The Sensor Wars, Cost Versus Tech

I started out writing a different article—which I may yet complete—but got side-tracked on something that struck me: Canon is using Nikon's old sensor strategy. 

What am I talking about? Well, Canon now has three mirrorless cameras and as many DSLRs all using basically the same 24mp APS-C dual-pixel sensor. There's rumor that the GX compact line will switch to that sensor soon, too. This is similar to when Nikon fixated on repurposing the 6mp DX sensor across quite a few models.

The reason to do this is simple: cost. Canon is now in the position of making millions of basically the same sensor, potentially reducing their sensor costs both from an R&D standpoint as well as benefiting from yield improvements and a single inventory item to stock/track. 

Nikon, for some reason, abandoned this approach, and at one time had different sensors in every one of their APS-C cameras. We've seen them back-track a bit with the 20mp sensor in the D7500 and D500, but you have to wonder what Nikon is doing here. By my count, we have six different sensors in their ILC bodies (20mp 1", 20mp and 24mp DX, 20, 24, and 45mp FX). And that's being generous, as it seems that Nikon has kept a few older models still in production—perhaps to use up sensor commitments. In terms of what Nikon currently sells as new, there are at least five different DX sensors in play, but strangely, not particularly differentiated.

Sony Imaging, meanwhile, seems to have gone fairly all in on the sensor tech side, with by my count at least nine different sensors in their RX through Alpha models, very differentiated by technical engineering changes in each one.

This is the conundrum.

Canon is clearly looking to optimize volume. By consolidating sensor use in the lower end products and developing automated factories, they're tackling the broader consumer market with some prodigious cost cutting. They've already got 45%+ market share in this realm, but now with even clearer cost advantages. They can use price as a lever to maintain that share. 

Sony, obviously, made a choice after NEX and SLT to back up and pursue high margin over volume, and they decided that meant that they had to be continuously vested in upping the sensor tech. They've been first to a lot of things on the sensor side, but their camera prices had to go up in order to do that. 

Nikon seems to have tackled a similar approach as Sony on a couple of bodies (D500 and D850 are definitely sensor R&D projects with new payoffs, while the D5 was really more of a simple refresh of the D3 sensor), but not all. The problem for Nikon is that their lower end products are just not selling. They made one of the most lukewarm update cycles with the D3400 and D5600 I've ever seen, and there wasn't really anything at the sensor driving those updates. The Nikon 1's haven't been updated in two years. The DLs were cancelled.  So what's Nikon's strategy here?

Of course, there will be those pixel peepers that chime in about differences in dynamic range, frame rates, and a bunch of other stuff. But frankly, I think Canon has got this right. There's a reason why the Canon EOS M5 is my pocket camera and not a Sony. The 24mp Canon sensor may not knock it out of the park in signal-to-noise ratio, but frankly, it's more than enough for almost any use I'd make of a smaller camera. 

Indeed, at the 1" to APS-C sensor sizes, people aren't really buying those cameras for deep dynamic range, but usually for something else. Which brings up a point: where is "entry level" for sophisticated compact and ILC now? 

Hmm. Right where Canon seems to be aiming: 24mp APS-C with excellent but not perfect image quality. 



Looking for gear-specific information? Check out our other Web sites:
mirrorless: sansmirror.com | general: bythom.com| Z System: zsystemuser.com | film SLR: filmbodies.com


dslrbodies: all text and original images © 2024 Thom Hogan
portions Copyright 1999-2023 Thom Hogan
All Rights Reserved — the contents of this site, including but not limited to its text, illustrations, and concepts, 
may not be utilized, directly or indirectly, to inform, train, or improve any artificial intelligence program or system. 

Advertisement: