**Note: 99% of this was written before the unfortunate announcement of Olympus selling its camera division. So, that affects some of what I’ve opined here and probably answers a few questions about the recent trends in the camera industry that I bring up - namely, how larger sensors have caught up to or surpassed the advantages of smaller formats.**
Nikon Z6 - Full Frame Sensor. Photo taken with an iPhone 8 in LR Mobile.
Once upon a time, we would see major innovations in sensor technology every other year or so: the shift from CCD to CMOS. The release of the D3 and D700 which, while also being around the time of the “full-frame” sensor explosion*, represented a significant boost in dynamic range, speed, and high ISO capability. There was the advent of full-frame 1080 24p video with the 5D Mark II – arguably the single most influential event in indie cinema history (along with the Arri Alexa and RED One). Then came the megapixel race: Nikon released the D800, bumping us past (for 35mm or smaller formats) the 20-24 megapixel ceiling; Sony soon followed with the same resolution in the a7R and the race was on.
*The first full-frame digital camera was the Contax N released in 2002. Canon introduced the full-frame 1Ds later in 2002 and Kodak unveiled the DCS Pro 14n in 2003. Canon then gave the world the 1Ds Mark II and 5D in 2004/2005, which saw significant IQ boosts over prior models. However, it wasn’t until Nikon released the D3 and Canon the 5D Mark II that the market began to shift toward full-frame as the digital standard.
Just a few years after the Nikon D800 and Sony a7R, Canon threw down a trump card with the 5DS/5DSR and their 50.6MP sensor. Sony followed months later with the 42MP a7RII. This was the peak of the resolution race – in fact, Canon wouldn’t lose the title until over four years later, upon the recent release of the a7R IV. Instead, camera manufacturers shifted focus to other technology: BSI sensors, blazingly fast frame rates, silent electronic modes that could compete with mechanical shutters, advanced autofocus features like eye AF, IBIS, improved EVFs, etc. We also saw an explosion of new lenses with seemingly everyone trying to one-up the others with sharper, faster, aberration free optical designs. Lenses became larger and heavier while cameras became smaller and lighter. Recently, this trend seems to have reversed itself: manufacturers realized there is such a thing as too small for a camera and engineers have worked toward a nice middle ground with smaller, slower lenses** that retained the similarly impeccable IQ (and succeeding, along with a higher price tag to match).
*The use of in-camera software to correct for various optical deficiencies – vignetting, chromatic aberration, distortion – has also been pivotal in the reduction of bulky lens designs.
While it may not seem like it, there’s been a massive shift in the camera world just in the past two years. Though unlike what we saw in the mid to late 2000s, it has been less of a shake-up and more of a subtle, natural evolution. Specifically, I’m referring to the disparateness between formats.
Not too long ago, it looked something like this:
Smartphone: terrible, usable only for what I’d describe as non-photography photographs; pictures to document something where IQ is irrelevant.
Compacts (ranging in size usually up to 1”): acceptable to very good image quality, particularly at or near base ISO. Many had impressive zoom ranges, some had fixed prime lenses (or limited ~24-70mm range) and large sensors. Many of these are still extremely good cameras (see: Ricoh GR series, Fujifilm X100 series, Panasonic LX100). Most of the compacts (excluding the large sensor outliers) hit a threshold of acceptable quality somewhere between ISO 400 and ~1600, depending on their sensor size and your standards.
Micro Four Thirds: A big leap from the 1” or smaller sensors, especially being the smallest format with interchangeable lenses.* Micro 4/3 built their reputation with a number of landmarks in imaging tech, starting in the late 2000s. They were the first mirrorless ILCs**; they matured and standardized the implementation of in-body sensor-based image stabilization (originally developed by Konica-Minolta); they cultivated the first truly thorough mirrorless lens line; and, due to some very wise choices with their mount, they had the distinct advantage of tiny, cheap, but optically excellent lenses.
*I’m ignoring the Nikon 1 and Pentax Q series because neither was ever a fully realized or mature system.
**Technically, the Epson R-D1 was the first MILC followed by the Leica M8, but for obvious reasons we’re restricting this to cameras with autofocus, live-view, and zoom/telephoto lenses.
APS-C / APS-H: Before full-frame technology matured, these were the standard sensors in professional and prosumer camera bodies. APS-H had a ~1.3x crop factor compared to the ~1.5x (Nikon) or ~1.6x (Canon) APS-C crop factor. It was used primarily by Kodak in a number of models, Canon in their 1D series, and Leica in the M8/8.2. It was eventually phased out in favor of the (slightly) smaller APS-C format – likely because the performance difference was negligible, improvements in APS-C were advancing faster, and APS-H required the use of full-frame lenses since lines like Canon’s EF-S couldn’t cover the larger sensor. It made no financial or technical sense for Canon to maintain three separate sensor-size lines at the same time (FF in the 1Ds and 5D series, APS-H in the 1D series, and APS-C in the Rebel line). So, Canon dropped it, Kodak dinosaured out of the digital camera market (though continued to make sensors), and Leica upgraded to full-frame in the M line.
