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.
We’re now deep into the firmware updates for most of the mirrorless systems. Because these systems all use electronics in the lenses, we have to update both camera and lens these days. This is becoming a hassle in more than one way:
- Every new lens tends to require that the camera bodies get updated (tables of lens conversion information and focusing performance, for example).
- Every new body tends to require that the lenses get updated.
Those two things alone start to become a real issue once you’ve got more than one body and two lenses. Every couple of months or so I find that I now need to sit down and do a full check of every body and lens firmware number for each of my mirrorless systems lest I manage to let something slip through the cracks and not get updated. We need a more monolithic way of updating products.
Why can’t we have one firmware update with everything on it? When you do the update procedure on your camera, it would then look in that file and see if there’s a camera update available and ask you to perform it. Once done, it would look at the lens mounted on the camera and ask you to update that, if necessary. Once that is done, it would prompt you to mount each of the other lenses you might have for firmware inspection and update, if necessary. Voila. One and done. No need to keep track of all the files and figure out what is what. Which brings me to:
- Firmware update names suck
I’ll use Fujifilm for an example since I just did a bunch of Fujifilm updates, but pretty much every maker has the same or similar problems. What the heck does XFUP0004.DAT refer to? Okay, I get the XF and UP bits (it’s a lens update for an XF lens). But 0004?
Because I teach photo workshops, I tend to carry a full set of updates with me into the field. Too often I’ve found that a student hasn’t had the latest update installed (in some cases, even I haven’t! ;~). Sometimes that turns out to be a factor in something we’re trying to do. Okay, so I have a folder with Fujifilm firmware updates in it. You guessed it, there’s XFUP0001.DAT, XFUP0002.DAT, XFUP0003.DAT, and so on. Which is which? Worse still, I had an XFUP0004.DAT file already and the new lens update was also named XFUP0004.DAT. If somehow I get both files onto my computer, how would I know which is the latest update?
Here’s the problem: the Japanese camera makers are still living in a DOS world. Why, I have no idea. Even Microsoft eventually managed to figure out how to extend 8.3 filenames into longer, more meaningful ones. Still, even if we assume that the filename has to be 8.3, a better naming convention would have been LUPxxyyy.DAT, where xx is the wide focal length or prime focal length and yyy is the telephoto focal length or nonexistent for primes. As in LUP1855.DAT. Hey, that would be a lens update for the 18-55! (By the way, why the file extension isn’t FUP for Firmware UPdate, I don’t know; that would give you back a couple of characters to use intelligently, and LUP [Lens UPdate] as an extension would give us three back.)
We need to abandon (or significantly update) DCF conventions. As long as we’re locked into 8.3 filenames, cameras are going to have increasing problems with the rest of the world, where we don’t have that restriction.
Here’s the thing: not a single one of the camera makers has realized that one of their users’ problems is recognizing what’s what with files. The old DSC_####.JPG convention tells us very little (it’s a still image in JPEG format). The firmware update conventions tell us very little (in some cases no more than it’s a firmware update!).
The DCF standards were wrong in the first place; if something is in the DCIM, or Digital Camera IMage, folder and has an extension of JPG or one of the known raw formats, we know it’s an image, so we don’t need the DSC part (Digital Still Camera). But the camera makers seem hellbent on not making anything even remotely user understandable. Heaven help you if your camera records AVCHD compatible video. The folder proliferation and naming conventions for that (derived from BluRay) are convoluted and arcane.
It’s time for DCF to basically say:
- Cameras create first level folders with meaning: STILLS, VIDEO, AUDIO, FIRMWARE, MISC (note that I made all those fit in 8 characters ;~), and then put the appropriate thing in the appropriate folder
- Extensions are used to determine file type, not names
- EXIF needs a field for longer, more meaningful names, controlled by the user, and these should be (optionally) substituted on ingest
Unfortunately, I don’t think things are going to change until the Japanese get completely disrupted by someone who just Thinks Different. As in “thinks like a user.”
Fujifilm made a number of updates to its camera and lens firmware today.
