR/ACR for now.
Pixel shift is real, trust me. I agree with Quentin, a good pixel shifted file will be very very close if not the same. The Fuji GF glass is excellent IMO and with the right raw converter you can get very good results. Pentax's mistake with Pixel shift was not to better support it. Currently the LR solution is terrible as it can't handle any motion, C1 can't figure it out or is concerned that it might compete with MF!! Silkypix does the best job, but I don't prefer their tools, and thus spend an inordinate amount of time with file that in C1 or LR would work much faster. The Iridient solution for Pixel shift is also very nice, but again lacks tools. (note if a one man show at Iridient could figure it out I would hope C1 could also by now). Silkypix Vr 8 is an improvement but not by much.
Paul Caldwell