Note: APS-H is still used by Sigma in their sd Quattro H body. As far as I know, this is the only camera currently in production to use an APS-H sensor.
Nikon Z50 - APS-C Sensor. Photo taken with Nikon Z6 & Nikkor 24-70/4 S
And thus, APS-C became the gold-standard for prosumer and consumer bodies. Advantages were plentiful over the larger 135 size: greater sensor yield (meaning cheaper to produce), smaller bodies, and smaller lenses – all while retaining good IQ and decent ISO performance. It was an excellent compromise between quality, cost, and size. Sony’s 2014 APS-C a6000 ushered in a mass appeal of mirrorless cameras among general consumers – it was the best-selling mirrorless camera of all time and the best-selling camera over $500 year after year. In fact, it still regularly remains toward the top. It may well be the most successful and profitable MILC ever released.
35mm/135 aka “Full Frame”: Since roughly 2007/2008, this has essentially been the standard against which all other formats are compared, right down to field of view or “crop factors” (part of this is of course inherited from the days of film). Full frame yields the best sheer IQ when you consider factors like size, weight, cost, flexibility of shooting, etc. Yes, you will have greater potential for superior image quality from a medium format system, but even today MF is still extremely limited by price and lens selection. And prior to the last couple years (i.e. the release of Fuji’s GFX and Hasselblad’s X1D), there was zero question about which format had the greater price-to-performance ratio, not to mention other factors like size/weight.
Medium Format (specifically 44x33): Until recently, it was very limited in application, but undoubtedly the best image quality under ideal circumstances. Improvements in technology were – and still are – slow to progress; the 50MP sensor used today (also the first CMOS medium format sensor) dates to late 2013 or early 2014. Earlier CCD based cameras like the Hasselblad H4D and Pentax 645D had incredible acuity, color, and tonal transitions at base ISO. But, once you pushed beyond ISO 400 or so, such advantages rapidly, and non-linearly, dissipated. Cameras like the Nikon D800/E existed at the same time and easily won out on dynamic range, color, and noise at anything above the lowest sensitivities. They also had significantly more latitude for hand-held photography, anything requiring a smaller footprint, fast glass, or access to a library of specialized lenses (e.g. tilt shift, super macros). And, of course, they were leagues more affordable.
Ok, let’s travel to 2017/2018.
Sensor technology tends to trickle upward; BSI (backside illumination) was originally developed by Sony with their Exmor R and used in some Sony phones as well as the iPhone 4s. It wasn’t until five or so years later that the technology appeared in a larger format sensor (Samsung’s APS-C NX1) and subsequently moved upward to the Sony a7RII and recently to medium-format with the GFX 100 (but oddly never to M4/3). Sensor-based image stabilization began with Konica Minolta’s Dimage A1 (2/3” sensor) onward up to Olympus (4/3 sensor) with the E-510, and now in 44x33 with the GFX 100.
The upshot of this that while under optimal conditions a larger sensor will always produce a higher quality image, there has almost always been various equalizers in the mix to close the gap (even if not fully so) under specific circumstances.
With the release of the Olympus E-M5 in early 2012, there now existed cameras with incredible in-body stabilization that could fit in a jacket pocket. But their sensors were 1/4 the size of full-frame. Given equivalent sensor technology, this gave full-frame a two-stop advantage. But it wasn’t really that simple: with 4-5 stops of image stabilization in the E-M5 versus zero in any full-frame camera (at the time), the Olympus could more than close the gap in lower light. Where the Nikon may require ISO 3200 to achieve a sufficiently fast shutter speed, the E-M5 may be capable of shooting at ISO 200 – all in a much smaller package to boot. There are limitations to this, of course: drop too low with your shutter and while you may be able to get a sharp image of a static subject, you may not for people or other moving objects. Conversely, at base ISO on a tripod, the Nikon wipes the floor with the Olympus.
Something odd has transpired particularly in the last two years or so. 2018 saw the release of Canon’s EOS R and Nikon’s Z6 & Z7, the first full-frame mirrorless cameras from either company – in fact, from any company other than Sony and the far more niche Leica. Looking specifically at the Nikon Z, we now had a full-frame, in-body stabilized sensor in a body the size of an Olympus E-M1. We could now shoot 50mm at 1/8 of a second with ease – even lower if we accept a dip in hit-rate. The Olympus still has better stabilization; physically speaking, on an equal playing field, a smaller (read: lesser mass) sensor will always be easier to move.
Of course, for a given field of view, aperture, and build quality, micro 4/3 lenses will always be smaller. You cannot get a full-frame 28/2.5 with autofocus at the same size of the Panasonic 14/2.5. But, since our playing field has been (mostly) leveled via stabilization, the other option is faster glass, which also applies if matching DOF is important to you*. Then we have something like this: Nikon Z with 24-70/4 S and Olympus E-M1 with a Panasonic 12-35/2.8. But now, we are once again back to almost exactly the same size and the Panasonic lens is still at a one-stop DOF disadvantage. These are simple rules of physics and the reason that lenses like the Fujinon XF 200/2 are gargantuan behemoths – they’re trying to match a 300/2.8 and thus ends up being about the same size and weight (and often, more expensive).
*DOF equivalency is a big, dumb topic of debate in internet forums and comments sections. Apparently, everyone on the internet shoots wide-open portraits or pictures of fence posts or something. If shallow DOF is that important to you, use full frame. It’s significantly cheaper and with more options. Alternatively, you could try to work past your obsession with bokeh and obliterating any sort of context behind or in front of your subject, but that’s your prerogative. Ironically, the same people who tear their hair out over f/1.4 lenses vs. f/1.8 also chase resolution, which is entirely wasted if most of your frame is out of focus.
Beyond this, we now have cameras like the Hasselblad X1D and Fujifilm GFX 50R that are approximately the same size as the Nikon Z, E-M1, or Panasonic G9 lines. With the exception of Sony’s a7/a9 line, the ubiquity of large format mirrorless cameras didn’t exist even just three years ago. Now it’s become not only commonplace but is clearly the direction of the future (nothing points to this more than Canon ceasing development of new EF lenses – I expect Nikon has or will do likewise). It’s quite possible that cameras like the Nikon D6 and Canon 1DX Mark III will be the last pro-DSLRs produced by those companies – consider the four or so year development time between new models… what will mirrorless technology in 2024 look like? My guess is the few remaining benefits of DSLRs will have been eclipsed by then, especially considering that depending on who you ask, they already have been. And, of course, I’d be remiss if I didn’t point out both Panasonic and Sigma’s entry into the full-frame mirrorless game – also both within the last two years.
Another issue is that prices have constricted inward; the lower end has gone up (a Fujifilm X-T4 is $1700, an E-M1 Mark III is $1800) and the higher end has dropped (a Sony a7III is $2000, a Z6 is $1800). Mirrorless medium format can be had for less than pro body full-frame. You can acquire mirrorless full-frame for $1,000 or less (Canon Rp, Sony a7II) – the same as an E-M5 Mark III and less than a Sony RX100 VII!
All of this is to say that while the capabilities of larger formats have largely caught up to the advantages of smaller formats in many ways, the price has seemingly done the opposite. This isn’t at all intended to say that Micro 4/3 or APS-C don’t still have advantages; you won’t find full-frame lenses with the same size, price, and quality as many of the Olympus of Panasonic lenses. This especially applies on the telephoto end (where I think M4/3 still retains a serious and very legitimate advantage); there is nothing in APS-C/FF with the same reach, size, or price that compares to the Olympus 75-300 or Panasonic 100-300. And there are very few larger sensor cameras that can do the frame rates of a Panasonic G9 or E-M1X in C-AF (only the Sony a9/a9II and Canon 1DX Mark III come to mind) and none can equal the speed in S-AF. Plus, all of this is to say nothing of the built-in features Panasonic and especially Olympus have developed over the years – arguably the closest thing to what we’d call computational photography in ILC cameras today.
A lot of what we’re seeing here is due to a natural paradigm shift in the camera world that, in large part, was brought on by the mass adoption of smartphones. More photos are taken and shared today than ever before in history – but unlike the 2000s or earlier, very few “regular” people even own a dedicated stills camera. The market caters mostly to amateurs, hobbyists, and professionals now – not mom and dad.
To close, I’ll simply say that this is a very interesting time in photography. Image quality and features are more than we could have dreamed of five years ago and our options are vast and expanding by the day (the Fujifilm GFX 50R was recently discounted permanently to $3500). There’s a good deal of volatility and uncertainty thrown into the mix due to COVID-19; it’s made for some nice discounts and offers out there*, but is also doing some concerning damage financially speaking.
I, for one, will be very curious what the next year or two holds for us.
*Panasonic is giving free 2 year warranty extensions (for a total of 3 years) on all of their cameras and lenses purchased between now and March 31, 2021. Make sure to register online!