News and commentary about the mirrorless camera world (latest on top). Click on News/Views in the gray menu bar above to see the full list of recent articles and folders containing older ones.
(news & commentary)
Back in the days when Fujifilm had (Nikon-based) DSLRs, they targeted the forensic photography niche with cameras that had wide, unfiltered spectral abilities, from UV to visible to near-infrared light.
Today Fujifilm announced the X-T1 IR, which is a version of its top-end mirrorless camera that does the same thing. Said to be responsive in light from 380 to 1000nm, the IR version is otherwise identical to the original X-T1. Cost will be US$1700 for the body.
What I don’t quite understand in the Fujifilm press release is “pairs seamlessly with each high quality Fujinon XF lens.” I find it unlikely that the existing XF lenses are very good at transmitting UV light. Generally, glass and the coatings used on glass for photographic purposes, tends to be a UV filter. Likewise, internal lens design is a potential issue at near and deeper IR values, as you get ghosting. It’s unclear to me whether Fujifilm is just saying “works with all existing Fujinon lenses” or they’re suddenly implying that the Fujinon lenses were designed for UV and near-IR work from the get go. Somehow I doubt that. It’ll take some testing to figure out what works best and what doesn’t work.
Many of us who explored UV shooting—including on the original Fujifilm forensic DSLRs—resorted to Nikon’s UV lens or other lenses where we stripped coatings off to increase UV transmission through the lens. Put another way, for UV in particular, yes, some UV will get through the lens, but your exposure may need to be higher to usefully capture that.
I’m also a little surprised that Fujifilm didn’t announce corresponding filter sets for the new X-T1 IR (they did with the previous UV/IR cameras). What you really want are four filters: (1) UV pass only; (2) visible pass only; (3) IR pass only; and (4) visible plus IR pass. This would allow you to use the X-T1 normally, and in the three additional spectral modes that are most useful.
Support this site by purchasing from this advertiser:
When set to perform Post Focus, the GX8 (and some other upcoming models and firmware updates) shoots 50 images at 30 fps with the focus point changed in each one. This function requires a DFD capable lens (Depth from Defocus). When a Post Focus image is reviewed, you tap on the point you want in focus, and the camera creates a new image with that point in focus. In essence, the camera is doing a focus rack during the 50 shot sequence, then letting you pick the appropriate image from any of the images shot.
It strikes me that this is the least useful of the possible uses of this function. A more useful option would be focus stacking, or using the individual shots to create a deeper of very specific depth of field, as we often do with series of shots in macro.
While Panasonic’s firmware is not yet complete, it appears that the 50 images do stay on the camera, so we can probably use them to focus stack after the fact. However, Panasonic really doesn’t have anything useful in the way of a macro lens that understands DFD, which apparently is necessary for this feature to work. It would have been more interesting to see a new longer focal length macro lens (say 60-80mm) being developed than yet another “normal” focal length lens (e.g. 25mm f/1.7).
It really feels to me that camera designers are tracking each other’s features rather than developing true photographic improvements. It’s the old Marketing Check Box game in action. The “touch to refocus” aspect of Post Focus appears to be a response to Lytro. Why? I don’t know. Lytro isn’t exactly setting the world on fire. Indeed, if Lytro is a match, someone dampened its head and then put it in the freezer. So why imitate that? Why not target a thing that photographers are actually doing (focus stacking) and more would like to do?
(news & commentary)
The m4/3 world just hatched another camera today: the 20mp Panasonic GX8, a rangefinder-style camera with a number of interesting aspects.
A lot of people are getting hung up on the 20mp aspect of the camera. While I always welcome a new sensor, given that sensors in general tend to make substantive improvements about every three years, note that this just nets you a 13% resolution gain, generally not enough to be visible to most people.
Compared to the previous GX7, the new magnesium allow body design features a more pronounced right hand grip, and a redefinition of the top plate controls with what I can only call a modern retro feel (similar dials to retro designs, but modernized). The body loses the built-in flash, but gets an almost lens-aligned hot shoe. The body is dust and water resistant. Note that the GX8 is bigger than the GX7 in every dimension, and a bit heavier, as well. It also uses a BLC-12 battery (330 shots CIPA).
