This is the eighth in a series of posts on the Fujifilm GFX-50S. The series starts here.
I got outside for this one. I put the Sony/Zeiss 55mm f/1.8 Sonnar (Zony 55) on a Sony alpha 7 R, Mark II (aka a7RII), and the Fujinon GT 63mm f/2.8 lens on the Fujifilm GFX-50S. Shutter set to EFCS for both cameras, which meant it really was EFCS for all the pictures with the Sony, and was EFCS at the narrower apertures for the GFX. Exposure compensation set to zero. AF-S used in both cases. 2-second self-timer used in both cases. Arca-Swiss C1 cube on RRS sticks. The Fuji images were about 2/3 of a stop darker than the Sony ones, indicating that the Fuji metering is a bit optimistic compared to the Sony. A more charitable way to look at that would be to say that he Fuji offers more highlight protection than the Sony. Base ISO for both cameras. Focus was in the center of the image. Tripod and head were not moved between series.
I developed all the images in Lightroom with default settings, except for Exposure. I adjusted that to roughly equalize brightness. ACR/Lr force lens distortion corrections for the Fuji 63, which probably reduce corner sharpness.
The scene, with the Zony set to f/2 and the Fuji lens set to f/2.8:
In the middle, at 250% magnification for the GFX, and the same sensor extent for the lower-resolution a7RII. This would give the same field of view if the lenses were equivalent focal lengths, but the 55 is longer than the full-frame equivalent of the 63.
Here’s how to use these highly-magnified crops. The dimensions of the GFX sensor are 8256×6192 pixels. If we make a full-frame print from the GFX on a printer with 360 pixels per inch native driver-level resolution, like the Epson inkjet printers, we’ll end up with a 23×17 inch (58×44 cm) print. The 318×246 pixel crop you’re looking at will end up 0.8333×0.6833 inches (2.12×1.74 cm). Let’s imagine that you or your viewers are critical, and will look at the 22×17 inch print from about 18 inches (conventional wisdom is that the distance would be a little greater than that, or 28 inches (the diagonal), but you did buy a high-resolution camera for a reason, didn’t you?).
The next step is dependent on your monitor pitch, which you may or may not know. Turns out, you don’t have to know it. Just take the 253% crops and view then at 1:1. How high are they? Get out your ruler and measure, or just guess. Let’s say they are 6 inches high. 6 inches is about 7 times 0.8333, so in order to view the crops the way they’d look from 18 inches on the print is to view them from 7 times as far away, or 10.5 feet.
Everything here scales proportionately. If the image on your screen is bigger than 6 inches, increase your viewing distance by the ratio of your image height to 6 inches. If you thin your viewers are going to almost get their nose to that print and look at it from six inches, divide that 10.5 feet by 3, and look at the image on the monitor from three and a half feet away.
The images from the a7RII are scaled to the same print height of 17 inches. Because it has fewer pixels vertically, that requires slightly greater magnification.
There in an argument in favor of comparing the f/2.8 Fuji image with the f/2 Sony one. That’s the equivalence approach. But you could also take the practical tack, and say that in a landscape situation, you’re gong to pick the best aperture for the lens and the scene, and use what ever shutter speed you have to. I favor the latter way of looking at things.
Either way, it’s clear that the Sony image is inferior. That is not to say that the Sony is a bad camera, or that the Zony 55 is a bad lens. In fact, in the past I have marvelled on how good my copy of the Zony 55 is, how well it does compared to spare-no-expense lenses like the Otus 55, and speculated that I got lucky when it came to serial number.
At f/4:
The Sony is getting better, but I’d sure rather have the GFX pic.
F/5.6:
Are you detecting a pattern here?
I’m convinced. How about the corners?
I’m not looking at the same part of the scene in these, because the fields of view (and the aspect ratios) are different.
I call the Fuji a modest winner over the Sony at f/2.8, but they’re real close.
I call that a tie.
Now I think the Fuji is doing better.
The Fuji is clearly better.
What if we don’t go all the way to the corners, just to the horizontal extremes?
Sony is softer, lower contrast.
Same thing.
There’s a pattern here.
Both corner and left-center crops I chose were of a part of the subject somewhat closer to the camera than a plane normal to the lens axis passing through the center subject. The usually helps lensse like this, but not alway. Let’s chose a left-center crop slightly behind that plane.
My take on this is that the Fuji 63 has a focal plane that is a bit closer to the camera than the Zony 55.
Completing the series:
So the Fuji is markedly sharper except at the very corners, where the Sony comes close. As I said earlier, ACR/Lr force lens corrections for the 63, which probably reduce corner sharpness.
The corner performance of the Fuji system doesn’t surprise me, but the center and edge do. The GFX is quite a camera.
A few notes: The tiltable EVF of the Fuji was remarkably convenient for making this series. I ended up not using the a7RII’s EVF, but relied on the LCD screen, which is harder to see in bright light. To get the camera to a convenient height for the Sony’s EVF, I would have had to use more tripod extension. It’s really nice to have tripod height to be non-critical. Lightroom performs camera and lens-dependent sharpening when converting raw images. Thus, tests like this one are always a little apple/orangey. However, I messed around with sharpening on the Sony images, and I couldn’t get them to look anything like the Fuji ones, except is the cases where they were very close to begin with.
Thanks Jim, that was a really interesting test, I’ve been curious as to how those two cameras compare. I’m genuinely surprised at how badly the Sony did here, I use my a7rii with the Zony55 all the time and I’m always blown away with the results! The Fuji is just one a whole other level.
You shouldn’t think of it as the Sony doing badly. The GFX just did very well. The Sony is every bit as good a camera as it was before the Fuji shipped. And look at the cost delta.
PS: see the update with the left edge images.