- 16-50mm f/3.5-5.6 version 1.12 — improves IS for stills, movies, and panning; fixed a bug that kept aperture value from being updated
- 18-55mm f/2.8-4 version 3.11 — improves IS for stills, movies, and panning; improves focus response with X-E2 and X-T1; added phase detect AF use on X-E2 and lens optimization modules; fixed problem with IS which could induce noise at low temperatures
- 55-200mm f/3.5-4.8 version 1.11 — improves IS for stills, movies, and panning; improves focus response with X-E2 and X-T1; added phase detect AF use on X-E2 and lens optimization modules
- X-A1 version 1.10 — added 56mm f/1.2 compatibility
- X-M1 version 1.10 — added 56mm f/1.2 compatibility
- X-E1 version 2.20 — added 56mm f/1.2 compatibility
- X-E2 version 1.20 — added 56mm f/1.2 compatibility
- X-Pro1 version 3.20 — added 56mm f/1.2 compatibility
(news & commentary)
At CP+ in Japan Olympus pre-announced two more m4/3 lenses, which brings to three the coming Pro lenses:
- 40-150mm f/2.8 (ships late 2014) — already announced and the companion to Olympus' 12-40mm f/2.8 already shipping
- 7-14mm f/2.8 (ships in 2015) — A fast wide angle lens that will fight with Panasonic's existing 7-14mm f/4 for the wide angle zoom spot in many bags
- 300mm f/4 (ships in 2015) — The first m4/3 exotic telephoto offering
So, sometime in 2015 an OM-D shooter would have a four-lens, high-quality lens set for their m4/3 camera: 7-14mm (14-28mm equivalent) f/2.8, 12-40mm (24-80mm equivalent) f/2.8, 40-150mm (80-300mm equivalent) f/2.8, and 300mm (600mm equivalent) f/4. For those of you who can't quickly equate focal lengths to field of view, that's 3.43° to 104° in four lenses.
Of course, today you can already go from 7mm to 100mm (14-200mm equivalent) in three lenses: Panasonic 7-14mm f/4, Panasonic 12-35mm f/2.8 (or Olympus 12-40mm f/2.8), and Panasonic 35-100mm f/2.8. So what we're really gaining is a bit more speed in the wide angle, and more length at the telephoto end.
It's a shame about the far off ship dates, though. I would have loved to have the two telephoto lenses on upcoming trips to the Galapagos, Africa, and Falklands and gone with a lighter kit.
This does up the ante for the other makers, though. Sony, for example, seems to have given up after doing a basic prime set and a few more mid-range or kit zooms for the E-mount, instead now focusing on doing the same thing for the FE-mount. Basically we've only got the 70-200mm f/4 FE coming that we know of. Nikon, of course, is resting on their FT-1 adapter, so is unlikely to give us native Nikon 1 telephotos of note, which is a shame. Fujifilm is promising a 50-140mm f/2.8 and a "super telephoto lens."
I'll repeat my "basic lens list" here just in case any manufacturers are reading:
- High performance, fast prime set that includes 24mm, 35mm, 50mm, 85mm equivalents
- Pancake prime set that includes at least one wide angle, normal, and telephoto of more modest aperture
- Wide angle zoom, preferably starting in the 14-18mm equivalent range, and preferably with corner-optimized performance
- High performance, fast zoom set that includes 24-70mm and 70-200mm f/2.8 equivalents
- At least one high performance, fast macro lens, preferably with a reasonable working distance, which typically means at least 90mm equivalent
- At least one relatively fast aperture telephoto of 300mm equivalent or higher
Sprinkle with as many kit and convenience zooms as you'd like. But to attract the Canon/Nikon DSLR shooters, you need at least that basic lens set. Ironically, neither Canon nor Nikon have achieved that with their crop-sensor DSLR lens sets. Talk about not covering a base when the other side is attempting a steal (I use a baseball metaphor so that the Canon/Nikon engineers reading this will clearly understand).
