By good fortune (mistake) not one but two samples of the Sony FE 24-70mm f/2.8 GM II showed up on my doorstep.
My review is going to get off to a slow start, but I plan on covering the lens in depth. Tonight my first foray into how it performs.
UPDATE July 5: I’ve shot material and analyzed it, but an auto-immune relapse has left me in a challenging cognitive state (brain fog, concentration), making it extremely difficult to concentrate. I will do it when I can. Update July 7: major rebound, I can resume work. I got a lot of sun exposure on July 6 (~2 hours, whole body), could it be related?
Immediate impression is of ultra high image quality, similar to the impressive results obtained with other recent Sony FE lenses, like the Sony FE 12-24mm f/2.8 GM, Sony FE 35mm f/1.4 GM, Sony FE 50mm f/1.2 GM.
All of which show stunningly high potential, and all of which also show that getting a really good sample is a major effort likely to defy most of us, including me. There is just a darned lot of variation. The Sony MTF charts are computer-generated fantasies; no real lens will deliver those results symmetrically across the field. OTOH, a good sample looks to deliver the results across ~3/4 of the field. This is not news and by no means specific to Sony.
Already it is apparent that the potential of the Sony FE 24-70mm f/2.8 GM II is world-class—eye poppingly crisp where it works. If it were possible to get a really superb symmetric sample, I’m all in on that one-lens-to-carry proposition.
And already it is apparent that one sample is considerably better than the other at the wide end, both have symmetry issues, some focals have major astigmatism seemingly caused by being “off” a little (aspheric elements are often intolerant of deviations from spec), and both have major outer-zone focus shift such that f/5.6 is worse than f/2.8 at some focal lengths and good luck getting f/8 to fix that fully.
If you focus in the area of interest you’ll likely be pleased—very pleased—with your new wonder lens. But for challenging subjects like a distant landscape scene, all the “warts” of your particular sample (varying by focal length) are going to crop up like mushrooms after a drenching. Optical tradeoffs in lens design yield both spectacularly good performance and annoying bulges and shifts in where sharpness impacts.
My recommendation is to find a sample that you like at one end and the middle. For me that would be the 24mm to 50mm range. For example, I was pretty happy with the Fujifilm GF 35-70m f/4.5-5.6 in the 35mm to 60mm range, and quite disappointed at 70mm (weak left side). That’s the rule for most all zoom lenses—a really good sample will be strong over 2/3 of the focal length range, and acceptable to annoying at the other end.