Don’t Worry, APS-C Isn’t Going Anywhere

With the explosion of full frame cameras over the last few years, speculation has mounted as to the future of APS-C and micro four-thirds. In this quick video, photographer Dave McKeegan offers his insights in the camera industry and gives his thoughts as to why APS-C is here to stay.

McKeegan makes some good arguments and I’m inclined to agree with him: if there is to be a casualty (which he doesn’t foresee in the immediate future) it will be micro four-thirds. For me, a trickier question to answer — and something I’d like to see McKeegan address as part of this discussion — is what Nikon and Canon will do next in terms of the inevitable transition to mirrorless. Sure, there is a huge number of APS-C shooters out there and there’s definitely a demand for more affordable cameras, but we wait to see how Canon and Nikon deal with this critical part of their customer base given the difficulties presented by the lens mounts. Full frame mirrorless required a new lens mount, and we assume that APS-C mirrorless means the same. Allowing users that smooth transition between EF-S and EF has been of great benefit to Canon over the years and I’m fascinated to see how this plays out now that we’re in an increasingly mirrorless world.

Another complication as we head into mirrorless is that Canon’s capacity to cripple mid-range camera models is going to become more of an issue. With the change in technology, it’s not just a case of leaving out some autofocus points to justify a lower price. Instead, you end up doing what Canon did with the EOS RP: the camera's hardware is perfectly capable of shooting 24 fps at 1080p, it’s just that Canon decided to switch it off.

As usual, leave your thoughts in the comments.

Log in or register to post comments


Crop sensors still have a place in some genres of photography like Wildlife, because of the pixel density and autofocus spread. The Nikon D500, which I own and use a backup to my D4s is still a great camera and definitely the best DSLR APS-C camera on the market.

Mirrorless APS-C does not require a new mount. Pentax had already proved that. (Although they did a terrible job at it).

Andy Day's picture

Interesting. I need to learn some physics. :D

With the mirror gone, one can make the box thinner, and that would require a new mount, but one does not have to make the box thinner, especially if one needs the space inside to put all that newfangled electronics, like EVF circuitry, Wi-Fi, GPS, electronic compass, accelerator chips, dual slots, intervalometer, et al, plus the big batteries to power all that and the heat sink to keep all that cool. ;-)

If the box is not thinner, no new mount is needed.

Alex Reiff's picture

True. But I did notice that Canon has been releasing lenses for their new mirrorless mount that have some pretty powerful zoom ranges and decent apertures, and they're fairly inexpensive. My guess is that allowing the lens enter the space that was previously occupied by the body is what makes that possible.

Come to think of it, Sony also proved it. Their A-mount system is descendant of the Minolta α SLR system. (Also known as, Minolta Dynex, and, Minolta Maxxum, in Europe and America, respectively).

Aps-c and Full frame have to convince the Smart Phone people that a real picture is better. Latest sales reports show real cameras are dropping by leaps and bounds because the smart phone throw away photos now rule. They do not take photos for art but for instant satisfaction and then trow them away. Sad state when art is available.

The smartphone is today's Instamatic. Most folks don't want art, they want images of themselves, their family and friends. Forty years ago you used a 110 or 126 camera to do that; the images weren't great but people had what they wanted. Now a smartphone gives better quality images of the same things people wanted photos of before.

So things haven't changed much in that regard.

Michael Dougherty's picture

If you do a quick search on the internet for the best iPhone photos, you'll find some "art" type images taken with iPhones that rival anything taken with a DSLR. It's incredible. Creativity knows no limits.

Ryan Davis's picture

Tools impose limitations upon the people using them. You can create art with smart phones, but not outside their limitations, which are some pretty severe strictures.

There are some people who have taken very good photos with smartphones. They understand the severe limitations of that device as a photographic tool, and work around its capabilities. When published as an 8-bit, 2 megapixel JPEG online, it certainly can rival an arbitrary good shot taken with a DSLR or mirrorless camera and similarly viewed online compressed and downrezzed.

You have not seen the full expression of that image in either case. You can, of course, see a good bit of the creative side, a smaller bit of the technical side, but you haven't fully experienced the image. Go see some Ansel Adams prints in person if you want that realization to hit you over the head like a ton of bricks.

Mirrorless Canon APS-C cameras have a mount: EF-M. Yes, not compatible with the RF mount and won't ever be. And yes, the ability to mount EF lenses on an APS-C camera is a major selling point (I do it all the time on my 80D). But if Canon decides to move the RF mount to APS-C they'll have some real engineering problems, and they'll have orphaned yet another system.

So I wonder if Canon's best move is to promote both EF-M and RF, and keep key EF lenses around that can be used with an adapter on each system.

there are cases when you want a small, light and high quality camera where an iphone is not good enough. for traveling, or something you can throw in your back or something with fits into smaller hands, something small and adorable... more boutique style. don't underestimate this. not every photographer is a 50 year old white guy with hands like excavator shovels ;)

Rk K's picture

Both apsc and ff (and mft) will reach diffraction limited resolutions in a generation or two. After that the smaller sensors don't make much sense if any. There won't be much of a size advantage and you can just crop.