So, Panasonic and Olympus have thrown down the gauntlet. They have everything on the list or have at least announced it. Thus, we can expect m4/3 to have a complete basic set and then some. Anyone else?
(news & commentary)
Sony today announced the A6000, the successor to the NEX-6 and NEX-7, blending attributes of both and moving several key technologies forward.
The big news is the change in the focusing system, which now uses 179 phase detect points on the imaging sensor that span a little over 90% of the frame. Sony is the latest to claim "world's fastest autofocus system"—I've seen that claim in four different press releases recently, from four different companies—with the footnote indicating that this is for interchangeable lens cameras with APS sensors. Technically, that means Sony is claiming that the A6000 focuses faster than their SLT DSLRs. Oops.
Overall, the new camera more resembles the NEX-6 than the NEX-7, as we no longer have the tri-dial controls. We do have the 24mp sensor of the NEX-7 updated inside, though. The price point is more NEX-6 (US$650 body) than NEX-7, as well. I'd say the A6000 is recognizably more NEX-6 than 7, and the new name seems to reflect that, too. We have the PlayMemories App support, WiFi with NFC, a 1.44m dot EVF downgrade from the NEX-6. BIONZ X joins the formerly NEX lineup, adding some of the JPEG improvements we've been seeing in a lot of the Sony models. Sony also seems to be standardizing on the new menu set we first saw in the RX1 and now has propagated through the RX line, the A7, and now the A#### models.
On the video side we do get one new interesting twist: uncompressed 1080P on the HDMI port.
Still, the big hoopla is going to be focus. At this price point, workable motion-following focus performance would be a bit of a breakthrough.
If there's a disappointment in the Sony announcement, it's lenses. It seems that all of Sony's scattergun development is stretching their lens resources (and Zeiss's) thin. We got no new lens announcements with the A6000, nor a new roadmap for the E-mount. While there is a reasonably broad range of lenses available (and you can always use the FE-mount lenses that are starting to appear), I'd say there are still some very serious performance gaps. The 24mm, 35mm, and 50mm primes are fine, as are the 10-18mm and 16-70mm f/4 zooms. But the kit lenses aren't really up to the 24mp sensor and we've got plenty of missing good telephoto capability still. Note that Sony now says that they're doing things in the BIONZ X processor to process out diffraction impacts, but I'll bet this is basically a more carefully crafted sharpening approach. In other words, we're getting deeper and deeper into the realm of "the pixels you see (especially in JPEGs) aren't actually the pixels that were captured": they're instead heavily processed to "look good." That's not the same thing as having true resolution in the first place. Sony, please fill out the lens lineup with some additions that are worthy of the 24mp sensor (and update that 16mm!).
Support this site by purchasing from this advertiser:
From time to time I write about where we are in terms of some desirable photographic traits in the mirrorless market. Since we've had a lot of cameras introduced recently, it's time for a quick update:
- Focus — The Nikon 1 cameras are still the winner here, as nothing else can quite match them with continuously moving subjects yet. However, the Olympus E-M1 is wickedly fast on static subjects and decent at some types of moving subjects. Everyone seems to be upping their game here, so if you tried a mirrorless camera two years ago and thought it just couldn't deliver for focus, you might want to check again.
- Dynamic Range — With a small caveat, the Sony A7 and A7r just run rings around anything else we've got. Those sensors that were so good in the Nikon D600 and D800 are still good ;~). The small caveat has to do with the way Sony compresses raw files. They use a lossy compression, and what gets lost depends upon subject detail (or lack thereof). It's a complex subject, and it rarely ends up as an artifact that shows, but I personally would rather have the full data available. I would also point out that, despite the DxO numbers, I don't find the m4/3 dynamic ranges to be all that deep. These cameras, from a practical sense, are about where we were with crop sensor DSLRs four years ago, before the latest adjustments to the Sony APS sensors. Do not expect D800 (or A7r) level dynamic range out of m4/3 cameras. Not even close. That said, there's plenty of dynamic range in virtually all of the mirrorless cameras these days. One other thing: be careful about metering. The new E-M10 seems to be metering to protect highlights far more so than the E-M5, for instance. This has the net result of technically underexposing the raw data (though it is lifted back up by the JPEG processing and raw converters), and if you shoot at metered exposures you're actually sacrificing some dynamic range in order to often protect just a few specular highlight pixels (see this article).