Thanks for this, Jim. I’m an a7rii user who is itching to upgrade– and I’ve been eyeing the gfx– but the biggest turnoff for me is that the gfx sensor is 3+ years old (I like shiny new things!). Any thoughts/predictions on what a more modern FF Sony sensor would bring to image quality (rumored to come out in a 7rii successor later this year)? Thanks!
Hi,
Jim’s test indicates that the Fuji lens is a better companion to the GFX sensor than the Zony lens is to the Sony A7rII. As Jim says, he has a good sample of the Zony lens.
So, how would a new sensor affect things, assuming more pixels? The MTF of the sensor would increase, due to smaller pixel size. MTFs multiply, so the Zony lens would give a better image with the higher MP sensor.
The same would apply to Fuji, with a new higher MP-sensor. Fuji indicated that all their lenses are calculated for a future 100MP sensor.
It is interesting to note that Jim believes that corner performance is lost because of mandatory software correction of the lens. I often check corners first and found that Fuji lens was a bit weak in the extreme corners in the DPReview test I have seen.
Best regards
Erik
Neither Fuji nor Sony favor me with any insights into their product planning. Sorry.
Certainly, with the lenses I’m using, a finer pitch in either the GFX or the a7RII sensor would help. My guess is the point of seriously diminishing returns on-axis is probably between 500 MP and 1000 MP.
Impressive, especially since the Fuji images were darker, so that the “more photons” advantage of the Fuji was sort of compensated.
I’m a bit concerned about the LR sharpening. I’m aware that there’s an argument for comparing using the software you’ll be using fo real photos, but carrying that further we should really be comparing prints. dcraw should be sufficiently camera and lens agnostic – or doesn’t it support the Fuji yet?
Noise doesn’t play much of a role in any of these images, so the photon deficit of the GFX is immaterial here. I will be testing whether this is a metering thing on the part of the GFX, or if the sensor is base ISO lower than the a7RII sensor.
Well, Jim….
First of all, apologies for being absent for a long time. Life intrudes, and I am incredibly busy right now and about to get busier.
Also, I’m officially out of Sony right now, a bit wistfully. Started my FF path with the A850, then an NEX7, and then the A7R as you may remember, so I was reading your blog a lot. But when I got my Z in mid 2014, I barely picked up the A7R. Held on to it for another 2 years, and then did another trade-in binge to go all Pentax with the K-1 as a backup to the Z, and with the adapter can use all my Z lenses on it (if I want to…). Made more sense, and actually the K-1 fits me ergonomically better than the Sony.
So, now I see you have the Fuji , and I can only say that I am not surprised to see what you are finding. With the right lenses that sensor is pretty darn good. Undoubtedly there will be better ones coming along soon, but I could easily stay in place with what I now have and not feel that the work I’m doing needed something more—which nagged at me before. I wonder if you will begin to feel the same way, or if your questing/testing nature will lead you on.
But I’m glad to be stopping back in.
Nice to hear from you, Tex. Good to know you’ll be stopping by from time to time. I always appreciate your perspective.
The inescapable issue I noticed again and again was the purple fringe from the Zony. I can’t stand it.
I got rid of my A7r and have a very good FE55 sitting on a shelf – guess it’s time to take my losses on that one too.
Oti (and Apo) all the way.
Max, lots of nice lenses have a fair amount of LoCA. Of course, the Otus lenses do not.
This comparison is really stupid.
I’m surprised that someone have courage to publish the comparison like that on internet.
If I would do comparison between these two cameras, I would choose static subject not a trees that can move in the wind. I would choose subject that has at least three well defined colors e.g. red brick, blue roof, green fence. I would choose subject that feels large portion of picture with object that is flat and in focus.
See the link
https://www.google.ca/search?q=photos+of+houses&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwi4w5G-7IHTAhVK0YMKHeBqB3UQsAQIGQ&biw=1280&bih=603#tbm=isch&q=photos+of+brick+houses&*&imgrc=kEkQ4YOS8jjwAM:&spf=475
It’s very bad that people who could provide valid and interesting comparison, do not have these expensive cameras.
I never had in my hands any of the cameras, but from technical specifications, I can tell readers the following, with 100% accuracy:
1/ sony has in body and/or in lens stabilization. Therefore it has to be concluded that for regular shots, taken by hand-holding camera, sony will deliver most of the time sharper pictures.
2/ sony 7RII sensor is technologically way more advanced (BSI stacked sensor) and it has fewer megapixels. Therefore it has to be concluded that PER PIXEL it receives about the same amount of light as MF Fuji as fuji sensor is not that much larger.
You are entitled to your opinion. I do not share it. Tree leaves and branches have a long history in informal lens tests, going back at least as far as Ansel Adams. Regular subjects have their own problems. If you’re patient, you’ll see future tests of the GFX with standard and non-standard test targets.
Your statements about IBIS are apropos to hand held photography, and they make a lot of sense. I think the GFX is a camera that is best used where there is a loto of light, or where the phtographer can make a lot of light. Another thing the a7RII has going for it is the change in conversion gain that happens at ISO 640 and above that gives it a photographic dynamic range that is greater than the GFX at those ISOs. The chocie of how much sharpness is lost due to camera movement is the photographer’s with both cameras, however, in spite of the fact that the shutter speed will need to be higher in the GFX, all else equal.
Your calculations about the GFX pixel pitch are inaccurate. The pitch of the GFX is actually coarser than the a7RII. The pitch of the GFX is 5.3 um, and the pitch of the a7RII is 4.51 um. Therefore, all else equal, the pixels in the GFX receive (5.3/4.51)^2 = 138% as much light as those in the a7RII. All else isn’t equal, however, and the microlenses on the GFX are not as large in relation to the pixel pitch as those in the a7RII, so the GFX per pixel light-gathering advantage will be less than that.