The Modular Debate — Nikon DX

(commentary)

In my previous article, I made the case for a modular D4-type camera. I believe a D5 that is communicating, programmable, and modular would rule the roost in the pro market, and attract quite a few prosumers, too.

Today I want to write about the DX cameras, because they are a different breed, and they are the core of Nikon's sales and profitability. 

Here's something you'll probably find surprising: all those DX cameras are already modular in design, and have been all the way back to the beginning (even back into the film era). Wait a second, modular?

Yes, modular. But not in the way you're thinking, and not in a way that you can see or is useful to a consumer. If you open up a Nikon DSLR, you'll find a warren of ribbon cables. I think on one camera I counted 22 wending their way from the main digital logic board (itself a module) to various sub-modules tucked into various corners of the camera. You probably know some of this already: the metering sensor lives up in the prism area, the focus system lives down at the bottom of the mirror box, and so on. (If you're interested in this, see the iFixIt teardown of a D600, and especially note the image on Step 27: see all those cables?)

Let me step away from the DSLR for a moment to a recent repair experience I had. My well-traveled MacBook Pro suddenly had fans that sounded like they were losing their bearing, literally. I'm not talking about the air whoosh sound that happens when you press the CPU cores to 100% and the fans have to work to cool the computer. I'm talking about a clickety clack sound every time a fan changed speed, and a severe rattling when the fans were running over 2100 rpm. 

Okay, off to the Apple Store. Prognosis (nearly instant): fan replacement. Time to repair: less than one hour.

Now technically my MacBook Pro comes with a lower "repairability score" from iFixit than a D600 (1 instead of 2). But that's for user repairs. Apple's thought carefully about what's likely to fail and made it relatively easy for their folk to just do a fast module swap. Yes, there are still some ribbon cables in the MacBook Pro, but surprisingly few, all things considered. Apple has engineered an easy-to-manufacture, easy-(for them)-to-repair product, and that centers around absolute modularity of subsystems rationalized to the nth degree. 

That's exactly what we don't have in consumer DSLRs these days. Indeed, when I consider what the insides of a 90's era Mac laptop looked like versus the ones today, compared to the insides of a 90's era Nikon DSLR versus the ones today, what I see is a vast difference. Apple has been engineering the heck out of the insides of their machines: simplifying, modularizing, optimizing, and hardening the guts of these laptops to the point where "major repairs" take very little time and are easy enough to be taught to thousands of service folk at 400+ locations worldwide.

Meanwhile, the Nikon consumer DSLR is still the same mess of soldered and non-soldered cables streaking all over the place to components that have been updated, but not rationalized, simplified, combined, modularized, or optimized (let alone hardened). Repairs take more tools, more time, more components to be in stock, and an attention to keeping everything carefully aligned, as well. 

Recently, Nikon has shown us that they see the Apple way of doing things: take a close look at Nikon 1 and you see interchangeable lens cameras with an order of magnitude fewer parts, far simpler and straightforward assembly/disassembly, and better modularization of critical components. So it's clear that Nikon can do what's necessary to move away from the old SLR/DSLR designs. But why that hasn't already been done for the D3xxx type cameras, I don't know. We're a long way from the one-hour repair-on-demand with DSLRs, and continuing to execute 90's designs is one of the reasons. 

Don't get me wrong. The internal engineering on a modern Nikon DSLR is generally clever and first rate. The problem is that it's just an incremental advancement of what are now out-dated design approaches. The Nikon 1 cameras must be enormous profit generators for Nikon (at least when they sell at list price ;~): fewer components, simpler and faster assembly, easier alignment, and solid housings that keep everything where they're supposed to be (much like the typical Apple aluminum chassis these days). But the sensor choice and the stripped control designs of those CX Nikons didn't make them best sellers at DSLR prices. Still, I'll bet when Nikon put the original Nikon 1 models on fire sale, they were still making a profit on them, that's how far Nikon went with the parts reduction and modularity of the internal design. 

What Nikon really needs is the same level of "rethink" in terms of DX DSLR design as they gave the Nikon 1. I think it's coming. For Nikon's sake, it had better be. 

If focus moves onto the imaging sensor as most of us believe it will (already true of the Nikon 1, OM-D E-M1, and several other mirrorless cameras, and coming soon to the Sony DSLRs), if global shutters get past their noise disadvantage, if OVFs get replaced by EVFs, we'd already be a long way towards that possibility. I imagine a DSLR with a mount/sensor assembly, a digital board assembly, an EVF assembly, plus the usual battery, flash, and LCD components, all hanging off a secure body structure (not a bunch of screw-on plastic pieces—the D600 has 36 screws that hold all those pieces together—with a modest number of connections out to connectors and controls. 

One reason why DSLRs have all those cables snaking inside has to do with those controls. I'm looking at my D7100 at the moment: 22 buttons, 4 switches, 5 dials, 3 displays, and 9 connectors/slots. If you have to run a cable to each of those, that's 43 cables, though I'm sure that you can group some of those to reduce the cable frenzy. Still, much of the assembly complexity of the modern consumer DSLR is due to all those controls. 

We've seen camera makers begin to address this in interesting ways. For example, no top LCD on the low end cameras (replaced by putting that information on the rear LCD). Or Command Dials that also can be pressed in for a button action, making the Command Dial multifunctional, not single function. Or touchscreen controls on the rear LCD. But frankly, we need more than experimentation here, we need someone to show the way for how a modern UI can work with fewer external components. Sony tried that with the much maligned early NEX models (three-button plus dial control of everything). It actually wasn't the simplification to three buttons and a dial that hurt Sony, it was that their menu system was a mess and that originally they didn't let users do any customization of the three buttons. I'll give them full credit for trying, but it took them a while to make that system fully usable, and it still could use some clean up.

Imagine a consumer Nikon DSLR with the following changes, though:

  • Both Command Dials are also press in buttons.
  • All the Playback* controls basically become gestures on the rear LCD touchscreen.
  • The top LCD goes bye-bye (but the EVF gets all that information and you still can bring it up on the rear LCD, too).


* why just Playback? Because we handle the camera differently during playback than shooting. When shooting, if we're looking through the viewfinder we don't have access to the touchscreen, so shooting controls on the touchscreen are only useful when on a tripod using a Live View type of action.

In just three changes we've eliminated a slew of separate controls and at least seven of those internal cables, probably more if the changes were really well thought through. Thus, I believe it possible to make a simpler set of controls that not only allow the same amount of access, control, and customization, but probably also make it less likely that you're moving your hand all the time to do something, let alone moving your eye from the viewfinder. This is the type of rethink that we need now, and which Nikon themselves need in order to cut out costs on consumer DSLRs (and regenerate interest). 

The Nikon 1 was an interesting experiment for Nikon. It showed that they can think outside the box in terms of future camera designs. We may quibble about their sensor choice and target customer (gotta have pink for those women to buy it ;~), but the engineering team proved that they can take a 2000-3000 piece product and make it a 200-300 piece product, and not lose much in the transition. (I'd argue they didn't lose anything that they didn't inflict upon themselves by making some bad calls on target user priorities: the V2 proves that they can put more [better?] external controls into the same engineering project without compromising the ideals.)

So I believe that the consumer DSLR of the future is going to be more modular, too. Only the modularity will be internal and mostly useful to Nikon and repair technicians, not something that the consumer can do anything with.      

© Thom Hogan 2013 / All Rights Reserved bythom.com  @bythom #bythom