Lots of interesting points were raised in another thread where I was considering trying a compact full-frame camera. That got me doing even more research and looking at all of the various ecosystems in more depth. I found some quite enlightening facts, such as that Sony’s A7C line of full-frame cameras are smaller than many APS-C cameras, and OM-System’s OM 1 micro 4/3 camera is larger than Fuji’s X-T5 APS-C camera. We’re also seeing Fuji’s medium-format cameras very close to the size of full-frame cameras from Canon and Nikon.
So, camera body size is less of a consideration these days for interchangeable lens systems. Which of course leads us to lenses, where there is still quite a difference in size and weight between lenses made for the various sensor sizes, as illustrated nicely by @garrett’s post here: Trying out new cameras and lenses - #47 by garrett
But with the high resolution of modern sensors, and with full-frame cameras able to shoot in APS-C crop mode (and medium-format able to shoot in 35mm crop mode), it got me wondering whether there is a “best of both worlds” scenario. For example, if you can use a smaller APS-C telephoto lens on a full-frame body in crop mode for wildlife, and then a smaller full-frame prime lens for street photography, can you keep your gear compact and still enjoy the benefits of a larger sensor? Is it really the best of both worlds? Or are there significant downsides?
Right now, it would seem Canon, Sony and Nikon have an advantage here because their mounts take both FF and APS-C lenses. The promise of a medium-format body, eg Fuji’s GFX line, able to use FF, APS-C and MFT lenses doesn’t really seem viable at this point in time because cumbersome adapters would be needed.
What are you thoughts? Does anyone have experience of this? Is it viable now or do you see it as viable in the future?
(I’m not interested in championing one sensor size over another, and I also think that there’s a lot more to the decision of what camera ecosystem you want to buy into, such as ergonomics and cost, for example.)
If you’re really thinking of doing this, it would seem prudent to do a lot of research about what lenses and bodies you want: Using Nikon DX Lenses on FX Cameras
I’d also note that overall weight shouldn’t be the only factor, but the size of the lens along with the weight of the body. The balance I’ve found to make a huge difference. I found it easier to hold the D850 with a tamron 70-200 than it is told hold the Z7ii with a Nikon 100-400, despite the Z7ii and lens being much lighter. The balance of the heavier D850 body made a lot of difference in the overall handling of things.
Pixel density comes into this. You are basically doing a crop of the image which could be done after on a computer later anyway. I recently bought a Canon R7 for wildlife photography and it has over 32MP in the crop sensor. Someone I know bought a full frame Nikon Z6 which has just 24MP. She would need over 80MP to match my crop sensor camera for ability to post shoot crop and get good enlargements.
My feeling is that full frames bring out the best with wide angle lens by giving wider coverage and crop sensors miss out on this because of the crop factor. On the overhand a quality crop sensor brings out the best with telephoto lenses because the crop factor tends to be about 1.5 and turns a 400mm lens into the equivalent of a 600mm on a full frame. It may be more important to decide which sensor size is more appropriate for your style of photography or have two cameras. BTW, I use a Canon G16 high quality compact sensor camera for night time street photography when I travel. The f1.8 6mm lens lets in lots of light and retains good depth of filed to allow candid images with a camera that is neither too large or threatening looking for the subjects being photographed. I can print great A4 images from the shots taken. Probably bigger if I wanted but A4 is fine for my needs.
Does crop mode on 35mm full frame or Medium Format give you the best of both worlds?
Having just bought a Sony full-frame camera after several years of using Sony apsc cameras, I’d say… No, not really.
I can put an aps-c lens on the full-frame camera. I’ll get the crop. (or you can have the whole circular image if you want). It will be, I think (not a numbers person) about 14Mpx or something like that. Won’t be as good as on my 24Mpx actual-aps-c camera. In fact, the aps-c lenses I’d like to use don’t seem to give very good results on my FF camera.
When I am using my 135 FF lens, I do a few crop shots. I have no complaints about the image quality of these. I do acknowledge the advice that it would have been better off to crop in post… but I’m just not that good at the necessary visualisation and remembering which I want to crop.
I do like my Sony a7iv camera a lot. AF and tracking stuff is amazing. I can get better shots at higher iso, eg going above 3200, which is what I wanted it for.