Once the diffraction limited resolutions are reached, that is when the larger, more expensive, more power hungry, sensors won't make sense. Remember, that to get a cropped F-type to have the same resolution of a 24Mpx D-type, it has to be a 54Mpx sensor. For a cropped F-type to have the same resolution as a 20Mpx MFT, it has to be a 80Mpx sensor.

So right now, my 24Mpx Pentax K-3 is rocking it, as is Edelman's Olympus OM-D EM-1 II.

Rk K's picture

You realise that diffraction limits are different on ff than on crop? Your comment makes no sense. People want more resolution. Once you reach that limit with ff (~80-100mpx) you also have an apsc and a mft camera with you for free. Besides, Sony has proven that bigger sensors don't necessarily mean much bigger or more power hungry cameras.

Diffraction limits are based on aperture diameter. A 100‪mm f/5.6 on an F-type would have the same FoV, DoF and diffraction issues as a 50‪mm f/2.8 on a MFT, (Because they both have the same apertue diameter). So, for telephoto shooting, it looks like my Pentax K-3 and Edelman's Olympus OM-D EM-1 II both are rocking it right now.

«People want more resolution.»
Yes, but YOU SAID, “they can just crop.” When one crops a 100Mpx F-type to a D-Type frame, they have a 44Mpx image. When they crop a 100Mpx F-type to an MFT frame, they have a 25Mpx image. It would be cheaper to make a 60Mpx D-type, or a 36Mpx MFT, than a 100Mpx F-type. They get more reach, more resolution, and spend less money. Your claim that, “the smaller sensors don't make much sense… and you can just crop,” is not matching the reality.

«…apsc and a mft camera with you for free.»
I don't understand this statement. But I do know that “Cropping” the FF is a worse idea than shooting the other kits, and I know that the larger sensor costs more to manufacture, so, yes, for the price of the F-type and one lens, I can buy the APS-C and three lenses. I did the maths. I priced them out at B&H and Adorama. I think I will stick with what I got for now.

«…Sony has proven …don't necessarily mean much bigger or more power hungry cameras.»

Sony has proven that? When it comes to “not bigger,” I think Pentax has the patent, copyright, and trademark on that, not Sony. When it comes to “power hungry,” it is an industry running joke that Sony has you running for more batteries all the time. (Mostly due to the consumption of power by the EVF and Live view, not the sensor size). But compared to smaller sensor cameras, per battery capacity, it is a known fact that the bigger sensors (and higher resolution sensors) use more power, (for various reasons), and that is usually offset by having larger capacity batteries. Sony has never proven otherwise.

Rk K's picture

You still don't get it, do you? Ultimately the maximum resolution is limited by diffraction at one end and the maximum aperture (and practical optical design) at the other end, once sensor tech reaches this balance point. But this is not the same for all sensor sizes. If for FF it's 100mpx, then for apsc it's 50 and for mft it's 25. After this point, there's no reason to go for small sensors. The extra cost of a larger sensor is insignificant these days, the power consumption and heat manageable (as the latest sony cameras prove). You don't get extra reach anymore over cropping, you don't get a smaller camera body, you could still mount crop lenses... It's a no brainer. And wtf are you talking about with Pentax? Their cameras are huge and outdated.

How many 645 sensors can fit on a wafer? How many MFT sensors can fit on a wafer? If the wafers have the same mean number of failure points, which wafer would have the higher percentage of bad sensors? assuming no bad sensors, which process will yield more sensors per 1000 wafers?

So how on earth can the 645 or the F-type sensor possibly cost the same as the APS-C or MFT sensor? The mathematics does not lie. Larger sensors cost more to manufacture than smaller sensors. End of story. I will not argue that one anymore.

«…talking about with Pentax?»
Canon 5D mk IV 800 g
Dimensions (W x H x D) 150.7 x 116.4 x 75.9 mm

Nikon D850 915 g
Dimensions (W x H x D) 146.0 x 124.0 x 78.5 mm

Pentax K-1 II 925 g
Dimensions (W x H x D) 136.5 x 110.0 x 85.5 mm

Sony α99 II 849 g
Dimensions (W x H x D) 142.6 x 104.2 x 76.1 mm

Although the Sony is shorter than the Pentax, that is due to lack of a Pentaprism. Pentax is the deepest in all cases, but that is by design, giving an arguably better grip to hold with the hands. When the cameras are compared side by side, Sony will still beat the rest in depth, due to the small grip, but Pentax beats them all in width, and all but the mirrorless Sony in height.

As for weight, the Pentax has a larger grip, GPS, e-compass, gyroscopes, Wi-Fi, dual card slots, glass pentaprism, fully weather sealed, et al.

For comparison, here are the D-types

Canon EOS 7D mk II
(W x H x D) 148.6 x 112.4 x 78.2 mm

Nikon D500
(W x H x D) 147.0 x 115.0 x 81.0 mm

Pentax KP
(W x H x D) 13.2 x 10.2 x 7.6 cm

Pentax K-70
(W x H x D) 125.5 x 93.0 x 74.0 mm

Sony 77 II
(W x H x D) 142.2 x 104.1 x 81.3 mm

I posted two for the Pentax, since every Pentaxian I know, hates the small grip of the KP (but it comes with three interchangeable grips of different sizes). Even then, both Pentax bodies beat the Sony in all fields including height.

To be thorough, and in the line of full disclosure, I use the Pentax K-3. Here are its dimensions….
131.5mm (W) x102.5mm (H) x 77.5mm (D)

«…and outdated.»
They also have 5-axis IBIS, Astrotracer, in-camera intervalometer, TAv priority mode, three composite modes (not counting PixelShift, which takes that number to five), et al.

You really should not talk of facts before actually having them. It is well knows in the industry that Pentax generally makes the smallest cameras in any given class, and have always been, and still is, an innovator in photography.

Rk K's picture

Of course they don't cost the same, that's what I said. Try and read carefully, even if it's difficult. I said the cost is insignificant. It doesn't matter that it costs 4x as much if it's just 2%(?) of the entire price.

Sony a7r3, 657g... It has less gimmicks than the Pentax, but it has class leading AF and image quality. You know, the things that actually matter. The Pentax isn't small or light and the tech is outdated, with a very limited lens selection. No good.

«…cost the same, …cost is insignificant.»
DUH!!! Same thing. (Not EXACTLY the same thing, but precisely the same thing).

«if it's just 2%(?) of the entire price.»
Is it?!?

Canon EOS 7D Mark II → US$1,649.00
Canon EOS 5D Mark IV → US$3,099.00

Nikon D500 → US$1,896.95
Nikon D850 → US$3,296.95

Sony α77 II → US$1,198.00
Sony α99 II → US$3,198.00

Again, get your facts straight first.

«Sony a7r3 [sic]….»
Is a fine MILC camera, for US$3,198.00, and dimensions (W x H x D) 126.9 x 95.6 x 73.7 mm with an E-mount. Okay. let us compare apples with oranges.

«…less gimmicks….»
That is what the jealous non-Pentaxians call it. (Except Olympus shooters. They call it their own). It is not a gimmick if it is useful and used. No matter what those who are unable to use it say. (Like that article on how IBIS is a game-changer, now that Nikon has it).

BTW, by less gimmicks, you mean, no e-compass, no astrotracer, no TAv priority mode, no composite shooting, and no DoF bracketing? Because, all that aside, It seems to have most of the Pentax “Gimmicks.” So astro-photography will require expensive, external equipment which needs calibrating, altering the exposure time and the aperture value simultaneously will require full manual mode, composite shooting will have to be done in post, and DoF bracketing will…. Well that just goes back to the full manual thing, doesn't it?

«…has class leading AF….»
Granted. So what does that have to do with this discussion?

«…image quality.»
If all you care about is Mpx, yes. (As long as you are not cropping, because your 42Mpx camera becomes an 18Mpx when cropped to D-type, and 10.5Mpxwhen cropped to MFT). But what does that have to do with this discussion?
It has nothing to do with diffraction limits, nor with Pentax being the leader in small size in its classes (F-type & D-type DSLRs), nor price of large formats vs small formats, etc. These are strawman arguments.

The fact of the matter is, cropping a large format to fit a small format, will give you the same diffraction limits, but worse resolution than the comparable, less expensive, smaller format system, for wildlife, sports, et al, where long reach is ideal. the smaller formats will still be the systems of choice for those photographers. Even Edelman, fashion photographer, uses Olympus OM-D EM-1 II, (and he uses the live-composite “gimmick” quite a lot, too).

You are pulling at straws, here. The Pentax systems are quite capable, modern, systems, with great IQ, more than adequate AF, less expensive and smaller than all others in its class(es), and, despite the wonderfulness of the Sony α7R III MILC, the greatness of the Pentax DSLRs are not diminished in any way. (But that is a digression).

The long and short of it is that, “Just cropping” does not make sense for the best possible image for many reasons, including that one will be loosing resolving power of an F-type lens.

Rk K's picture

It's getting hilarious, your complete inability to parse text... Obviously it's an insignificant part of the manufacturing cost, not the purchase price. Cameras are low volume products, so they aren't priced according to manufacturing cost.

We were talking about cropping when the sensors reach the diffraction limits of their respective formats, not cropping on current cameras. Idiot...

And yes, they are gimmicks. It can have all of them, if the other camera takes better images (because of the af and modern sensor) I'm going to choose that 100% of the time. It's even lighter and smaller as a bonus! Like the f* astro tracer?! How many people take full astro photos (and not nighttime landscapes), but don't have the necessary equipment for it?! Who's the one grasping at straws?

