Looking at the level of detail (even if those JPEG were a bit over sharpened) that 36Mpx Sony sensors without AA filter can produce in optimal conditions (high resolution glass, accurate focus, and absence of camera shake blur or diffraction):
I wondered how they compare to a digital back designed for studio work. Comparing the A7R sensor (36Mpx) with the ultimate backs: Phase One IQ 180 (80Mpx) and Hasselblad H3DII 50 (50Mpx) regarding different sensor parameters:
ISO:
ISO values in the backs are typical for their intended use: very low ISO values to maximize image quality under controlled lighting situations where getting a good exposure will never be a restriction.
The Phase One begins at an amazingly low (for those of us used to a typical DSLR scale) real
ISO29, keeping that value up to a ISO100 user setting. From that point ISO pushes up to ISO3200, surely through software RAW level scaling.
Hasselblad sensor is even more honest, it keeps a real
ISO47 in the whole user scale from ISO50 to ISO400.
The A7R however begins at a real
ISO73 when ISO100 is set, simulating ISO50 via metadata and reaching ISO25600 in the right adjustable end.
So regarding ISO no surprise, beginning at lower real ISO gains will help the backs to get the best from their sensors in the controlled situations they were designed for.
Noise in areas with a good exposure:
We can see that performance for a given real ISO, and once different resolutions have been weighted (DxOMark 'Print' mode), is the same for the three sensors.
Since the lowest available real ISO is lower in the Phase One (29 vs 47 and 73), those applications where a correct exposure can be obtained at the lowest ISO will benefit from an almost negligible 1/2 stop improvement in the Phase One over the Hassel and the A7R.
This parameter is important regarding noise when dark areas are not going to be lifted in post processing (in other words, when the JPEG straight from the camera is fine in terms of exposure), which is typical in studio applications to which these digital backs are aimed.
Noise in dark areas:
When comparig Dynamic range, or in other words the ability to lift shadows preventing noise from appearing, the A7R Sony technology clearly beats the backs by 2 stops (and that is much) for any given real ISO.
Again, if the application allows correct exposure at the lowest ISO, the lower ISO in the backs will reduce that advantage: Phase One will only be 1/2 stop below the Sony performance and Hasselblad 1,5 stops worse than the A7R.
In tripod applications with high contrast scenes (landscape, architecture) is where this parameter matters, and places the A7R sensor as the best performer. If the optics used allow to get the same real effective level of detail in the FF sensor or close (and here the diffraction/DOF trade-off plays a role in reducing the initial pixel count advantage of the backs), we could even conclude that the A7R could be the best candidate in these large DOF applications.
Please note that this was only a
sensor performance comparision. It doesn't take into account output Mpx, optics, the advantage of larger formats in DOF control or the so called "3D look" in shallow DOF images using large formats. It doesn't take into account any value for money consideration though
Regards