Family photography with the X2D/XCD90V and the GFX 100 II/GF110

I reported a year or so ago that I’d had trouble with the autofocus on the Hasselblad X2D-100C at family gatherings. That was using the XCD 38V lens. This year, I spent two days making those kinds of images with the X2D and the XCD 90mm f/2.5 V lens. Then I spent two more days making the same kinds of images with the Fujifilm GFX 100 II and the GF 110mm f/2 lens.

Similar outfits. Same format. Heck, it’s the same sensor! Both putatively with on-sensor phase detection autofocus (OS-PDAF). Similar focal lengths. Apertures only differ by two-thirds of a stop. Shouldn’t be much difference, right?

Wrong. The experiences were completely different, and necessitated different techniques. The pictures that both cameras could make look very similar, even if there were different hoops to jump over to capture them. However, the Fuji could make some images that the Hasselblad couldn’t. The Fuji camera has continuous autofocus, and the Hasselblad one does not. But the differences extend well beyond that.

As I had experienced with the 38V in similar circumstances, the X2D hunts a lot in dimmish light with low contrast subjects, and when it does focus without hunting, it is significantly slower than the GFX 100 II and the 110. I was trying to make unposed images, although I’d accept it if someone noticed that I was taking their picture and struck a pose for me. That means I couldn’t tell people to freeze while the Hassy was taking its own sweet time about getting the subject in focus. As I did in the same situation a year ago, I gave up on the X2D autofocus, and ended up using manual focus and the excellent Hasselblad focus confirmation indicator. This feature really is something special. It tells you in what direction to turn the focusing ring and indicates when the image is in focus in the desired area, but most such systems do that. It tells you how far out of focus you are so you know how far to twist the ring, and its tolerance for focus error is tighter than most such systems. Without it, I would have been completely lost. When the autofocus system works on the X2D, it’s as accurate as that on the GFX 100 II. It’s just that the GFX can quickly achieve that level of accuracy in situations where the X2D lies down on the job.

The Hassy X2D autofocus system is confused by bright lights in the scene, either in the foreground or the background.

Here’s a situation where the X2D refused to focus at all:

In a similar situation, the GFX 100 II had no trouble at all focusing accurately and quickly.

Same thing with backlight.

Hasselblad:

Fuji:

The backlighting problems with the Hasselblad are made even worse if you cut the exposure to keep from blowing out the background, but the Fuji has no problems even in extreme situations like the one below, which took heroic measures in post to recover the dark subject:

And then there are some times when you really need continuous autofocus:

If you’re going to be shooting semi-posed images like this one:

then it probably doesn’t make much difference which camera you use.

But the next time I’m in this situation, I’ll pick up the GFX in preference to the X2D.

Biggest surprise of the comparison? The GFX EFCS shutter is a lot quieter than the leaf shutter in the 90V. There is a firmware upgrade for the 90V that is supposed to reduce the acoustic noise at low shutter speeds, but I haven’t tried it yet. In the spirit of completeness, I used a prerelease copy of the 90V. SInce my experience with it is so similar to what I experienced with the 38V, I doubt if a new 90V would improve the autofocus situation much. I’ll try to get a new 90V and do a comparison.

Addendum.

There are people drawing conclusions from these images that, IMO, aren’t appropriate.

The first is comments about the relative noise in the Hasselblad and Fuji images. I don’t see how you can do that from these greatly downsized, heavily compressed (first with JPEG’s DCT algorithm, and then with Imagify’s proprietary one) images which were made with no attempt to equalize exposure. I did tend to use lower ISO settings in the Fuj images, because not only was the Hasselblad AF worse when the subject was dark, the manual focus assist feature didn’t work well under those conditions. Therefore, faced with bright backgrounds, I used a lower ISO setting but the roughly the same  exposure with the Fuji than with the Hasselblad, which allowed be to keep the background from blowing out.

The second is judging color differences from these images. There are color differences between the two cameras and among different raw developers and color profiles. I have written about these before, using controlled conditions to make the test images. The images in this post were made under anything but controlled conditions, in different situations, on different days. Adobe Lr’s Camera Standard color profile for the Hasselblad images, and Adobe Color for the Fuji ones. These are the Lr and ACR defaults. Adobe Color with the GFX is not as accurate as the Camera Standard with the X2D.