«…Obviously it's an insignificant part of the manufacturing cost….»

the typical IC is less than the size of your little toenail. An Intel Core i7 is smaller than your little fingernail. A 4/3rds sensor is at least 20 times bigger. It should costs at least 20 times more. It is not insignificant, but it actually costs more than 20 times more to manufacture, because it has a higher failure rate.

If every wafer has four flaws, and a typical wafer can produce 400 ICs, then we have about a 1% percent failure rate. If we can now fit 20 4/3rds sensors on that wafer, we still have about four flaws, we now have about a 20% failure rate. An F-type sensor, being four times bigger, will only fit about five on a wafer. Now we have about an 80% failure rate.

Yes, I picked these numbers out of thin air, but the mathematical principle still holds. But, from a Canon white paper from 2006, an 8-inch wafer holds about 200 APS-C sensors, but only 20 F-type sensors, (because wafers are round, not oblong). This is why larger sensors are unproportionately more expensive than smaller sensors, and why smaller sensors are unproportionately more expensive than a typical IC; Size and failure rate.

Notice, I said, “Cost,” not “Price.” It is not an insignificant part of the manufacturing cost at all. It is quite significant. (That same white paper put the cost of that 8-inch wafer at US$5,000, for a quality sensor).

Also, bear in mind, that [EDIT] ̶a̶ ̶¹̶/̶4̶0̶0̶0̶s̶ ̶s̶h̶u̶t̶t̶e̶r̶ ̶o̶n̶ ̶a̶n̶ ̶F̶-̶t̶y̶p̶e̶ ̶s̶e̶n̶s̶o̶r̶ ̶h̶a̶s̶ ̶t̶o̶ ̶t̶r̶a̶v̶e̶l̶ ̶f̶a̶s̶t̶e̶r̶,̶ ̶a̶n̶d̶ ̶b̶e̶ ̶m̶o̶r̶e̶ ̶p̶r̶e̶c̶i̶s̶e̶,̶ ̶t̶h̶a̶n̶ ̶o̶n̶ ̶a̶ ̶D̶-̶t̶y̶p̶e̶ ̶s̶e̶n̶s̶o̶r̶.̶ a shutter with a sync speed of ¹/200s on an F-type sensor has to travel faster, and be more precise, than on a D-type sensor. (The difference between “shutter speed,” and, “exposure time.”) [/EDIT] If one is talking DSLRs, then they also need bigger, yet faster, & better dampened mirrors, and bigger screens & pentaprisms. They will also need faster processors, accelerators, etc., to handle more data in less time. All of this means much more manufacturing COSTS! (…NOT price).

You ought to learn how to parse text.

«…cropping when the sensors …respective formats.»
Yes. I got that. But you fail to see that diffraction limits are NOT of the sensor, but of the lppm of the lens for a given resolution. That is to say, if you want a far reach for sports, one might use a 200‪mm lens on an F-type at f/2.8, —a 171.4mm aperture— and cropping by a factor of two, or a 200‪mm on a MFT at f/2.8 —a 71.4‪mm aperture and not cropping. Now, the MFT lens can have a greater lppmm rating, —due to smaller image circle— thus producing a sharper image, and, with a better fabrication pipeline, the MFT system will be cheaper to manufacture, including the lens, (for obvious reasons, which I will NOT break down for the uninitiated).

This means that, between using the F-type and cropping, or the MFT and not cropping, the MFT will be cheaper and give better images. …But that is only important if one cares only about the image.

«…if the other camera takes better images….»
My point, exactly. Better images. …But you brought up the alleged gimmicks, which are still not germane to this discussion.

«…and modern sensor….»
Hmm?!? I wonder who makes the sensors for Sony and Pentax, respectively? Oh, irrelevant, since that is NOT germane to this discussion. (I already admitted to the better AF. Why you keep hammering that nail is beyond me. It is deader than dead).

« How many people take full astro photos …but don't have the necessary equipment for it?»

Well, Pentaxians have the equipment for it. How many don't? You tell me. I won't do your homework again. (Sick and tired of doing it). Again, though, not germane. Just another straw-man argument.

«Who's the one grasping at straws?»
That would be the one asking rhetorical questions without data, making statements without data, making assumptions without knowledge,…. the one who is NOT a trained systems analyst.

«…your complete inability to parse…. Idiot.... …the f* astro….»
Yeah! Great arguments, there. This is where my conversation with you ends, as I cannot abide the vulgar, disrespectful tones that you bring. That is usually the sign of someone who lacks reason.

Rk K's picture

Haha, mr systems analyst, what a moron... You have a rudimentary understanding of semiconducturing manufacturing and some old, incomplete information from Canon and instead of trying to learn something you reach a completely wrong conclusion and post it on the internet!

Modern sensors are fabbed using 300mm wafers, but on older nodes which are very reliable (way more so than the cutting edge "compute" nodes that are used for modern cpus). Also, a sylicon imperfection doesn't kill a sensor. Most of them are too small to do anything (pixels are way bigger than transistors after all) and if they do ruin a pixel, those just get mapped during testing, written into the firmware and interpolated from neighnouring ones. All in all, Sony can probably make a FF sensor for less than $100. (also, modern consumer chips are way bigger than what you suggest, over 800 mm2 in the case of some nvidia gpus)

Diffraction limits the maximum resolution you can achieve at a given aperture and format(!), with an ideal lens. That's all. As I said, you can mount lenses with smaller image circles on FF cameras, if you know you'll have to crop anyway. The processing also gets lighter if you crop in camera. More importantly, why on Earth would mft have a better fabrication pipeline?! It's a low volume, niche system, their lenses are way more expensive to make for what they are. Especially because they need to be brighter to keep up with ff, further complicating things. They are not cheaper to make, that's why Olympus has never managed to generate profit. This is also why mft lenses are often bigger and more expensive compared to equivalent FF ones.

There's no reason for mft lenses to have a higher resolution. They often do (for now) because they have to. But you can find FF lenses as well with the same linear resolution.

You brought up Pentax and their gimmicks, not me. I don't know why, because it's a completely inconsequential, has been system, but here we are.

«…what a moron... You have a rudimentary understanding….»
Why am I even responding to you? Oh, because I do not want the less informed to follow your stupidity.

«…300mm wafers….»
Almost 12 inches. Back in 2006 (yes, I gave a date), they used about 8" wafers. Now they use 12" wafers. Does not change the maths. 𝗠𝗼𝗿𝗲 𝘁𝗶𝗻𝘆 𝘁𝗵𝗶𝗻𝗴𝘀 𝗳𝗶𝘁 𝗼𝗻 𝗮 𝘁𝗵𝗶𝗻𝗴 𝘁𝗵𝗮𝗻 𝗯𝗶𝗴 𝘁𝗵𝗶𝗻𝗴𝘀 𝗳𝗶𝘁 𝗼𝗻 𝗮 𝘁𝗵𝗶𝗻𝗴. 𝗜𝘁'𝘀 𝗮 𝘁𝗵𝗶𝗻𝗴 𝘄𝗶𝘁𝗵 𝗺𝗮𝘁𝗵𝗲𝗺𝗮𝘁𝗶𝗰𝘀.

Also, if the failure rate for an 8" wafer was about eight points, then for a 12" wafer it would be about eighteen points, assuming the same process, because the failure rate is per square area. (actually, it is per cubic volume, but since each wafer is the same thickness…). I am not actually assuming the same process, but, let us once again assume a failure rate of merely one per wafer. 𝗧𝗵𝗲 𝘀𝗺𝗮𝗹𝗹𝗲𝗿 𝘀𝗲𝗻𝘀𝗼𝗿𝘀 𝘄𝗶𝗹𝗹 𝘀𝘁𝗶𝗹𝗹 𝗵𝗮𝘃𝗲 𝗮 𝘀𝗺𝗮𝗹𝗹𝗲𝗿 𝗳𝗮𝗶𝗹𝘂𝗿𝗲 𝗿𝗮𝘁𝗲. 𝗢𝗻𝗲 𝘀𝗺𝗮𝗹𝗹 𝘀𝗲𝗻𝘀𝗼𝗿 𝗽𝗲𝗿 𝘄𝗮𝗳𝗲𝗿, 𝗼𝗻𝗲 𝗯𝗶𝗴 𝘀𝗲𝗻𝘀𝗼𝗿 𝗽𝗲𝗿 𝘄𝗮𝗳𝗲𝗿, 𝗯𝘂𝘁 𝗺𝗼𝗿𝗲 𝘁𝗼𝘁𝗮𝗹 𝘀𝗺𝗮𝗹𝗹 𝘀𝗲𝗻𝘀𝗼𝗿𝘀 𝗽𝗲𝗿 𝘄𝗮𝗳𝗲𝗿.

«…sylicon [sic] imperfection doesn't kill a sensor. …pixels are way bigger than transistors….»

Oh! 𝗜 𝗱𝗶𝗱 𝗻𝗼𝘁 𝗸𝗻𝗼𝘄 𝘁𝗵𝗮𝘁 𝘀𝗲𝗻𝘀𝗼𝗿𝘀 𝗮𝗿𝗲 𝘁𝗼𝘁𝗮𝗹𝗹𝘆 𝗱𝗲𝘃𝗼𝗶𝗱 𝗼𝗳 𝘁𝗿𝗮𝗻𝘀𝗶𝘀𝘁𝗼𝗿𝘀, 𝗲𝘁 𝗮𝗹. 𝗠𝘆 𝗯𝗮𝗱. Someone ought to tell Sony, Canon, and other sensor makers that. Apparently, they are not aware of that. Sensors die on the table all the time, 𝗲𝘃𝗲𝗻 𝘁𝗼𝗱𝗮𝘆, 𝗶𝗻 𝗺𝗼𝗱𝗲𝗿𝗻, 𝟮𝟬𝟭𝟵 𝘄𝗮𝗳𝗲𝗿 𝗳𝗮𝗯𝗿𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗹𝗮𝗯𝘀.

«…over 800 mm2 in the case of some nvidia gpus….»
A Large Scale Integration IC has about a 1cm² area (100mm²). the die size of the nVidia RTX 2080 is actually 754mm². Nevertheless, the an Intel Core i7-8700K has a die size of a mere 151‪mm². That would be about the size of your little fingernail. So I was mistaken. I said smaller. (Apparently, they are not using the 9nm process yet ;-) ).Not the point. The point is that far more of those fit on a 300mm wafer than 4/3rds sensors, and far more of those than an F-type sensor.

The maths still holds up, even if I got the size of an Intel IC die wrong by a couple of millimetres. (…And a typical IC is still smaller than the size of your little toenail. An nVidia GPU chip is NOT typical, being far more powerful than an Intel Core i7).

«Diffraction limits the maximum resolution you can achieve at a given aperture….»
No one disagreed with that.

«…and format(!)….»
This is what we are discussing. I put it to you that at resolution limiting size, the MFT is better than cropping an F-type due to resolution and costs. That is all we are discussing here.

To have an 𝗲𝗾𝘂𝗮𝗹𝗹𝘆 𝘀𝗶𝘇𝗲𝗱 𝗽𝗶𝘅𝗲𝗹 𝗱𝗲𝗻𝘀𝗶𝘁𝘆, to get the maximum lppm that is possible, the costs of the F-type 𝗦𝗬𝗦𝗧𝗘𝗠 will be far more so than the MFT 𝗦𝗬𝗦𝗧𝗘𝗠, due to costs to manufacture the larger 𝗦𝗬𝗦𝗧𝗘𝗠 in all its glory.

«…As I said, you can mount lenses with smaller image circles on FF cameras…, if you know you'll have to crop anyway.»
Oh, you said that? Good. So, if you own an F-type camera, 𝘄𝗵𝘆 𝘄𝗼𝘂𝗹𝗱 𝘆𝗼𝘂 𝗯𝘂𝘆 𝗮 𝗗-𝘁𝘆𝗽𝗲, 𝗼𝗿 𝗠𝗙𝗧, 𝗹𝗲𝗻𝘀? To save money, or to get better images? If it is to save money, why not buy an MFT system? If it is to get better images, why not buy an MFT system? 𝗜𝗳 𝘆𝗼𝘂 𝗸𝗻𝗼𝘄 𝘆𝗼𝘂 𝗮𝗿𝗲 𝗴𝗼𝗶𝗻𝗴 𝘁𝗼 𝗵𝗮𝘃𝗲 𝘁𝗼 𝗰𝗿𝗼𝗽, 𝘄𝗵𝘆 𝗻𝗼𝘁 𝗯𝘂𝘆 𝗮 —(I absolutely HATE this meaningless phrase, but here goes)— “𝗰𝗿𝗼𝗽” 𝘀𝘆𝘀𝘁𝗲𝗺? Oh! Because “REAL PHOTOGRAPHERS” use F-type systems, because they are pros! All this time I thought real professionals use the best tool for the job. Pardon me.

You just made my point! If what you are shooting is wildlife and sports, and you need a narrow AoV, and you already know this, 𝗮𝗻𝗱 𝘁𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗵𝗮𝘀 𝗿𝗲𝗮𝗰𝗵𝗲𝗱 𝗮 𝗽𝗼𝗶𝗻𝘁 𝘄𝗵𝗲𝗿𝗲 𝘀𝗲𝗻𝘀𝗼𝗿𝘀 𝗿𝗲𝘀𝗼𝗹𝘂𝘁𝗶𝗼𝗻 𝗮𝗿𝗲 𝗻𝗼𝘄 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲𝗹𝘆 𝗱𝗶𝗳𝗳𝗿𝗮𝗰𝘁𝗶𝗼𝗻 𝗹𝗶𝗺𝗶𝘁𝗲𝗱, why not buy a MFT system?. After all, if one can crop a 100Mpx F-type, one can also crop a 36Mpx MFT image, right? Wrong! Because shooting a MFT at 500mm is better than shooting at 250mm and cropping. Shooting an F-type at 1,000mm is better than shooting at 500mm and cropping. And 𝗯𝘂𝘆𝗶𝗻𝗴 𝗮 𝗠𝗙𝗧 𝘀𝘆𝘀𝘁𝗲𝗺 𝘄𝗶𝘁𝗵 𝗮 𝟱𝟬𝟬𝗺𝗺 𝗵𝗶𝗴𝗵-𝗿𝗲𝘀 𝗹𝗲𝗻𝘀 𝗶𝘀 𝗰𝗵𝗲𝗮𝗽𝗲𝗿 𝘁𝗵𝗮𝗻 𝗯𝘂𝘆𝗶𝗻𝗴 𝗮𝗻 𝗙-𝘁𝘆𝗽𝗲 𝘀𝘆𝘀𝘁𝗲𝗺 𝘄𝗶𝘁𝗵 𝗮 𝟭,𝟬𝟬𝟬𝗺𝗺 𝗵𝗶𝗴𝗵-𝗿𝗲𝘀 𝗹𝗲𝗻𝘀.

