Comparing lenses for full frame vs. APS-C sensors

[ajax]

Higher quality images for large format printing, David, not for viewing on the computer / posting on the internet. Smaller sized sensors are more than adequate unless you plan to make large prints.

As an aside, if I were getting into a new camera system today, I would likely not go full-format, but would go directly to either a FujiFilm GFX100S or the Hasselblad 907X, medium format cameras. Both are at an appropriate price point versus full format and with the medium format sensor, I would get an even larger image to work with. As the Fuji is a true 100MP, 16-bit camera that is likely the direction I would go versus the Hasselblad. I understand that Hasselblad uses rebranded FujiFilm lenses.

MY MOTIVE DEFINITELY IS MAKING LARGE PRINTS! This is the basic reason for me saying "I think my motive is the same as his." when referring to some of Manfred's replies.

I'm certainly open to looking at those other options and will include consideration of related comments from others.

[DanK]

How large, and will you print yourself, as you do now?

All of my photo printers have been Canons, but I suspect these data are similar for Epsons. To step up to 17" wide (the Prograf Pro 1000, which I use) costs about $1300. It's also quite large, about 29" across and 70 pounds. The next step up is 24 inches, which normally costs bout $2300 but is currently on sale at B&H for about $2100. I made the decision to cap this at 17" and to use a lab if I want something larger (which I haven't done yet).

If you are interested in printing at the sizes I do, the key thing is lenses, not megapixels or sensor size. The first four images in the built environment gallery I pointed to were taken with a 12-MP Lumix micro four thirds camera. The second one from the left, the first of Bergen rooftops, is nearly a 40% crop from that. The final image is less than 8 MP. An 11 x 19 print (the aspect ratio is the reason for 11 rather than 13) is hanging on my wall, and I've had one in a gallery. If you stare very closely, you can see that it is not quite up to the levels of a 30 MP image, but the fact is that no one who looked at it, including the gallery director, found the print lacking. I wouldn't go larger with it, however. The reason is that modern software does a pretty good job of interpolating to increase resolution.

Personally, I find 30 MP, plus or minus, to be a nice sweet spot; I can comfortably print 17 x 22 with moderate cropping. Literally no one has ever looked at a 17 x 22 print from my 30 MP Mark IV and commented that they see too little detail. I bought my Mark IV just after the R5 (45 MP) came out. One reason was a very large difference in price. But even leaving that aside, I didn't see the 50% increase in MP all that beneficial for me. It has other features that are, however, and if I went with the R series, I would probably not get the R6 because it's only 20 MP.

Others may differ, of course. Manfred does a lot of work where camera size is less important (landscapes and studio work) and prints larger than I do, I think, and he does value additional MP more than I do. And I'm not advising you to buy any particular sensor. I'm just making the point that more MP isn't the main bottleneck in getting crisp prints.

By the way, all of the bug photos in my gallery were taken with crop-sensor cameras, either a 50D or a 7D (first generation). I think the 7D may have the same sensor your camera does; in any case, it's the same number of megapixels. I use that because it has a much higher pixel density than my 5D, which would be relevant for wildlife as well. Moreover, because of limited depth of field in macro work, they are mostly shot at a nominal f/13--which gives an even larger effective f/stop (smaller aperture) at macro distances, which softens the image because of diffraction. I've never printed them large (other than cheap canvas prints for my grandkids), so I can't give you detail about that, but I can try to create a link to a a full-resulution JPEG available if you'd like to see one.

I know this is redundant, but as both Bill and I posted, the single most important thing is lenses. That doesn't mean necessarily buying L lenses, but it does mean buying good lenses. Fortunately, you have a wide variety of options. All of the major manufacturers have superb OEM lenses, and the aftermarket producers, in particular, Tamron and Sigma, have really stepped up their game in recent years. However, not all options are available for all bodies. For example, I don't think there are many good aftermarket options for the Fuji X system, and the aftermarket manufacturers for the most part haven't yet shifted from Canon's EF mount (DSLR) to the R mount (high-end mirrorless). they will in time because that's where the market is.

Good luck sorting this out.

Dan

[ajax]

Not true, sorry. Many cameras apply pre-processing e.g. tone curves to what the sensor outputs before writing to the so-called "raw" data file - usually at a greater bit depth than the camera ADC (analog to digital converter).