I’m absolutely going to carry on using the Sony a7iv for my concert pics, which is my every day niche. I’m also going to use my a6500 aps-c camera for my much rarer travel or carry-around camera.
But I am a confirmed fan of aps-c. Lighter, cheaper gear, especially the lenses! I
probably should have stuck with that.
So the issue is lens size. We have already established that bodies don’t need to be (much) larger for larger sensors. But larger sensors require longer lenses to cover the same FOV and longer lenses are heavier.
Except, that is not true either: for instance, the full-frame Canon 600 f/11 or the MFT 14-140 lenses are long lenses that are still very small; because they are dim!
As far as I understand, the reason why my APS-C 16-80 f/4 is smaller than a FF 24-120 f/4 is that it’s a stop dimmer. They could build a FF 24-120 f/5.6 that would be a similar size as my 16-80 (and equivalent in every way). But “f/5.6” would be a marketing stigma, it would be perceived as “cheap” or “dim” or “not premium”, which is why they don’t do it.
Or more cynically, perhaps these things would be too cheap and useful, thereby cannibalizing sales of more expensive lenses. So they only do it for exotics like the Canon 600 and 800 f/11.
Fun fact, I once sold a 10-18 APS-C lens to a photographer using a FF Canon RP. The RP automatically switched to crop mode when mounting the lens, so the buyer didn’t notice it. I wasn’t brave enough to tell them…
TL;DR Instead of cropping into a sensor and using a crop lens, you could natively design a dimmer lens. This would give you the same weight/size savings, without needing to discard half your pixels.
BTW, I like to use the crop modes on my Fuji X-T5 (“sports finder mode”) and Ricoh GR III. In good light and where the depth of field is sufficient, they produce great images. The X-T5 additionally has dedicated crop modes, but they only crop the JPEG, whereas the aforementioned modes produce cropped raws. (Although darktable tends to struggle a bit with these cropped raws).
Fun tidbit, the 24 MP GR III in 23mm crop mode still outresolved the 16 MP Fuji X100T, which is part of the reason I didn’t gel with the latter.
I was surprised at how nice the files from the GR iii looked when I first got mine a few years ago. If Ricoh ever put out another one that is an actual upgrade, I’d buy it instantly.
Thanks for all the replies. Just to add some numbers, the Sony A7CR has a 61MP sensor and so shooting in crop mode makes it a 26MP camera. I don’t think it will be long before we have a 90MP full frame sensor. These kinds of numbers are obviously great for any kind of computer viewing and printing. The pixel density will be the same as well because in crop mode you are only using a cropped part of the sensor. So I’m thinking this negates all the arguments about megapixels and pixel density not being sufficient.
As for lens size and balance, a compact full-frame body with an APS-C lens should be the same as an APS-C body with an APS-C lens, right? This is why I’m thinking you can maybe have the best of both worlds. You can have a full frame sensor and enjoy all the benefits of full frame with full frame lenses; but then when you want a smaller, lighter lens with extra reach, you can use APS-C lenses without suffering a penalty that really matters. It seems most arguments about why it’s not a great idea are now becoming obsolete with smaller bodies and higher MP counts. But maybe there’s something I’m still missing here.
Incidentally, I’m not on the verge of swapping out my APS-C system or planning to buy another camera any time soon. I’m just putting this out there for discussion and considering it for future purchases, perhaps.
For me personally, any combo where the body out weighs the lens and the lens is relatively short, it feels fine.
The Z7ii is a pretty compact body for what it is, and was not much larger than the X-T3, so when the 100-400mm is on it, I find it very unbalanced and not super easy to hand hole. The 50mm f1.8 Nikon lens, which is my most used lens now, is pretty large for a 50mm and feels just slightly front heavy on the Z7ii. The newly acquired voigtlander 15mm f4.5 is quite compact and feels fine.
The 50mm is still somewhat problematic for me, I twitch a decent bit and have found hand holding that lens, even at what I’d think to be shutter speeds that are fast enough, broad daylight, f8 @ 1/160 sec, I still sometimes get blurry photos. I have not figured out if this is my shaking/twitching, my use of back button focus and just missing focus, or something else. The photos don’t look like obvious motion blur, but are rather just unsharp.
A tripod, which I use most of the time, makes this all moot
For me, the grip is really important, and it’s something I’m not thrilled about with the X-T5. I much preferred the X-S10 grip. It makes a big difference for me when I have larger lenses. But I don’t have an extensive lens collection, and the biggest I have ever used is the Fuji 70-300mm, which is quite compact for its focal range. I bought the Smallrig grip for the X-T5 and it makes it more comfortable for sure, but it also adds to the size and weight.
But I agree, a larger body is probably a bit more versatile because smaller lenses on it probably won’t be a problem, even if it’s a little unbalanced. Whereas a small body with long lens will be hard to hold. I think for me a sweet spot where the body feels fine with both full frame and APS-C lenses, would be something like my current X-T5 with a better grip, so maybe an X-H2, A7RiV kind of size.
When I see someone like Simon d’Entremont carrying around his Canon 600mm F4, I really wonder whether I’d have the motivation to carry around such a thing, and that’s why I’m not sure I’d ever want to go completely full frame. But a full frame body with full frame wide angle lenses and APS-C telephoto lenses could be something I’d consider.
Not sure I fully understand what you mean here. An F4 lens is an F4 no matter what sensor size you are on, and has the same light gathering capabilities, so I’m wondering what you mean by being a stop dimmer?
The equivalent of an APS-C 80/4 lens is a FF 120/5.6, as it will capture the same field of view and depth of field. Of course ISO will need to be raised by one stop to counteract the loss of light, thereby producing the same noise as well.
That’s why I said, somewhat inaccurately, that an APS-C 80/4 lens is “dimmer” than a FF 120/4. My claim is that a fully equivalent lens would have the same size. Hence, a FF 24-120/5.6 could be made as small as an APS-C 16-80/4.
Ok, I see what you’re saying now, thanks. I’m not a lens designer, but I assume that the mount and flange distance also affects the overall size of the lens, but not sure how much of an effect that would have.
But going back to the original question, it seems there are two compromises to be made to get a smaller lens on a FF body. Either make the lens have a smaller max aperture, as you said, which makes the lens dimmer; or reduce the size of the image circle, which would then require using crop mode on the camera.
The former would probably not be popular, as you rightly said, because marketing an F/5.6 lens is much harder than an F/4, etc. The second option has historically not been popular because the compromise of losing megapixels was too great. Enthusiasts and pros don’t want a 12MP image anymore. But with FF sensors now at 61MP and probably soon higher, this argument is not so compelling. And with the greater depth of field and reach from a crop lens, it makes it an attractive proposition for certain genres like wildlife without sacrificing light gathering capabilities. There are plenty of smaller FF lenses for shorter focal lengths, so a “best of both worlds” kit would have FF lens for wide angle and standard shooting, and APS-C lenses for telephoto shooting, with only one body needed.
I’m thinking it’s viable, but maybe not the ideal solution for everyone. I don’t know, I just think the arguments for it being a bad idea are maybe outdated now.
I think you are getting ahead of history: we are not there yet.
Mine is 33MP. Sony a7iv. The MEGA mega-pixel cameras are still top-of-the-range, top of the price list. A long way away from being any kind of “usual” for FF cameras.
The Sony a7R5 has enough MP to give a crop with as many as an APS-C camera. But many of us FF owners are going to get not much more than 12MP in crop mode. Probably for some years to come.
That’s a fallacy, though. You do lose “light gathering” by cropping. By capturing a smaller area, the noise floor increases (APS-C has higher noise than FF). To compensate, you’d need to lower your ISO by one stop. There is no free lunch.
That’s noise per image, not per pixel. Cropping doesn’t change anything about the pixels. But more pixels make less noise per image area, and so cropping increases noise.
Noise per pixel is a univariate statistical distribution (per channel). It is relevant if I want to extract/estimate/reconstruct the “underlying” image. It is neat because it is local and well-defined, and depends on the sensor and the camera circuitry.
In contrast, how would you even quantify “noise per image”, and what purpose would it serve?
Noise per image area is well-defined and useful. It’s what you perceive when looking at the image.
I would argue that pixel-level noise is pretty irrelevant to most images. What counts of how visible that noise is in the rendered image. Obviously the visibility depends on reproduction size. But the noise level in general is easy to measure and see.