If one is NOT using expensive high-res lenses, then one does NOT need a 100Mpx F-type camera, and can go to a 36Mpx F-type, or a 24Mpx D-type, or a 20Mpx MFT system. Right now, my 24Mpx D-type is rocking it, as is Edelman's 20Mpx MFT.

«…mft have…?! It's a low volume, niche system, …way more expensive….»
I should have thrown those prices into the mix. Did not, because I was sticking with the same class of systems, Sony being the partial exception (due to the fact that it is mirrorless, which is why I stuck to the A-mount system, originally a DSLR system). Besides, this is all a strawman argument. Nevertheless…. (Don't know why I am, yet again, doing your homework).

Olympus OM-D E-M1 II → US$ 1,599.00
Still cheaper than the F-types by a significant margin.

With Sports/wildlife lenses?

Nikon AF-S NIKKOR 600mm f/4E FL ED VR → US$12,296.95
Olympus M.Zuiko Digital ED 300mm 𝗳/𝟰 𝗜𝗦 PRO → US$2,499.00

Really? Maybe I got that wrong. Let me double-check….
Nope! Got it right!

Wide open, 𝘁𝗵𝗲 𝗹𝗲𝗻𝘀 𝗶𝘀 𝘁𝗮𝗰𝗸 𝘀𝗵𝗮𝗿𝗽 𝗮𝗰𝗿𝗼𝘀𝘀 𝘁𝗵𝗲 𝗲𝗻𝘁𝗶𝗿𝗲 𝗳𝗿𝗮𝗺𝗲. Looking at our blur charts, the𝗯𝗹𝘂𝗿 𝗰𝗵𝗮𝗿𝗮𝗰𝘁𝗲𝗿𝗶𝘀𝘁𝗶𝗰𝘀 𝗮𝗿𝗲 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹𝗹𝘆 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲𝗹𝘆 𝗳𝗹𝗮𝘁 𝗮𝗰𝗿𝗼𝘀𝘀 𝘁𝗵𝗲 𝗳𝗿𝗮𝗺𝗲 𝗮𝗻𝗱 𝗲𝘅𝘁𝗿𝗲𝗺𝗲𝗹𝘆 𝘀𝗵𝗮𝗿𝗽 𝗰𝗼𝗿𝗻𝗲𝗿-𝘁𝗼-𝗰𝗼𝗿𝗻𝗲𝗿. Exceptional sharpness is maintained as you stop down, though we do see minor diffraction-related softness come into play around most around ƒ/16-ƒ/22, but it's quite minor. All said and done, the 𝗢𝗹𝘆𝗺𝗽𝘂𝘀 𝟯𝟬𝟬𝗺𝗺 ƒ/𝟰 𝗣𝗿𝗼 𝗶𝘀 𝗳𝗮𝗻𝘁𝗮𝘀𝘁𝗶𝗰𝗮𝗹𝗹𝘆 𝘀𝗵𝗮𝗿𝗽.

The Olympus 300mm ƒ/4 Pro exhibits 𝘃𝗲𝗿𝘆 𝗹𝗶𝘁𝘁𝗹𝗲 𝗰𝗵𝗿𝗼𝗺𝗮𝘁𝗶𝗰 𝗮𝗯𝗲𝗿𝗿𝗮𝘁𝗶𝗼𝗻 𝗼𝘃𝗲𝗿𝗮𝗹𝗹. There's slightly more the wider the aperture is, but it's an extremely minor increase. Maximum measured CA was 𝘂𝗻𝗱𝗲𝗿 𝘁𝗵𝗿𝗲𝗲 𝗵𝘂𝗻𝗱𝗿𝗲𝗱𝘁𝗵𝘀 𝗼𝗳 𝗮 𝗽𝗲𝗿𝗰𝗲𝗻𝘁 𝗼𝗳 𝗳𝗿𝗮𝗺𝗲 𝗵𝗲𝗶𝗴𝗵𝘁, but on average, 𝗖𝗔 𝗶𝘀 𝗻𝗲𝗮𝗿𝗹𝘆 𝘇𝗲𝗿𝗼.

Similar to 𝘃𝗶𝗴𝗻𝗲𝘁𝘁𝗶𝗻𝗴, 𝗴𝗲𝗼𝗺𝗲𝘁𝗿𝗶𝗰 𝗱𝗶𝘀𝘁𝗼𝗿𝘁𝗶𝗼𝗻 𝗶𝘀 𝗽𝗿𝗮𝗰𝘁𝗶𝗰𝗮𝗹𝗹𝘆 𝗻𝗼𝗻𝗲𝘅𝗶𝘀𝘁𝗲𝗻𝘁 on this lens. Both with and without the 1.4x teleconverter, the level of barrel distortion sits just a hair above the zero mark.

