Focus or Image Quality?

As I look at post after post and email after email concerning little nits in image quality between possible camera purchases, I wonder if people are actually understanding what’s important to their shooting. Maybe focus is more important. Or maybe it’s something else, like fps and buffer? Or tethering stability? 

But let’s stick with the headline, shall we? 

The critic in me says this: if the image isn’t in focus it won’t much matter what the camera’s basic image quality is. That’s true of landscapes, of sports, of events, of studio shots, of pretty much any photography you do. Focus—and its relative depth of field—needs to be nailed if you want ultimate “image quality.” 

I only write this because recent cameras are changing the equation of how we think about this a bit. 

Go back even a few years and focus was more the responsibility of the shooter than the camera. In the earliest autofocus systems I often was tweaking focus manually after letting the camera do its thing. If you saw me on the sidelines of a sporting event in the 90’s I can pretty much guarantee you that my left hand was on the focus ring and overriding the focus from time to time even on the best autofocus Nikkors. 

The progression in the 00’s of D1 to D2 to D3 wasn’t particularly about focus. Sure, these cameras each had different focus systems of varying ability, but my trust factor on them was still on the low side: I was still micromanaging focus and overriding the camera decisions quite a bit of the time. Thus, in that progression, it was actually more important to me that the quality of the basic raw data was being improved. We’ve seen an enormous jump in that from, say, the original D1 to the D810. 

Low level data integrity is pretty darned good these days in almost all APS/DX and full frame/FX sensors. It now takes a lot of testing and a truly trained eye to see any difference in raw data between similar cameras. 

Yet focus performance is seeing an interesting resurgence of improvement. The mirrorless systems are superb at nailing static subject focus planes (assuming you know how to direct the system as to where to put that plane). Some are near magic. Sony’s face-detection mode on the A7’s may be better than I am at nailing focus on the eyes for static portraiture, at least in some situations. 

But what really prompts this discussion are the D5/D500 changes. 

My level of confidence in the autofocus system has gone up with the D5 generation. I’m finding less reason to override the camera—assuming that I’m controlling the focus point and focus settings correctly for the situation. It’s even easier to configure the camera to jump from one type of focus to another with the programmable controls Nikon has added this time around. 

So here’s where the question I ask really comes up: D4 versus D5. 

Clearly the D4 has better image quality measurements in the lower ISO range. Eventually, at higher ISO values, the D5 wins. If I were evaluating the two solely on low level image quality, my decision would depend upon how much shooting I did below ISO 2400 (D4 wins) versus above it (D5 wins). 

But that’s without considering focus. There’s no situation where I’d prefer the D4’s autofocus system to the D5’s. None. 

Not that there was anything wrong with the D4’s autofocus system. Obviously, I’ve used it with great success for a number of years now. But it simply isn’t as fast or as intelligent or as accurate as the D5’s is proving to be (once learned). 

So we’re back to what I said earlier: if the image isn’t in focus it won’t much matter what the camera’s basic image quality is. Thus, if the D5 helps you nail focus more often, it doesn’t really matter what it’s low ISO dynamic range is. 

Over time, I’ve noted that my use of many mirrorless cameras really fall vulnerable to this same evaluation: if the camera I’m using doesn’t allow me to always nail focus for the situation I’m facing, I get frustrated with it. It doesn’t matter whether it’s a 42mp state-of-the-art full frame sensor (Sony A7rII) or a smaller 16mp m4/3 sensor that’s not going to match my full frame sensors in dynamic range: focus is what determines whether I use these cameras or not. 

I’m probably more demanding at this than most of you reading this, as I’m shooting more sports—already committed to several college events in the fall—plus I shoot wildlife in wild, and these subjects tend to move fast and unpredictably (especially true considering my close-in positions to the subject). But I also think more of you should be more demanding about focus than you are, even shooting somewhat static subjects.

More and more the camera industry looks like the old Hi-Fi industry (including the eventual collapse of that market and the reasons why). An awful lot of folk are too busy arguing about esoteric measurements that they don’t understand or how the data was generated, and thus are not focusing on the right things (intentional pun).

Let me remind everyone: photography is a huge pile of decision making. Those decisions start with buying and selecting what to bring/use for a shoot. The decisions continue during the shoot, and if you’re shooting raw, they continue long after you put the camera down. In every decision there are compromises that you will have to make. If you emphasize X in any decision, Y will be compromised. That’s why you’re making the decisions: to optimize to your intended result. 

So consider this: what are you compromising if you absolutely insist on purchasing the camera that has the highest engineering dynamic range? And what are you compromising if you don’t buy a camera with true state-of-the-art focusing? Obviously, those of you shooting static subjects with manual focus lenses will answer those two questions differently than someone shooting moving subjects with autofocus. But they’re still questions worth asking.

Looking for gear-specific information? Check out our other Web sites:
mirrorless: | general:| Z System: | film SLR:

dslrbodies: all text and original images © 2022 Thom Hogan
portions Copyright 1999-2021 Thom Hogan—All Rights Reserved