Cameras and Photography Explained

I keep getting asked the AI question: "do I use AI?" Or sometimes "do I think AI will render photography meaningless?"

My answers are "yes" and "no" respectively. And in those answers lies the headline: I sometimes use AI tools to assist in my conversion and processing, but I'm not into generating completely new scenes. (There's also a "middle area" between the two, such as generative expand, where you use an AI tool to mimic an area to extend a scene, which isn't so much making up something new as using what's known to project further.)

But let's examine that second question a bit. The implication behind it is that our Insta and Tok doom scrolls will become full of deep fake landscapes, selfies, portraits, and more. That instead of capturing the news—e.g. photojournalism and broadcast journalism—we'll be illustrating someone's opinion of the news. Real Estate listings would be 100% fake. Food ads will go from their current highly manipulated to totally fake. You wouldn't need to buy new fashion from a clothing store, as you could just swipe it in on your selfies from the Generative Fashion store. Heck, you could probably skin your Kia with some sort of system that allows it to chameleon into a Porsche. 

That's not a real world. That's living in a fantasy world. As much as Hollywood seems to want to have us live in fantasy worlds two hours at a time, I don't believe such a universe is livable 24/7. Way back in 1978 I wrote a screenplay called Labor Day where I envisioned such a existence that was so reversed from reality, that Labor Day was celebrated by actually doing some (unnecessary) labor. The realm the characters lived in, however, was antiseptic and monolithic because ultimately, you can't change the fantasy on everyone every day. That would not only be exhausting, but counterproductive.

I'll remind everyone that painting went through this same crisis (caution: I'm going to extremely oversimplify). Cave art slowly evolved from absolutely crude, symbolic, and mostly unrecognizable to us to something that we do recognize, though still crude. Over time the notion of realism became mainstream. Then in the 1800's we started to see the rapid creative branching that led to impressionism, cubism, and all the other isms. Realism, though, is still clearly present in painting, though (e.g. social realism, new realism, and particularly photorealism). 

Ultimately, the visual arts are all about "made you look." Preferably: "made you look, pause, and study." There's no doubt that Generative Photography will make you look, especially as it gets more and more extreme. However, Real Photography done well and right has always been in the "made you look, pause, and study" category, particularly because it relates a story/place/event that existed and is meaningful, not a fantasy that will never exist. Thus, I believe we'll continue to see high practitioner photographers who understand this and continue to do the Sontag thing ("stand on the shoulders of those that came before..."). 

And those photographers very well may use AI assists to try to ferret out things that mask the underlying reality, such as noise (and particularly color noise). They'll use AI-based autofocusing for times when they can't manually follow a subject fast enough. They may even use generative remove to take something out of a scene as they would have done had they had the ability (e.g. that white RV parked way out there in your otherwise natural scene). I'm all for those helpful uses of AI. 

But just typing words into a browser to get back an fantasy image? No, not for me, and not a photograph.

As I continue to decipher and analyze the 2024 year end data for cameras, both from company financials as well as other sources such as Nikkei, one aspect of the numbers that doesn’t get talked about much is compact cameras. 

CIPA reports that 1,880,414 compact camera units—cameras with integrated lens—were shipped in 2024. Doing some calculations on all the Nikkei, TSR, and company numbers, I come up with:

• Canon — 430,000 units
• Fujifilm — 130,000+ units, mostly X100IImodels
• OMDS — 30,000 units, probably all OM Tough TG-7 models
• Ricoh — 60,000 units, probably all GR3 models
• Sony — 470,000 units

Note that an OM Tough TG-7 is US$550, a Ricoh GR3 is US$1130, and the Fujifilm X100VI is US$1800 (Fujifilm has already tariff adjusted their pricing, OMDS and Ricoh are about to do so). These are prices that rival lower-end (typically APS-C) mirrorless cameras, which seems to indicate to me that there’s still reasonable demand for the shirt-to-jacket pocket camera. Canon has restored some of their PowerShot line, though the most easily available one in the US is the V1 model, which is more video oriented (a new pocket Elph is rumored to come soon). Nikon still seems to sell a few P950 and P1100 superzooms. Panasonic’s ZS99 and FZ80D cameras also seem to be selling. Sony’s RX line was just given a high-end refresh in the RXRIII. 

The question I have is this: are any of these options the “right” design for a serious photographer? 

Maybe?

The just announced (but more expensive at US$1500) Ricoh GR4 seems pretty good in terms of performance and function, though it still looks a little too old-school compact with it’s controls. The Fujifilm X100VI is my current choice, though it has a lot of overkill for my needs and I almost never use the fancy optical side of the viewfinder. Since the US$1500-US$1800 price seems to be where “highly functional, strong results, but still pocketable” seems to live now, is there a better camera definition that fits the serioius user space? 

Probably. 

The “base” that feels right to me is 26mp APS-C, large articulating OLED Rear Display (brighter in daylight situations), and 24-50mm (equivalent) f/2.8 zoom (a short zoom answers the 28/35 and lens adapter questions). We can argue about whether an EVF and flash is required, and what storage is used. But state-of-the-art subject detect focus and as much user control/customization as possible need to be present, too. Speaking of which: buttons you can find and distinguish with your fingertips, even with thin gloves on are also required. I’m not going to do pro-level video with this product, but it should produce good looking 4K/30 when needed. 

There’s clearly a 50-100k annual volume that would be obtained by such a product in the price range I note. That’s a very viable product. So why haven’t Canon, Nikon, or Panasonic produced one? 

Ironically, the 2013 Nikon Coolpix A (16mp APS-C) these days sells for US$500-600 (before any tariffs if it’s coming in from overseas) in good condition, and is still a highly viable compact camera. If Nikon simply stuck their current 20mp sensor and EXPEED6 chip in and changed nothing else, a Coolpix A+ would be a best seller at US$1200, maybe more.