- JPEG Quality — Olympus and Fujifilm have long been known for their pleasing out-of-camera JPEGs. Panasonic and Sony have come a long way recently and are basically close to that level. Note that I wrote "pleasing," not "accurate." Moreover, most people talking about JPEG quality actually tend to not get in and do much, if anything, with the sub settings. What they're really talking about is "the default settings produce really nice images." Yes, we've come quite a way with this. But frankly, we've always had a lot of tuning ability, and there hasn't been a camera yet I couldn't get "pleasing" JPEGs from.
- Low Light Capability — This is a slightly trickier thing to write about, because we have some shenanigans we have to watch out for. For example, if you boost contrast so that shadows drop to black fast, you can mask noise. What I'm seeing more and more is that camera makers are changing their "curves" in how raw data is interpreted to gain a little boost in noiselessness. This is a trend that started in the DSLRs and now has spilled over into all cameras. That said, ISO 1600 isn't to feared on any mirrorless camera, and many now do quite well into the ISO 3200 to 6400 range. Moreover, we've gotten some interesting new noise reduction in post processing solutions that can be amazing. The DxO Optics Pro PRIME technology, though it will probably take many minutes to run on your computer, does a really good job of dealing with pixel level noise.
- Handling — The current trend is towards making mirrorless cameras into "mini DSLRs" in terms of both looks and handling. That's not a bad thing. We do, after all, have 50+ years worth of tuning the SLR-type handling (sometimes for the good, sometimes into strange not-so-good realms). A well handling modern DSLR is faster to use and easier to set than older film-type SLRs. Yes, I know you all love "retro" dials, but that doesn't make a camera necessarily faster and easier to set, it mostly makes for more visible feedback when the camera's display are off. Don't get me wrong, I'm not against dials, but they have to be done right, not done randomly. The Olympus E-M1 is probably the camera that most handles as I'd expect coming from a Nikon DSLR (twin control dials, plenty of controls accessible with eye at viewfinder, etc.). But more and more I'm finding that the others are similar. The Sony A7's are more DSLR-like than mirrorless, for example. The recent Fujifilm X bodies are all trending that way. Panasonic's GX7 and GH3/GH4 are definitely mini-DSLR like, even though the GX7 has a more rangefinder-style body design overall. For the most part, handling has gotten far, far better on mirrorless cameras in the last couple of years. Olympus, however, really needs to rethink their menu system and the way they name options. It still gives me a headache every time I have to set up a new Olympus camera the way I want to shoot with it.
- Price — If you want to know what's holding mirrorless back from gaining significant market share against DSLRs, this is the item, with maybe one exception. Canon and Nikon just dominate the US$500-800 price range with DSLRs. Plenty of DSLRs. Highly competent DSLRs. DSLRs that use lenses that people may already have in their closets, and for which almost any lens you'd ever desire has been made. Meanwhile, the best of the mirrorless bunch, like the E-M1 and X-T1, sell for more than the Nikon D7100, which is a supremely competent DSLR. Basically, you pay a lot of money for smaller size and weight. Meanwhile, for less money than mirrorless you can get more and better pixels generally (24mp in crop sensor DSLRs at reasonable prices, for example). Mirrorless hasn't cracked the DSLR defenses yet. The exception might be the Sony A7 and A7r. These are lower-than-DSLR price full frame mirrorless cameras. They have a size and weight advantage over the full frame DSLRs with no image quality penalties. Where they fail at equalling the DSLRs is in focus and frame rate performance, particularly with moving subjects. And where Sony is particularly vulnerable for the time being is in lens choice.
- Lenses — Every mount has a different weakness, but they all have a common weakness, too: good telephoto choices, especially performance telephoto.