In our testing, the Olympus 300mm ƒ/4 Pro's image stabilization using the E-M1 v4.0 was fantastic. Our lens technician Rob has quite the steady hand to begin with and, as you can see from our graph, was able to get about a 40% keeper rate of "Good" shots down to 1/60s of a second without any image stabilization. With image stabilization, however, not only was he now getting 100% keeper rates at 1/60s, he saw 100% of shots rated "Good" down to 1/15s! Going even slower, we still saw a handful of "Good"-rated shots (~30%) at a whopping 1s exposure time. At 300mm. That is incredible!

…But, like I said, 𝗶𝘁 𝘄𝗮𝘀 𝗮 𝘀𝘁𝗿𝗮𝘄𝗺𝗮𝗻 𝗮𝗿𝗴𝘂𝗺𝗲𝗻𝘁 𝗮𝗻𝘆𝘄𝗮𝘆. It is still fun when strawmen burn, though.

«…Especially because they need to be brighter to keep up with ff….» That is a strawman, and totally false. They keep up with F-types just fine. Both those lenses above are fixed aperture f/4 lenses, performing equally well (or equally badly), in low-light.

«You brought up Pentax….»
Yes! Yes, I did! (…Because you made a claim about Sony which made me laugh! Don't get me wrong Sony systems are great! If I had to switch systems today, my first two choices would be Olympus, then Sony. …Then Nikon, then Canon, then Fujifilm. Actually, it would stop at Sony. Nikon, Canon, and Fujifilm have nothing to offer me. Olympus and Sony does).

«…and their gimmicks….»
Nope. That was you. That was your strawman argument to say that I was wrong because all Pentax has are gimmicks, almost all of which one will find in an Olympus, and most of what one will find in a Sony, anyway. Yep. That was definitely your strawman. I remember burning him down and mentioning that it was not even germane to the conversation.

«…because it's a completely….»
Yep! That was the strawman! I see you are trying to resurrect him. Don't bother. Still not germane to the conversation.

That being said, and having made things clear to the uninitiated who may be following this thread, 𝗱𝘂𝗲 𝘁𝗼 𝘆𝗼𝘂𝗿 𝗰𝗼𝗺𝗽𝗹𝗲𝘁𝗲 𝗱𝗶𝘀𝗿𝗲𝘀𝗽𝗲𝗰𝘁, 𝗜 𝗮𝗻 𝘁𝗵𝗿𝗼𝘂𝗴𝗵 𝘄𝗶𝘁𝗵 𝘁𝗵𝗶𝘀 𝗰𝗼𝗻𝘃𝗲𝗿𝘀𝗮𝘁𝗶𝗼𝗻.


Larger format 𝗦𝗬𝗦𝗧𝗘𝗠𝗦 will always cost more than smaller format 𝗦𝗬𝗦𝗧𝗘𝗠𝗦, no matter what size wafers, no matter how much better the fabrication process of sensors become.

Using a proper lens will always be better than cropping.

Mathematics does not lie. …Ever! (They say that statistics lie, but it is not the statistics which lie; it's the statisticians).


Rk K's picture

You keep wanting leave, because it's embarassing how wrong you are, but instead you come back and rehash your disproven points, demonstrating your lack of reading comprehension.

The size of the wafer matters. Larger ones mean much less wasted space at the sides. Because, as you keep saying, mathematics. You still don't understand sensor manufacturing and keep comparing it to compute chips, which is completely inappropriate. It's fundamentally different technology, with different nodes, tolerances, requirements and even business models for the fabs. As I said, Sony can make a cutting edge FF sensor for $100. An mft one is cheaper, obviously, but when cameras cost $2-3k (including mft now), it simply doesn't matter.

You still don't seem to grasp that the diffraction limit for FF is 4x larger than for MFT. That's why you'll be able to crop in and get the same resolution. You also keep ignoring that you don't have to put ff lenses on ff sensors if you know you'll crop. After we get the 100mp FF sensors cropping won't be worse than using a smaller sensors. What you're saying is complete nonsense. Your examples aren't even equivalent setups.

Now stop lying, you brought up the astro tracer and compass and other useless gimmicks to try and justify your investment into an outdated, dying system. It certainly wasn't me. Pentax has no relevance to any discussion at this point.

"Nikon AF-S NIKKOR 600mm f/4E FL ED VR → US$12,296.95
Olympus M.Zuiko Digital ED 300mm f/4 IS PRO → US$2,499.00"

Hahahaha, funny guy. Seriously, this explains so much! Did you drink the Olympus marketing coolaid? Don't you understand basic optics, signal theory and equivalence? A Tamron 150-600 can do what that Olympus lens does.

So much darkness in one person's head...

Of course it's not going anywhere. APS-C is still a fantastic sensor size. APS-C cameras today are more than capable of producing great results. Look at the XT-3. it's bomb AF