Cameras and Photography Explained

LEDE ON

Who was it that wrote that the Trump-mandated tariffs were unconstitutional and that they wouldn't reduce the trade deficit? Oh wait, that was me, back when the "tariff war" was first declared. Now that the Supreme Court and the full year trade numbers has confirmed what I wrote, much of it looks pretty prescient. Yet it was just applied basic macroeconomics and an actual reading of our Constitution. Of the nine bulleted items I listed would happen, six have become arguably true, the other three are just still arguable.

So now we can expect all the camera makers to reduce their prices and you get issued refunds, right? That's where human behavior takes over from macroeconomic theory. Congress has this way of, once they have your money, not giving it back. Meanwhile, the camera companies didn't see the end of the world after having to raise prices, just a lowering of their profit margins, so what's their motivation to lower prices again? 

Don't get me wrong. I do expect further price adjustments moving forward, but I'll bet those will come mostly in the form of constant and repeated sales rather than the permanent lowering of list prices. I had a boss once whose mantra was "price high, and discount if you have to, because it's near impossible to unilaterally raise prices, but real easy to unilaterally lower them."

——————————
Analysis

Taxing Times

You didn't think I'd stop with just the Lede, did you?

We're almost a year after all the original tariff turmoil and guess what, we're back in choppy waters. Let's start with de minimus. That's the part of tariffs that applies to <US$800 items sent directly to customers. Until last year, the idea was to eliminate cost at the Customs department by eliminating the tariff on low-priced items coming in directly to the customer. Then the Trump administration cancelled de minimus and imposed new tariffs by country. While you might think that the Supreme Court threw all that out with their decision last Friday, Trump said that he reinstated tariffs (see also next paragraph). What you probably don't know is that the Big Beautiful Bill that Congress passed last year officially ends de minimus in 2027, so tariff collection will likely start on small overseas purchases regardless of what's happening with the "by edict" Trump tariff declarations.

While the Supreme Court was pretty clear about the way tariffs were declared last year and said, no, that's a Congressional authority, not Executive Branch one, Trump within hours imposed new tariffs, using a different mechanism. Section 122 of the Trade Act of 1974 allows the President to impose up to 15% tariffs for as many as 150 days when there are large trade discrepancies. Trump used that authority to immediately impose 10% tariffs on everything. Then someone in the White House actually read the section in question, realized that they had already agreed to 15% tariffs with a number of companies such as Japan, and the next day Trump said, "oops, I meant 15%." 

So, two things about that: (1) Congress will have to act within 150 days to extend these new 15% tariffs, and that's unlikely in the current environment because it would happen just before the next election; and (2) the reasoning is once again problematic, and will likely get tested by the courts as once again exceeding the authority given by Congress. Section 122 doesn't actually refer to "trade deficit" it refers to "payment deficit", of which the US has very little (the US heavily exports "services", which balances out "trade"). In essence, we have an autocratic regime that can't read and has indicated that they will do what they want and keep trying to find some methodology that might make tariffs stick. In other words, businesses still have no idea what their costs are (and will be), as they keep fluctuating as the tariff battle continues. 

I mentioned before that supposedly one reason for imposing the tariffs in the first place was to reduce the US trade deficit. Other than a brief dip in October, we seem to be pretty much back to "normal" with our physical goods deficit. The only real change that is clearly due to the tariffs is where the goods are being imported from (e.g. not China, though how much of the increase from places such as Vietnam is really just "shuffling" is unclear). We're still paying more for all those goods at the retail side, because, well, someone has to pay the tariffs, and guess what, it isn't who Trump says it is, it's ultimately US consumers.

From the camera maker standpoint, things are back to chaos. Sure, 15% may be less than what they had been paying (or not in several cases). But it's also far less than the previous China tariffs, as well, reintroducing the potential for stronger competition with Chinese lenses again. And, no, no one is going to be reimbursed for unconstitutional tariffs any time soon (if ever), so it's no different than if a parts supplier changed their pricing (and may change it again) and insisted on keeping the money you've already sent them. Note that every camera maker said their profits were reduced by tariffs in their last fiscal year, but that's not going to go away in the foreseeable future.

As much as I've maligned the bean counters in Tokyo in the past, they currently have my sympathy, because they're trying to run their spreadsheets when the inputs keep changing in (mostly) unpredictable ways. Coupled with Japan's 30- and 40-year bond yields now spiking upwards—which has implications on corporate borrowing—the accounting side within all the Japanese camera companies has to be nervous and twitchy at the moment.

For those of you outside the US who probably think that the whole tariff situation doesn't impact you, you'd likely be wrong. With a quarter of sales for companies directly impacted, plus the net impacts of some negotiated deals that have been made, it's probable that pricing is being impacted worldwide, just not nearly as much outside the US as inside. Product decisions are being delayed, as well, because pricing impacts aren't certain. 

My advice to this site's visitors really doesn't change. First, are you sure you need new gear, or are you just a GAS (Gear Acquisition Syndrome) addict? If you need something, then buy on sales if you can or perhaps consider refurbished gear, but don't believe that "things will get cheaper in the future." They might, but that's far less likely at the moment than likely. 

Meanwhile, I'm reminded of a Churchill quote: "We contend that for a nation to try to tax itself into prosperity is like a man standing in a bucket and trying to lift himself up by the handle." Unfortunately, no one seems to be able to pry our leader out of the bucket.

——————————
Analysis

I'm Bullish and Bearish on AI

The mainstream media and the big tech companies investing in AI certainly are making loud noises, much like elephants fighting with hyenas. Most of what you hear and what you're afraid of because of those terrible sounds is going to result in a lot of nothing, however. 

The trillions of dollars being invested in server-farm-based, one-company-owned AI systems is not likely to pay back returns. I see distinct parallels to the so-called Web 2.0 run up at the turn of the last century. A lot of cacophony followed by a lot of merrymaking players going poof. 

On the other hand, the serious investment and training of AI models is already paying off big. I, like many of my Silicon Valley friends, run my own LLM (large language module) on my laptop. Yes, I said laptop, and yes, I mean it runs entirely on my laptop. What do I use this for? Sort of as an intern/assistant. While working on my new Web sites, I've had my LLM offer up coding suggestions and verify code, for instance. Sometimes I'll throw an idea at it and see how it responds. But I'm doing the work, I'm just looking at ways to do it more productively. And I can say that it works. Quite well, actually.

If you're a betting person, you can make a bear or bull bet on AI today. 

Meta is a bear bet for me. I see no clear way they can pay back their growing investment in AI, just as I saw no clear way for them to pay back their huge investment in VR. The closest they come to a payoff is the Rayban/Oakley glasses, but because that's an isolated product for them, I don't see it generating the ROI that is needed. Meta's primary source of revenue is selling you advertising on their social media platforms, stet.

Apple is a bull bet for me (and my laptop with the LLM on it is a MacBook Pro). Apple makes hardware, and specifically, they make hardware that takes advantage of technical advances to give users clear new and useful capabilities. While they poke around with centralized AI with Siri, et.al., the real benefit for Apple is if they make hardware that runs AI efficiently and user-useful ways. Which is exactly what is happening and I predict will continue.

What does this have to do with photography? Well, Topaz Labs is an example of them making a bet on my bull side: they want their AI models to run on their centralized servers for which they charge a toll. Frankly, their local model AI (Topaz Sharpen AI, for example), worked more than well enough for me. If anything, they've muddied their models and made them near impossible to use in any productive way, but they'll be happy to sell you use tokens if you want to try using it.

Adobe, meanwhile, is dipping toes in both waters, with both some on-machine modeling as well as centralized. That probably is going to work out okay for them, because, yes, there will be times when my laptop just isn't going to be able to handle some really sophisticated requests and needs a big GPU farm somewhere to do all the computing. 

PetaPixel recently published an article titled "The First Mainstream Manufacturer to Put AI in a Camera Will Regret It," apparently not realizing that all of our current cameras have localized AI models working in their autofocus system (and in other areas, as well). Which gets to part of my point: there's much that AI can help with that's very useful. It can now focus my camera faster and more accurately than I can (though I still override it when it gets it wrong, a much easier process for me to manage). My laptop AI can whip up a Javascript function and test it faster than I can. Adobe Photoshop, both on machine and off, does a better initial selection than I can and does it faster, too. These are the types of AI uses that make sense, and they also often make more sense on local devices (can you imagine your camera waiting for the server to tell it where to focus?). 

Much of the AI hype and fear running around the mainstream media these days is really marketing (and anti-marketing) messaging. Sam Altman is a used car salesman, trying to get you to buy the premium model you don't need. Marc Andreessen is essentially trying to get his investments to make more money, not help you. Elon Musk thinks he can control you by jerking his X controller back and forth. Many of those who helped develop our current AI and LLMs in scholarly venues have cashed in where both high salaries and unrestricted research investments allows them to "live better" both at home and in the lab.

Reality will hit home at some point. It always does with tech. All the promises and gee-whizzing eventually have to have real payoffs for users, or else a lot of (venture) capital disappears in the night. So be careful where you place your bets. New shooter coming out...

——————————

CP+ Announcements

Here come the lenses (more in sight)

▶︎  Tamron 35-100mm f/2.8. At US$900 (E-mount) and US$930 (Z-mount) this is going to be a popular lens, methinks. The MTFs suggest it is very good out through the DX corners, but the real desired trait is the modest size (4.8" long) and weight (20.3 ounces) for a fast aperture zoom that covers all the standard prime focal lengths (35mm, 50mm, 85mm). This is a lens the Z5II and A7V crowd probably can get behind, even though the initial reviewers all seem a little meh about the focal range. But here's something to chew on: with a Z50II this lens is an effective 53-150mm f/2.8, and appropriately sized. While it's no 50-100mm f/1.8 like the old Sigma DC, it does give a reasonable "fast (not quite) 70-200mm alternative for DX users. To bad it doesn't have VC (stabilization). The curious aspect to this is that Tamron also has a 28-75mm f/2.8 lens, so Tamron is seriously overlapping their own offerings now. 

Along with the lens comes the new, optional Tamron Link accessory, which plugs into the lens' USB-C socket and allows you to control it from your mobile device. This US$50 dongle has a 16' range (Bluetooth) and almost no specifics about it in Tamron's marketing and promotions. So let me explain: using the Tamron Lens Utility app on your smart device and using this new dongle you have quite a bit of control over the lens. This is useful for doing repeatable focus pulls in video, setting precise stopping points for any of the control rings, setting how much rotation of a ring is necessary to get full effect and which direction it goes, switching between linear and non-linear ring responses, jumping between two focus positions, focusing during time-lapse recording, setting precise astrophotography focus positions, setting a focus limiter, and assigning functions to the lens buttons. Videographers are going to like this little tool better than still photographers, but given you can also do firmware updates via the Tamron Link and the Lens Utility app, it's probably worth having for anyone with multiple Tamron lenses.

▶︎  Voigtlander 40mm f/2. And if the Tamron is too big and heavy for you (sic), the new Septon from Cosina will attract your attention, as it's a half-pound pancake. Sure, it's manual focus, but with the focusing aids we have these days—including Nikon's MF subject detection—that's not the liability it used to be, particularly for wide angle optics.

▶︎  Nikon 70-200mm f/2.8 VR S II. The second in Nikon's lens remakes for the Z-mount (and not the last), this lens is much like the 24-70mm f/2.8 remake: smaller, significantly lighter, better optically, faster focusing, and with a plethora of small changes that will make you like it more, such as the Arca-Swiss dovetail support on the removable Tripod Collar and the polarizer friendly lens hood. The only thing not to like is the new, higher, US$3199 price. 

▶︎  Nikon 28-75mm f/2.8?. Here's a real curiosity: B&H is listing two versions of this lens (20107 and 20137) at a US$200 price difference, with the higher priced one saying "New Item - Coming Soon."  People are interpreting that as a new version of the lens, but it isn't. The problem is that Nikon had to introduce a new version number based upon the change in country of manufacture (that tariff thing raising its head, again). The only difference between 20107 and 20137 is which country the lens is made in, and Nikon took the opportunity to reprice the lens based upon "current" tariffs. Of course, no sooner did Nikon do that the US changed the Vietnam tariff rate ;~).

-------------------
Wrapping Up

And in other news

▶︎ Cage Fighting. SmallRig seems to be fighting with themselves, as they just announced a new HawkLock cage for the ZR. One key difference is that it leaves the top right plate controls completely uninhibited. Another is the V-mount power option that mounts at the back. 

▶︎ WANDRD Unrolls. Their popular PRVKE pack now comes in two non-rollup models, the Zip and the Pocket. No rollup on top, and the Zip has a zippered back pocket, the Pocket has a larger back pocket for "more stuff." These are generally well-made and well-thought out packs, replete with a lot of nice features. The PRVKE now comes in Roll, Zip, and Pocket versions at 21L and 31L sizes. 

▶︎ Feeling Colorful?. Want to test how well you remember a color? Try dialed.gg. Good luck with that. I had difficulty getting above 98.5% correct. But if you think during image processing you're going to dial in the actual color you saw last week, guess what this little game is telling you? ;~) Sure, you've been capturing reality. But which reality is that?

I'm transitioning all of my News/Views into a new format. For the time being, news will continue in this digest form on this site. Enjoy.  (p.s. If you're interested in potentially subscribing to my new offerings, be sure to click here to receive further updates as that idea gets developed and closer to launch.)

--------------------
LEDE ON

Did the Chinese puncture the Japanese lens tire? The Japanese have a term, attachment rate, which defines how many lenses they sell each year per camera body. The attachment rate was within ±0.02 of an average of 1.67 lenses per body for 11 of the last 20 years. There was an understandable dip in the two post quake/flood years when production was constrained, plus an odd, short upward spike during the pandemic. However, from 2022 to the present there's been a strong and consistent downward trend in the attachment rate: 1.64, 1.61, 1.56, and now 1.51 for 2025. Not coincidentally, the number of prime lenses the Japanese have been selling during that four-year dip has also fallen, and mostly for full frame. Indeed, the drop is more dramatic when only looked at in terms of full frame (a 21% drop in four years). We currently have 55 full-frame Chinese prime lenses that are available for the Nikon Z-mount and even more for the Sony E-mount. While no individual Chinese lens seems to be stealing the attachment rate on its own, collectively the Chinese lens armada is deflating the attachment rate the Japanese camera companies have relied upon for so long. If the ~8% overall decline is really due to Chinese lens sales, that implies that the Chinese sold about 850k units, by the way.

——————————
Analysis

Republishing data is not the same as interpreting it

Once CIPA published their year-end results for 2025, it seems that—at least according to virtually every photo Web site—the big news was the resurgence of compact cameras. Yes, compact cameras went from 1.88m units in 2024 to 2.44m units in 2025, but it seems no one actually got around to trying to answer the question of why that happened.

It happened because the Japanese camera makers simply started making them again. The demand was always there (and there's still more demand than the 2025 numbers would indicate, probably something at least in the 3-4m unit range). The real reason compact camera sales withered starting in 2020 and continuing through 2024 has to do with the pandemic and parts availability than anything else. Each camera maker tended to deal with the supply chain problem differently, and the two that were the most extreme at that were Canon and Nikon.

Canon simply stopped making—but did not discontinue—compact models. Thus, you might have noticed in the last year (plus rumors for this year) that some Canon Powershot cameras are seeing resurrections (though not updates). To my knowledge, Canon never closed or repurposed factories the way others did. For instance, Nikon. Nikon did discontinue pretty much everything except for the top Coolpix models (P950, P1100, and for some reason, the P1000 still seems to be sold). There appears to be little prospect of other Coolpix coming back, though the jury is still out on whether Nikon might add a Coolpix model in the future.

Fujifilm had one of their biggest hits with the X100 in 2013, and the 2020 iteration (V) was the point where it really went viral, so they kept their production up despite parts shortages that hurt other model sales. They doubled down with the X-half and GFX100RF in 2025, too, at a time when most of the other camera producers were not investing in new models.

I don't believe the demand for US$1000-2000 compact cameras is anywhere close to being sated, as the Ricoh GRIV and Fujifilm X100VI backorders show (I've got a GRIV I'll be adding to my compact camera reviews soon). But I also believe there's room in the US$500-1000 space for a really competent compact camera, too, and that's not really being well served by anyone at the moment. 

Let me go out on a limb and suggest that compact camera sales will again show dramatic "growth" in volume when the 2026 results eventually get published. How much growth? Not as much as the market demands, as parts shortages are still being triggered by the unavailability of fab time, plus most of the camera companies have made their design teams more lean. But with Canon back to producing Powershots, we'll see more units shipped in 2026 than 2025. 

——————————
CP+ Announcement

Canon is first to announce

On Wednesday Canon announced two new RF lenses and a reappearance of a Powershot compact in the run-up to the CP+ show later this month in Japan.

▶︎  14mm f/1.4L VCM. This new wide angle optic lists at US$2599 and represents an attempt to lock up the extremely wide, high optical performance prime position before the Chinese start sniping at the RF mount. Some of the early examples from the lens are astrophotography work, so I expect it to be well controlled in terms of coma and other deformities that show up in corners.

▶︎  7-14mm f/2.8-3.5L VCM. And if you want to go more extreme—into the fisheye realm—Canon has a new US$1899 zoom that helps round out their high end zoom optics. At 7mm you get a full circle fisheye (at full frame), while at 14mm you get a 114° rectilinear view.

▶︎  PowerShot G7 X Mark III limited edition anniversary special. This US$1299 compact camera is a 20mp 1" sensor camera with a 24-100mm f/1/8-2.8 equivalent lens. So what do you get for US$400 more than the regular model? A new paint job, a fancier knurled front ring, and a little 30th anniversary (of PowerShot) mark on the pop-up flash.

-------------------
Wrapping Up

And in other news

▶︎ Petapixel isn't an economics site. Once again I have to point out that in their rush to get news out, Petapixel managed to produce clickbait headlines and statements. This time it was "Nikon Posts Big Losses and Cuts Projections." This, of course, always raises the social media negativity towards Nikon Imaging, even though they're not really part of those "big losses," the projection cuts were modest and likely the result of the delayed Z9II, and the group is not in any danger. 

The primary source of those "big losses" for the whole company was a 90.6b yen write down against the acquisition of SLM Solutions to help create Nikon's new digital manufacturing business. Indeed, none of the 103.6b yen loss for the quarter was the result of lack of profitability of Nikon's primary businesses. Cash on hand, receivables, and inventories didn't really drop; what really dropped was 60.5b yen in "goodwill" that was on the books and 26.2b yen in intangible assets. 

What that tells me is that Nikon was over optimistic in how they booked the original SLM acquisition. The write downs are mostly just saying that the 3D metal printer company Nikon bought in 2023 wasn't worth what they paid for it and then booked into their assets. Thus, this past quarter they wrote down their assets to reflect that. The actual cash long ago left the building. FWIW, get ready for a repeat in three months from Petapixel, as Nikon has already said they'll take another significant write down in the fourth quarter (again, none of it having to do with their biggest business, cameras and lenses). 

In the fiscal news that is important to the photography market, Nikon Imaging is showing signs of a slowdown in sales (part of Petapixel's "cuts projections"). Foreign exchange rates and tariffs coupled with the fact that Nikon volume shifted to lower priced models due to recent camera updates is the real story. Overall, that means that Nikon is estimating that they'll still reach the same total sales figure by March 31st as they earlier forecast, but it will produce less profit by about a third. Since Nikon is also forecasting 50k fewer cameras and 100k fewer lenses, for sales to hit their new expected fiscal year result they need to sell more higher end gear and not give back more money via foreign exchange and tariffs. What that suggests to me is fewer sales on lower end gear and more on higher end products. 

I do note that Nikon is going to slip slightly in market share for the year, but as I've already reported, the next generation of top end cameras has been delayed, and that's now starting to show up in the results. I believe Nikon originally thought that they'd drop the Z9II towards the end of their fiscal year (i.e. this quarter), which produced their original estimates. I do not expect the Z9II until this fall (in Nikon's next fiscal year) at this point, and even then it may be a development announcement first, followed by frantic testing to get it out as quickly as possible. 

Nikon Imaging is fine. Not quite where management hoped it to be, but actually still on the path of the long-range forecast they made several years ago. But as I've written several times lately, early 2026 is going to mostly be about lenses, as I see no new Nikon camera ready to to hit any time soon. Frankly, that's fine by me as I'm still trying to catch up with the Z9 generation onslaught Nikon made in the past few years.

▶︎ Godox keeps it going. The perennial Chinese flash cloner has introduced the V1mid, a reasonably small round-headed US$179 flash unit with a built-in Li-ion battery. As usual, there's fairly full functionality (including hi-speed sync) with Canon, Fujifilm, Nikon, OMDS, and Sony specific models available. Also as usual, Godox refuses to publish Guide Number information, which is useful in determining which flash unit you might actually want. In fact, there are a number of "unprovided" notations in their specifications, as if this is top secret information that could fall into the wrong hands. All we know is that the V1mid is not as powerful as the V1. I guess the dox in Godox doesn't stand for documentation. 

▶︎ Next week will get busy. The CP+ announcements start in earnest this coming week, so expect to see a lot more news and commentary coming shortly. 

------------

As you can see, this weekly or bi-weekly "newsletter" style for News/Views can be quite elaborate and lengthy. But by putting everything in one spot less often, in a single format, it saves me time and allows me to spend more time on the commentary than the news itself, which you can get pretty much anywhere.

byThom MAX is still coming, but for the time being I'll be doing news this way. I'll have more about byThom Max when I kick it off later in 2026. In the meantime, if you're interested in subscribing, click here to receive updates.