It was my premise that the 60-megapixel Sony A7R IV with its 3.74 micron pixels* would place an unprecedented demand on lens performance, demands never before experienced except when using Multi-Shot High-Res mode of the 45MP Panasonic S1R.
I was not disappointed—or rather I was quite disappointed. Last night, I shot 22 different lens combinations in order to assay what to expect (I’m counting zooms and their focal lengths as a lens).
Today in going over the files I was shocked at how disappointing most of them were—yikes, this is going to require careful selection and usage to really use the Sony A7R IV to its best. The gains for 60MP are real, but you’d better have the best lenses and perfect shot discipline and be aware of focus errors, focus shift and field curvature and lens asymmetry, with most of the lenses showing symmetry issues of varying degrees (and I’m beginning to question Sigma’s quality control, looks like the claims are bogus as to MTF testing).
As just one example in the quality control area, the Voigtlander 21mm f/1.4 Aspherical which I praised so highly on the Sony A7R III... well this latest copy is a dud, unable to make sharpness in the outer zones at f/5.6 or even f/8—what the hell? A bummer, as I was hoping to use it a lot. Update: I received another Voigtlander 21mm f/1.4 and it looks to be as good as any I’ve yet tried (best of three so far)—excellent symmetry and edge sharpness.
Here’s a headache: stopping down is not much of a solution for focus shift and field curvature and lens asymmetry because f/4 is the last aperture (for a high-grade lens) that avoids diffraction effects! Using f/5.6 shows subtle micro contrast losses with the best lenses, and f/8 visibly degraded. It’s exactly analogous to f/8 vs f/5.6 on the Sony A7R III—f/8 is the new f/11, so to speak.
* 9552 x 6360 images in a 35.7 X 23.8 mm form factor— not quite full frame sensor, about 1.6% smaller.
Total benefits are still a plus
While the resolution gains require the best lenses and shot discipline, the benefits of higher resolution are there in other ways for “free”: reduced moiré, reduced color aliasing, reduced spurious detail, all of which produce a superior image compared to the 42MP of the Sony A7R III. Image quality is a lot more than just resolving power!
Gordon S writes:
Very interesting update and glad to see you are finding similar challenges with the Sony A7R IV.
I was beginning to question wether I had some issues with my A7R IV or just my techniques. With IBIS on, I am finding the shutter speed for sharp shots needs to be almost 3X the focal length for consistency.
My simple backyard testing parameters are showing serious flaws in some of the lenses I would take on a family vacation that allowed for both easy walking around plus ability to have some more serious shot taking time.
Retried the same shots on the Panasonic SR1 and shots were sharp - so while I had hoped for the smaller footprint for travel I am not ready for spending a few weeks on the road and constantly having to chimp to see if things are in focus and sharp. (this was comparing the 24-105s to each other). I think my spouse would shoot me! She is normally a good sport but if I began taking extra time on every shot it would end up pushing her over the edge. Just the couple hundred frames I tested over the weekend in the backyard was getting me some serious annoyed looks.
DIGLLOYD: that IBIS comment raises serious concern but the other comments are consistent with what I am seeing.
I have now had to exchange four lenses, hoping for better samples:
- Sigma 35mm f/1.2 DG DN Art — 2nd sample is better, but has issues. Both have symmetry problems and both are unimpressive at f/1.2 and f/1.4 with 5-pixel haloes in blue light.
- Sigma 14-24mm f/2.8 DG DN Art — 2nd sample is definitely better, and more symmetric. First sample had a soft left side that required f/8 for sharpness approaching right side.
- Sigma 45mm f/2.8 DG DN Art — waiting for replacement, 1st sample had obvious lens skew.
- Voigtlander 21mm f/1.4 Aspherical — waiting for replacement, the sample I tested in July was superb, though it had a slight lens skew. This latest sample has obvious skew that only f/8 overcomes.
Per K writes:
My experience is that Sony A7x IBIS is below average compared to the competition. But not as poor as the reader experiences on this brand new mk4. My advice to him would be to use the warranty.
IBIS performance definition used is, to me, rather worthless: What is "5 steps" really? staring from 1/8000 sec? I have some personal (being not very stable myself) tests with non stabilized and stabilized 85mm on A7R2: The gain with stabilized lens was about 1/30 sec. With the stabilized Batis 135/2.8 I can get sharp images (if I am lucky) at 1/30 sec. The gain on just IBIS is in practice far from 5 steps that Sony brags about. I'd say "two steps" or rather 1/30sec. One has to bear in mind that results heavily depends on how you hold the camera and wind conditions.
(Last year I was trekking in Nepal and I was really happy, especially above 4000m, with the short and light Batis 135! It is 614g, the Sony 135 weighs 50% more and the Sigma 100% more. Weight and balance actually contributes a lot to image quality!)
DIGLLOYD: my impression is that Nikon Z7 and Fujifilm GFX100 IBIS are better than Sony.