- Canon EOS M — Three lenses, not all available in the US. Overall rating: inadequate.
- Fujifilm X — Nice lineup of primes up to 60mm. Work in progress with zooms. Overall rating: adequate.
- Nikon 1 (CX) — Quickly gave us a small, basic set, then decided to start putting them in waterproof housings. Overall rating: barely adequate.
- m4/3 — Plenty of options in primes and zooms and even third party lenses. But basically we're topped off at about 200mm equivalent for high performance optics for the time being. Yes, I know you can stick the old 4/3 lenses on the E-M1 and get reasonable focusing, but this isn't a solution for m4/3, it's a solution for people who have 4/3 lenses (e.g. buy an E-M1 and adapter). Overall rating: more than adequate. Could be great with more and better telephoto options.
- Samsung — A very nice set of the basics, with promises for more. Overall rating: adequate.
- Sony E — Hey, everyone took a break while trying to work on FE lenses. Basically Sony built a basic set of lenses, but now has moved on for the time being. Overall rating: adequate.
- Sony FE — Two great lenses (35mm and 55mm), one mediocre lens (28-70mm f/3.5-5.6), one lens still not shipping everywhere yet (24-70mm f/4). Overall rating: inadequate.
(news & commentary)
Today Panasonic introduced the latest push of the GH series into serious video: the GH4. "4" in this case means 4K. If you had any doubts that the GH design was video-oriented, they should be fully erased by now. That doesn't mean that camera isn't capable of taking still images, but that virtually all of the technology development is centered on the video side.
To some degree, the GH4 is a videographer's dream. 4K video capability is what a lot of folk will probably talk up, but the 200Mbps bit rate using ALL-I intra mode compression and faster off-sensor movement of data (less rolling shutter) are more interesting to me. (For the reasons why, read my article on Why No 4K.) We've also got an optional mammoth YAGH interface adapter for XLR audio input and SDI output, plus 4:2:2 color at 100Mbps, and even 120 fps at 1080P. That's a lot of attention to the video side, and the specs will probably make most videographers drool when they see the body only price that should be under US$2000 (official price will be announced in mid-March). You can't come close to getting that mix of features at that price, plus the GH4 is a m4/3 camera, meaning it has a reasonably large sensor (for DOF isolation with fast lenses) and a very healthy stable of lens choices to mount out front. I can see a lot of guerrilla videographers with a GH4 and a pair of Blackmagic Pocket Cinema Cameras, which would provide a pretty nice set of features and B-roll capability in a small bag.
One other useful video feature comes with the YAGH interface unit: the SDI output for 4K video is uncompressed 10-bit, which is a very high end specification.
Of course you will need a new type of card to get all the goodness out of the video side (basically Panasonic's P2 card technology in an SD shape and size). Panasonic also has apparently paid much more attention to heat dispersion, and juiced up the focus system some as well (additional performance and good peaking features).
From a still camera standpoint, things haven't really changed much, though there are some significant changes. The new camera is still basically the GH3 body and feature set, and the stills side is still 16mp. The EVF gets a boost to the same 2.36m dot system everyone else seems to be using, and a few other features like frame rate get similar boosts to current technology, as well. The rear LCD is now 1.04m dots, the mechanical shutter supports 1/8000 and 1/250 flash sync and is rated to 200k activations. But frankly, the GH3 is big for a m4/3 still camera these days, and I really think that most folk will be looking at the GH4 as a video camera, especially given it's price and specifications. That you can take nice stills with it seems to be icing on the cake, not the cake itself.
Taken with the recent GM1, with the GH4 Panasonic seems to now be taking a "we'll try different" approach in m4/3. Certainly they've effectively steered away from the Olympus "mini-DSLR" onslaught.