The sensor upgrade also nets 4K video at 30 fps, as well as improved DFD style focus that now includes a starlight mode capable of shooting at -4EV. The sensor features an electronic shutter, so silent shooting is possible, as is shutter speeds up to 1/16,000.
Like the GX7, the GX8 features sensor-based image stabilization, and integrates with Panasonic’s lens IS system for a “dual IS” approach.
Rumors that the GX8 would be LX100-like in its controls were a bit off the mark, as we have a Mode dial on the GX8 and no shutter speed dial. Still, I get the reason why folk who used the prototype spoke this way: the GX8 has a fairly simple, direct, and photographer-centric control system. The tilting EVF has been redesigned, and the LCD touchscreen is now a swivel instead of tilt-only.
Price went up a bit, too, something we’re going to have to get used to. At US$1199 the GX8 is US$200 more expensive than the GX7 it succeeds. I suspect that these price bumps we’re now seeing are mostly artificial. Since the expectation now is that any camera that’s been out for awhile has discounts on it, by starting at a higher price the camera companies can still appear to be giving large discounts, but better retain their product margins when they do so. Here in the US, everything looks very artificial in that respect though: in the time between the GX7 and GX8 the implied price should have gone down 20% due to the yen devaluation, not up 20% as it appears here.
As part of the GX8 announcement, Panasonic also announced development is in progress on two new m4/3 lenses: a 100-400mm f/4-6.3 and 25mm f/1.7. These lenses introduce a new function called “post focus”, which is similar to Nikon’s BSS: the camera shoots a burst in 4K resolution while shifting the lens slightly, then after the shot you can select the one that is best in focus. I’m more intrigued by the statement that the 100-400mm lens is DFD capable with 240 fps image data. This means it should be on par with the fast focus we see with Panasonic bodies with the shorter focal length lenses. The other nice thing about the new telephoto zoom is that it has a tripod collar (the current 100-300mm does not). No details on other specifications, date of release, or price were mentioned.
Support this site by purchasing from this advertiser:
(news & commentary)
Fujifilm today produced a 4.0 firmware update for the X-T1 mirrorless camera that has quite a few improvements to it, and makes the camera a more responsive one, especially with focus on moving subjects. Specifically:
- Zone and Wide tracking modes were added to the continuous autofocus system
- Autofocus accuracy was improved
- Eye detect autofocus has been added to Face detection, with the ability to recognize the closer eye, as well
- Autofocus during video recording has been improved in a number of ways
- You no longer have to enter into macro mode manually to get close focus with lenses that support it
- Exposure compensation control is allowed in Manual exposure mode
- The grid lines in the viewfinder have been scaled down to be more subtle and less intrusive
- Silent mode is now named Sound & Flash Off
- Shutter speed can now be set from 30 seconds to 1/32,000 via the front command dial
- Shutter speed can be set to electronic or mechanical
- Spot metering can now follow the focus area
- You can lock or selected settings
- Shutter speed and aperture can be adjusted while recording video
- 50, 25, and 24 fps options were added to the video settings
- The Quick menu can be edited
- Classic Chrome has been added to the film simulations
- Multiple white balance adjustments can be made
- You can change the hue of the display
- A Connection Setting option was added to the Setup menu
- A USB mode was added to the Setup menu
As others have reported, this big firmware update basically makes the XT-1 a better camera. Indeed, the camera's personality for autofocus is far better than before: it seems to track subjects with more accuracy as well as more speed. Most of the time for moving subjects you’re going to want to stick to mechanical shutter only and turn off face detection (which enables the contrast detect system and adds a bit of delay). Interestingly, the camera now beeps to confirm focus before it can actually update the EVF status to green for focus.
Time to stir up the hornet’s nest a bit. We can’t move the nest away from the house unless we get all of its occupants out and buzzing about, after all.
I’ve heard all kinds of reasons why people buy mirrorless cameras. Some are reasonable, some are a bit misstated, some just aren’t correct.
Most of the time, the mirrorless camera purchaser makes their comparisons to DSLRs, so I’ll stick to that here, but I’m seeing more and more “convenience mirrorless shooters”—one lens with a lot of focal range stays on the mirrorless camera all the time—so technically to defend a mirrorless camera purchase you’d also have to compare against all-in-ones, such as the Panasonic FZ1000 or Sony RX10, for example.
Here’s the thing: the primary reason why someone buys a mirrorless camera is that the DSLR makers are perceived as not providing something that the purchaser wanted. That’s it. It’s not because mirrorless is “better” than DSLR. It’s certainly not that mirrorless is “cheaper” than DSLR. It’s not because the mirrorless makers have been in the ILC camera business longer (though Leica might make that claim). It’s not that mirrorless cameras are simpler and easier to learn to use than DSLRs (I’m looking at you Olympus).
I’m sorry to say that Canon and Nikon, the two primary remaining DSLR vendors are moving about like a sumo wrestler trying to run a marathon: slow and awkward. Occasional oddly positioned experiments aside—the Nikon 1 comes to mind—the duopoly in ILC cameras just isn’t delivering everything shooters want. The seven dwarves correctly figured that out and came out with a way to appear to create a niche that delivered many of those things: mirrorless.
Here are the primary reasons why people buy a mirrorless camera:
- Size and Weight. Everything is conspiring against large size and heavy weight these days. Airlines. The aging population of SLR/DSLR shooters. The young not being quite as consumption-happy as the Boomers. The list of reasons is fairly extensive. One of the things that high tech does exceedingly well is put additional capability and quality into smaller and smaller products. That's pretty much an expectation from consumers these days: products get more compact, yet better. The DSLR makers ignored this for far too long. Even when Canon tried the smaller, lighter SL1 DSLR, they got something incredibly wrong: the SL1 was perceived as a less capable camera than its larger Rebel brothers. Nope, you have to get both things right: better and smaller. Indeed, that’s one of the things that held m4/3 back for awhile: it was perceived that the small 12mp sensors just weren’t delivering what the constant parade of more pixels in APS and full frame were.
- Size of System. This is related to the first, but a little trickier. I’ll get to the tricky part in a moment. First, let me point out that an m4/3 body, 24-200mm (equivalent) worth of fast zoom lenses, and a couple of useful primes fits in an incredibly smaller bag than does a full frame DSLR with the same things. If you really want the advantage of an interchangeable lens camera, that means that you’re carrying lenses with you besides the camera. So while the size of the camera itself is important, what’s actually more important is the size of the system you’re likely to carry. A large, heavy backpack is less convenient than a smaller, lighter shoulder bag, for example.
Of course you give something up in getting the smaller system: smaller sensor, sometimes crammed hand positions for controls, and the need for faster aperture lenses to keep various equivalence factors at bay, amongst other things. But I can attest to the fact that even going out to shoot wildlife with an Olympus E-M1 produces a significantly smaller and lighter bag than doing so with my D810.
I said this point was tricky, and the Sony A7 is exactly where that tricky bit comes into play. The A7 series uses a full frame sensor. It needs lenses with larger image circles, which means more glass (larger and heavier). Technically, a 24-70mm f/2.8 zoom for the A7 is pretty much going to be as big and heavy as a 24-70mm f/2.8 zoom for the Canikon DSLR full frames. Sony decided to produce a smaller 24-70mm f/4 instead. I’d also argue that this lens shows some further design compromises to keep its size down. it’s a decent lens, especially in the central area, but it relies on software lens corrections fairly heavily to rid it of corner issues.
That said, the fact that Sony understood that it wasn’t just body size, but system size too, that was important to consumers shows that Sony might have been listening better to the customer base than Canikon. Actually, I personally think Canon and Nikon heard the same things, but they’re just acting like lazy duopolists.
- Better autofocus. Surprised? Let me explain. Contrast detect autofocus systems (most mirrorless cameras) have one advantage over phase detect autofocus systems (DSLRs): they are extremely precise in focus plane. Phase detect is fast at moving focus—especially with tracking fast subjects in motion—but it suffers from a lack of absolute precision. For a static subject a DSLR gets focus very close to perfect extremely fast. A mirrorless system gets that same thing precisely focused but not quite as fast (especially as we move into telephoto focal lengths or long focus moves). It also doesn’t help that most DSLR focus systems are sophisticated enough that it takes dozens of pages to figure out exactly how they work and a half dozen settings you need to dial in for any given subject.
Thing is that most people are taking photos of non-moving (or at least not fast moving relative to the camera) subjects. Aunt Betty standing at the rim of the Grand Canyon, for instance. Cousin Billy climbing his backyard jungle gym. The new friends you met on the cruise that are sitting at your table. In almost all the most common photography situations, just using Single Servo focus on a mirrorless camera is going to dial in the focus plane with more precision and fast enough. That’s especially true when you consider you can position the focus point almost anywhere—and often via touching the LCD these days—whereas the DSLRs limit your focus point choices to a smaller central area.
To be honest, it isn’t really “better autofocus,” but “more usable autofocus for what the customer typically wants to record.” While some of the mirrorless cameras are closing the gap on overall focus performance, the top DSLR autofocus systems are very fast, very flexible, reasonably precise, and track even fast-moving subjects exceedingly well when set right. In some cases, shockingly well now that some DSLRs are mixing color information with focus information.
- EVF. Some of you reading this will be surprised to see this in the list. You’ll argue that the EVF has a lag to reality, and that even in its current XVGA form it is a bit crude, especially in low light. But the reason that EVF is on this list is the same reason that DSLRs took over so fast from SLRs: the broad consumer base is terrible at making good photographic decisions without help. With film, they’d take pictures and get them back a week later, and they had no idea what they did wrong. With DSLRs, they could at least chimp the shot afterwards and see that they got something wrong. With EVFs they can see things like exposure problems in real time, closing the feedback loop completely (at least on well designed mirrorless cameras). They can even see special effects instantly, picture styles, and a host of other things they might want to set. My Olympus OM-Ds, for example, can be set to show where shadows are blocking up or highlights nearing clipping in the viewfinder in real time. I can’t get my Nikon DSLRs to do the former even while chimping the shot ;~).
Not everyone wants to spend hours, days, weeks, or years learning how to make all the right settings the right way without any instant feedback. Indeed, the majority want just the opposite: show me instantly what I’m doing so I can fix it immediately.
I’d also point out that the viewing screens in DSLRs have slowly but surely been stupidified by camera makers. You wanted better focus, so they gave us autofocus systems. Then you complained that the viewfinders were too dim, so they took out all the focus aids and concentrated on just brightness. Frankly, it’s easier to see if the camera has actually focused with a mirrorless camera than it is with a DSLR these days. Mirrorless cameras with peaking settings make that absolutely obvious, too.
Again, I’d say that the DSLR camera makers were lazy. The Nikon Df is a good example. If there’s any DSLR made lately that’s likely to be manually focused, that’s the one. Nikon made a slight adjustment to the viewing screen, but did any of the manual focusing aids return? No.
Which brings me to…
- Technology. The mirrorless camera companies are now perceived as being the ones incorporating the most new technology into cameras. Sensor-based image stabilization, multiple-shot high resolution, 4K video, focus peaking, real time view of long exposures, and more (including that EVF I just mentioned and all its new capabilities).
DSLRs are now being perceived as iterating dinosaurs (hmm, Jurassic World, anyone?). All the “cool technology” is happening in mirrorless first.
Okay, but I’m going to temper that a bit. First, the mirrorless camera companies have to innovate and move into new technologies fast. They only have 23% of the market (2014 numbers), and it’s actually worse than that when you consider that Canikon have some of the mirrorless market, too. If the mirrorless camera companies competed solely on price or performance, they’d be crushed by the two big companies. They have no choice but to lead the R&D parade and discover the technologies that really should be in cameras.
But consumers are easily fooled by technology, too. Take image stabilization. This is a “good enough” solution not a “best practice” solution. I’ve written for over a decade the following: turn off IS and stabilize your camera (e.g. well chosen and set up tripod) and you’ll almost always get better pixel level results. IS/VR systems are moving something (lens elements, sensor) during the exposure. And they’re not moving those elements perfectly in sync with the motion. Not even close. While you might not be able to easily measure it, you’re losing edge acuity with IS systems turned on. Of course, if the alternative is to get totally smeary results because you can’t hold the camera steady, then IS is certainly useful.
Fujifilm, Olympus, Panasonic, Samsung, and Sony are pushing interchangeable lens cameras forward with their mirrorless entries, no doubt. They’re doing so faster and better than Canon and Nikon are with their DSLRs. (Pentax is a mixed case, somewhere in between, but probably closer to Canon and Nikon given how slow they’re moving, even with new technologies.) The reason why you consider and possibly purchase a mirrorless camera is generally because of this progression. The DSLR makers aren’t giving you what you wanted, or not doing it well enough or fast enough, so you look at the alternative.
Whether you buy a mirrorless or DSLR camera really has very little to do with the presence or lack of a mirror, as it turns out (witness the Sony SLTs, which are in between). It’s really about whether a camera company is providing you with what you want.
In contradiction to that, there’s the inertial pull of 200m+ EOS and F-mount legacy lenses out there. After all, you’re buying an interchangeable lens camera whether you go mirrorless or DSLR. Thus lenses are an important consideration. This is why both Canon and Nikon have made incredible mistakes in their crop sensor DSLR systems: they didn’t fill out the lens lines. That’s allowed the mirrorless companies time to not only create new lens systems of their own, but in the case of Fujifilm and m4/3, arguably fill those lines out pretty well. Samsung and Sony aren’t far behind.
Realistically, you’re buying mirrorless systems because Canon and Nikon aren’t giving you what you want, simple as that.
(news & commentary)
Sony today announced the 42mp A7r Mark II camera, the expected update of the original high-megapixel full frame mirrorless body. Along with the bump in pixels came a bump in price. Indeed, I thought at first it was a typo, as the A7r price was US$2300 and the A7rII price is now going to be US$3200. But it appears Sony is trying to push the price up as they improve the camera.
What’s improved? Well, like the A7II, we get the new 5-axis sensor-based image stabilization and basic body redesign. Also added is phase detect focus points on the image sensor (399 of them) and the ability to create 4K video (100Mbps XAVC S), and from the full frame width. That said, the camera also supports a crop (Super35mm) mode that doesn’t use pixel binning, reducing moire. 720P shooters can set 120 fps for slow motion. Sony’s S-Log2 gamma control comes along for the video ride, too. Curiously, the sensor uses BSI (back side illumination), a first for full frame sensors, but not likely to provide much in the way of a tangible benefit in terms of image quality.
The mechanical shutter—a source of significant vibration issues with the original model—has been redesigned, with Sony stating that it produces 50% less vibration than before, and is rated to 500k shots. An electronic shutter mode is available for silent shooting.
The new hybrid phase detect/contrast detect focus system coupled with the faster data refresh off the image sensor allows the A7rII to shoot with focus at 5 fps (the previous version was limited to 1.5 fps with focus, 4 fps without).
It seems odd that Sony not only announced this camera a full two months prior to actually shipping it, but did so at 2am Tokyo time via press release (apparently because of a 1pm New York briefing to press). Thus, much of what we know about the camera is still basically just from the scant information Sony released at the moment. The A7rII announcement came with the RX100IV and RX10II announcements, but both of those cameras are expected to ship in about a month. Frankly, I think Sony actually goofed on this, as the A7rII is quite likely to bury the RX100IV and RX10II in the photography media coverage. Deal with the 1” sensor cameras this week, and the full frame later. Update: as someone pointed out to me, it’s possible that Sony thought to try to produce some FUD just as Canon was shipping their two 50mp 5D’s to customers. I’ll stand by what I wrote, and if Sony starts playing the FUD game, they probably can expect to get FUD right back at them when they eventually ship the A7rII in August ;~).
The news and views for 2013 by month from sansmirror.com:
- December 2013 Mirrorless Camera News
- November 2013 Mirrorless Camera News
- October 2013 Mirrorless Camera News
- September 2013 Mirrorless Camera News
- August 2013 Mirrorless Camera News
- July 2013 Mirrorless Camera News
- June 2013 Mirrorless Camera News
- May 2013 Mirrorless Camera News
- April 2013 Mirrorless Camera News
- March 2013 Mirrorless Camera News
- February 2013 Mirrorless Camera News
- January 2013 Mirrorless Camera News