How To Design a Perfect Mirrorless Camera

Oh that we could mix and match features and parts. 

Imagine for a second that you were appointed Full Frame Camera Czar, and that all the current full frame product groups now reported to you. Your job is to take the best of all the existing products and build the ultimate full frame mirrorless camera. What would that look like?

Here's my take:

  • Nikon Z mount (possibly Canon RF mount; I'd need to know more about the communications of each). Why? Because it's the most flexible mount for optical design. If you can't design it for the Z mount, it can't be designed for any mount, basically.
  • Sony sensor (we'll argue about which one, but it's still likely to be a Sony). Why? My answer might surprise you a bit: because Sony Semiconductor is willing to license from anywhere to make a better sensor. The Nikon/Sony relationship brought a lot of technology to the Sony sensors, and Sony Semiconductor found plenty of other sources for "better tech" along the way. Canon's going it mostly alone. No one else has the volume to do it at all.
  • Nikon Picture Controls (e.g. EXPEED over DIGIC and BIONZ when it comes to JPEG processing). Why? For the same reason as the last bullet: Nikon has been licensing things longer and from more parties in building up their JPEG rendering engine's intellectual property and capabilities. Now, you may want to tune the defaults slightly differently (e.g. apply the Canon double hue shift), but I'd argue that the little bits and pieces doing the black box work in EXPEED are more mature and subtle than in DIGIC or BIONZ. We can argue about this one, but make sure you bring real arguments and not just a bunch of default choices that were made for the out-of-box settings.
  • Canon dual pixel focus system. Why? The Nikon/Sony approach simply has too many short axis gaps, nor can they build a depth map to store for interesting later use. Technically, I'd want quad pixel focus to get truly equal focus discrimination, but of what we have today, dual pixel is the best choice.
  • The Sony SD/CFe dual card dual slot. These are the two card types that have a future, why not have two of both in the camera?
  • Nikon's menus (with maybe some idea inclusions from some of the others). Why? First of all, there's the clear differentiation between still and video menus. But this menu system is fast, supports touch, (mostly) well organized, fairly clear in language, has depth, uses graphics well, and has (mostly) usable help. 
  • Sony's AI. Why? Well, because they're the ones with the best machine learning algorithms so far (not actually artificial intelligence). This also ties into the sensor, as well, as that's where the "AI" is going next.
  • Nikon's physical controls with Sony's customization. Why? Nikon's still using the same basic control-and-command as they did with the N8008 back in 1988. The buttons and dials have been considerably rethought and refined over time, but the basic interface wasn't broken originally and Nikon never broke it trying to be clever (well, okay, the N50 did). But Nikon is a nanny and tries to minimize what you can customize in this interface (and for no logical, sound reason). Sony pretty much allows you to assign any function to any button. The two things combined would be awesome. 
  • SnapBridge with image.canon. Why? It's taken awhile, but SnapBridge is flexible and works well up to the point where you'd want it to talk to Nikon Image Space and do things automatically. It's like Nikon is getting the front end figured out while Canon is getting the back end figured out. Add some things in the middle and make it all one, and that's what we really want.
  • Nikon's metering. Canon still overshoots based upon scene contrast in complex situations. Sony does better, but tends to run hot. Nikon's many years of improving, refining, adding to, tweaking, and doubling down on color has made their metering the most consistent and accurate I've found. 
  • Sony's top EVF. (with Nikon's optics, Canon's eyeglass user mode, and a few other goodies). Why? Well, Sony's already basically making everyone's OLED panel. So the hardware is easy to pick. It's what you do with the hardware that needs a bit more work.
  • Canon's Image Stabilization. Why? Well, they're all pretty good, but Canon seems to have gone further (at least with the R5 and R6 version). Throw in Nikon's Electronic VR for video, too. 

We do have to dip into some other formats to get a few final pieces, though:

  • Olympus Pixel Management (Live Composite, Pixel-shift, etc.). Why? While some of this is leaking into the full frame cameras, the implementations aren't as good as Olympus's. And add in Fujifilm's focus shift shooting.
  • Fujifilm's tilting/flipping Rear LCD (or: a better variation of it). Why? Videographers want total articulation, still photographers want tilting. It's possible to make something that does both, but the only ones that have sort of tried that are Fujifilm and Pentax, and neither make a full frame mirrorless camera.

I'm not going to delve into the video aspects of the camera. We all just want no crop, 10-bit, Log, with a high bit rate codec and some form of raw support, and at least up to 4K 60P now. Do better than that and we're all good.

So there you have it: the new ARZ1 Mark I. Coming never to your no-longer-existing brick and mortar camera store. 

text and images © 2020 Thom Hogan
portions Copyright 1999-2019 Thom Hogan-- All Rights Reserved
Follow us on Twitter: @bythom, hashtags #bythom, #sansmirror