Support this site by ordering this product from the following advertiser: (not currently available)
Here in the US, Fujifilm has introduced rebates (applied as instant savings) on most of their older model X cameras:
- US$200 — X-Pro1 body through end of March
- US$200 — X-M1 body or kit through end of March
- US$100 — X-A1 kit through end of March
In addition, the following price adjustments were made:
- US$600 — X-E1 body
- US$800 — X-E1 kit
Starting next week, the following lens rebates will begin:
- US$250 — 27mm, 60mm
- US$200 — 14mm, 18mm, 55-200mm
- US$150 — 23mm, 35mm
Finally, there are rumors that there will be a couple of other small adjustments to prices or rebates later in March.
Just a reminder: camera companies are trying to game you. All those leak campaigns we're getting now are basically attempts to generate buzz around "the latest and greatest."
For example, less than six months ago the buzz was all about the Panasonic GX7. A month later, the Olympus OM-D E-M1. A month later, the Sony A7 and A7r. Now after a short break for Christmas-related selling, the buzz is back and centered on the Fujifilm X-T1.
If you were to put some sort of measurement device on positive Internet messaging (rumors, posts, comments, unboxings, snap reviews, etc.), you'd find that there's an awful lot of energy being focused on hyping product launches. One reason why that is has to do with affiliate links (disclosure: this site has affiliate links for Amazon, though they're not linked by product, just a Support this Site button in the sidebar). First one to post an affiliate link for a new product tends to win the bulk of the pre-orders, and first one to post a positive review tends to win the slightly hesitant crowd's ordering.
But that's not really what I want to write about today. Let's tackle a slightly different question. So, with the E-M1, A7, A7r, and X-T1 all taking over the buzz, is the GX7 a bad camera? Obviously, no (see my review). It's just lost its buzz. The Internet, as it shows up in top ranked searches and high volume sites and user comments, is very fickle. Only the latest need apply, and it better have something that makes it different enough to buzz about (swivel EVF, fastest AF, full frame, D800 sensor, fastest AF in the case of the cameras I'm writing about today).
We've had other cameras launched during the GX7/EM-1/A7/A7r/X-T1 buzz sessions. The Sony A3000, NEX-5T, and A5000, Nikon AW1, and Samsung NX30, for example. Curiously, at least three of those cameras will outsell the cameras that generated all that buzz, and it's possible that all five will outsell at least one or more of the buzzed-about group.
I have nothing against iteration and new and innovation. I actually applaud that. Despite 50+ years of development, serious cameras still have a lot of immaturities and lapses in their designs and abilities. So progress is good. But whether a camera is the right one for you is about a lot more than buzz, especially since that buzz is usually so fleeting. You should be buying on suitability to task and value more than anything else. I'd have to say that there are plenty of mirrorless cameras that probably fit those parameters for most of you reading this.
Yes, I know that owning the latest buzz maker is appealing. If nothing else you can hang your buzz badge around your neck and claim bragging rights, which makes you feel good. In the end, however, cameras are tools, not trophies.
It's getting more difficult to evaluate what's just buzz on the Internet in promotion of commerce versus what is useful information and reasoned analysis. I don't always manage to get the latter perfect, but that's my aspiration with this site. That often means that my reviews appear long after the buzz has settled, but I'd rather that my reviews reflect real use and longer term evaluation than just climbing on the Must Publish Now bandwagon.
Am I excited about those recent cameras that generated so much buzz? Yes, I am. Mirrorless has moved further and faster forward in a very short time, and there's much to be impressed with in virtually all the recent introductions. Ultimately, though, the proof is in the pictures. Are these cameras allowing me to do things I couldn't do before, are they achieving new levels of performance and image quality, and are they worth the big dollars most of the buzz candidates are asking? That picture is more mixed, I think.
That said, expect reviews of the Olympus and Sony cameras to appear soon. The buzz may have died down, but I'm still interested in them, and so should some of you be.
Zeiss officially announced the 50mm Touit macro lens for Sony E-mount (shown at right) and Fujifilm X-mount systems. This lens will perform 1:1 macro (life-size) work, and is the third in Zeiss' Touit lineup of lenses. The Fujifilm version of the lens comes with an aperture ring. The lens is US$999.
See the data page for more details.
Support this site by ordering from this advertiser: