I’ve never shot an Olympics. I’ve shot pro, collegiate, and high school sports of all kinds, but never the Big Event.
As I start shooting more and more sports again, I began contemplating what I’d have to do to earn a spot on someone’s photography team so as to be one of the chosen few shooting the next couple of Olympics. Given my age, I’d have to guess that the next two Olympics are my last chances.
But as usual, my mind quickly went elsewhere.
The 2020 Summer Olympics are in Tokyo. Both Canon and Nikon would be expected to bring out completely new generation pro bodies just prior to that, probably in early 2020 so that there’s time to get used to them before the games begin.
What the heck are those cameras going to be?
That’s an important question to ask right now, because both companies would already have their deep technology teams beginning to try to solve the problems necessary to deliver an incredible new flagship product by then. The 2020 Olympics is a home game, and if Canon and Nikon don’t deliver something great for that, it’s game over.
If I were a video shooter and these were video cameras we were expecting, the answer would be easy: 8K standard, with perhaps some sort of higher not-quite-ready ability (e.g. 16K, or maybe VR support of some sort). Absolutely great 1080P (120 fps minimum) and 4K (60 fps min) slow motion. Edit marking in camera would be the norm, as would raw file creation and an edit-ready compression like ProRes. Shape focus recognition might even be there (put the cursor on an object, the overall shape of the object is analyzed and then follow-focused, even as the shape morphs with motion).
But stills? Oh dear, that’s a far tougher challenge to predict.
I’ll tell you what I want: chimp-and-send. By that I mean that as I’m shooting I’m using pauses in the action to chimp my shots—preferably on a display larger than 3"—and pick the images that need to be immediately sent to the service or publication I’m working for. If connectivity isn’t available as I shoot, those chosen images are still all queued, and then squirted out just as soon as a connection is found. Any connection.
But am I using a DSLR, a mirrorless camera, or something else (e.g. multi-sensor quasi-light field)?
That, my friends is the US$64,000 question in Tokyo labs right now. Fujifilm and Olympus will absolutely say “it’s mirrorless.” They’re all in. Sony would probably say the same thing, though the recent A99II shows that they see an alternative. But Canon and Nikon? Magic 8 Ball says “Ask again later.”
If I had to guess, both will stick to DSLR. The last DSLR to be built will be at the high end, after all. That’s generally always the case with dying technologies: the remaining practitioners retreat to the highest ground as their volume goes away.
And yet, I can clearly see someone standing next to me at the Tokyo 2020 stadium with some non-DSLR camera that just runs the sensor at absurd speeds and generates a second or two of stills at 60 fps or even more, all without flapping a mirror. It’s a real pain-in-the-editing-butt to deal with all that data downstream, but I’m sure there will be someone willing to take up the challenge for the benefit. I have to do better than that person, but what’s the camera that allows me to do so?
The tricky part of this is lenses. Sports and wildlife and a few other photography disciplines still revolve as much around lenses as they do camera bodies, if not more. If you’ve ever seen the photo bullpens at the Olympics, they’re not necessarily close to the action. What you end up with is 100 still photographers with 300mm or longer lenses all piled into a small space really suitable for only 24, and all trying to take the same (or better) picture. A smaller mirrorless camera almost works against you in this instance, as those big cannons of lenses will almost certainly block your smaller lenses' view as they swing around in front of you. I once tried to take a 24-70mm shot in a restricted box like that, and most of what I got in the frame was other photographers and the sides of lenses.
Focus systems on the 1DxII and D5 are already wicked good if you know how to set and control them. Anything better than those will seem a bit like magic, but I’m sure that deep in Tokyo there are engineers trying to accomplish just that. It’s going to all be about pattern and color recognition and being able to not get distracted by competing elements as the subjects all move.
Image sensors aren’t going to push incredibly further in four years, I believe. At current rates of sensor improvement we’re already up against the treachery of random photons, and nearing some boundaries. Sure, if we could boost quantum efficiency up to 100% we’d get a clear benefit, but I’m not exactly seeing where a sports photographer in low light would get much additional benefit from current sensor tech iterated as normal. 24mp at ISO 12,800 seems about the extreme we can expect to get truly useful results from four years from now. (Yes, I know some of you are shooting that today, but you don’t have my standards. ;~) I’m currently maxed out at 20mp and ISO 6400.)
And do I want a mirror flapping at faster than 14 fps? I’m going to take an off-the-wall guess at this one: what if the mirror flipped every other shot? In other words, shot, shot, flip, shot, shot, flip… That would give us near 30 fps with current technology, though there would be some issues with blackout and timing that need some finessing to make this look right to the photographer and get a usable sequence. (Side note: this is a way the still cameras can get around the video rights issues; having a 30 fps burst that isn’t perfectly regular is a great way to argue to the rights owners “it isn’t video.”)
Still, the big thing is the dependence on sneaker net for getting images off the camera and to where they need to go. That surely needs to change. This is where Canon and Nikon need the most help. SnapBridge is currently about 1% of the solution ;~).
Look, I want to shoot raw so that those files can be be honed later if I get something really good. I really don’t want to be shooting RAW+JPEG because I end up having to deal with two sets of files in post. What I want is this: I chimp a shot, select a crop, add a caption, then mark it for sending. If I’m shooting raw, obviously, the camera does all that to produce a JPEG. Heck, add the SnapBridge “make a 2mp version” option, too.
“Sending” happens as my choice of (a) immediate; (b) batched; or (c) stays marked (and cropped) in some way so that I can find those from my laptop quickly if I don’t have immediate connection ability from the stadium.
The question is “what is the send method”? I suspect that it’s going to have to be in-camera cellular (and lets hope the Tokyo networks are 5G by then ;~). If things have to go through my phone first, that just adds a step. I’ve got a data plan, and I’m sure that for immediacy whoever I’m shooting for would pay for data usage of selected, ready to Web/print publish JPEGs.
So I'm getting closer to the definition of a D6 (or 1DxIV): 24mp, a bit better low light capability, 24-30 fps, more magical focus with even better user control, built-in cellular data connect. Better support for selecting, cropping, and annotating in camera. All in a DSLR that uses my current lenses.
Is that enough to hold off the mirrorless competitors one more time? I think so. But mirrorless is going to be closer in terms of camera capabilities than ever before at that point. So once again where the Canikon advantage lies in in lenses for this type of work. If Canikon wants to stay ahead of the game, more DO/PF type telephoto lenses need to show up (and Nikon needs a better 200-400mm f/4). Dare I ask for a 70-200mm f/2?
Okay, with that out of the way I can get back to thinking about how I might be able to get my ticket to the Olympics as a photographer.