Not true. Programs such as RawDigger can send "raw composite" data from the file to a screen unchanged by conversion to RGB e.g. demosaicing.



Please pardon my pedantry.

Well, I do concede that I'm no expert on cameras which has something to do with why I started this discussion. However, I have been developing raw files for several years now. I'd argue pretty strongly that what I said is the purpose of raw files. I know it is why I prefer raw files and I'd strongly resist buying a camera that does what you say. Of course the main problem with all of this artificial intelligence is that we have NO way to tell what is really going on.

The real raw data must be put through the demosaicing process in order to produce pixels which are what portrays color. That has to happen before tone curves can be applied. The camera has to do all of that and more to make something that can be displayed. If that data is really raw you get to do the demosaicing in post processing and there are lots of different ways to do it.

I guess what I meant to say was that the raw data is NOT what the camera is displaying when you are viewing a picture with it.

[Manfred M]

David - I think it all depends on what you look at as a "large print". I'm working on a couple of prints right now that I hope to get out to a printer next week that are going to be in the 44" x 66" range; these are definitely the largest prints I will have made. These are going to a show in one of the galleries in town in about 6 weeks.

Like Dan, I do my own work for anything up to 17" wide and I have an Epson P800 printer that I generally make 1 - 2 finished prints a week on. Most of my work is either 13" x 19" or 17" x 22", although I make smaller prints (generally no smaller than 8" x10") from time to time. My printmaking started decades ago in the traditional "wet" colour and B&W darkroom. I have been making inkjet prints for over 10 years, initially with the Epson 3880.

In my experience, when it comes to large prints, the most important factor is sensor size, closely followed by the number of MP on the sensor. In large prints, we end up enlarging (upsampling) the data, so a larger sensor needs less enlargement, hence improved quality. The same is true for the number of MP on the sensor, again, more data is better. I have printed image from micro four-thirds, 1.5x crop sensor and full frame cameras. I can see a difference in image quality between the different sensor sizes in those larger formats.

Unless you are shooting in near-ideal conditions (i.e. a heavy duty tripod), any reasonably modern lens produces excellent results. I try to use a heavy-duty tripod whenever I can, to minimize camera movement. While image stabilization works well, I find that it is not as good as shooting from a very stable platform. Most of my work is landscape, street photography, portraiture and still-life work (the latter two are primarily studio based and even when I do location shooting, I use studio lights). I don't do any macro work or a lot of flowers, so I defer to Dan on those topics.

You can see the types of images that Dan and I make in our postings, so you can judge our experience from our work here, although they are not prints. Another good print maker here is Donald MacKenzie (another one of the moderators) and he specializes in B&W work almost exclusively.

When it comes to Ted, he offers a lot of opinions, but tells us he does not do any print making, so he is writing more based on theory than any practical knowledge on this topic.

The only reason I mentioned the medium format cameras is that if you are looking at moving from APS-C to full frame, you are likely free to look at other makes than Canon as most (if not all) of your current gear will be usable on a Canon full frame body. Sony, FujiFilm, Nikon. Panasonic, etc. should be considered as you look at upgrading to larger format gear.

[xpatUSA]

Well, I do concede that I'm no expert on cameras which has something to do with why I started this discussion. However, I have been developing raw files for several years now. I'd argue pretty strongly that what I said is the purpose of raw files. I know it is why I prefer raw files and I'd strongly resist buying a camera that does what you say

That may leave you with quite a small choice of cameras, David, IMO. I know of no digital camera that simply converts the sensor per-pixel output voltage via the ADC to a per-pixel 16-bit number in the raw data segment of a "raw" file. Anybody?

Of course the main problem with all of this artificial intelligence is that we have NO way to tell what is really going on.

The real raw data must be put through the demosaicing process in order to produce pixels which are what portrays color. That has to happen before tone curves can be applied. The camera has to do all of that and more to make something that can be displayed. If that data is really raw you get to do the demosaicing in post processing and there are lots of different ways to do it.

Again, I claim that what you call "the real raw data" only exists briefly in camera's buffer memory and that what ends up on the memory card has had some pre-processing applied. That would be in spite of what dumbed-down articles say.

I guess what I meant to say was that the raw data is NOT what the camera is displaying when you are viewing a picture with it.

Now that I agree with.

[Manfred M]

That may leave you with quite a small choice of cameras, David, IMO. I know of no digital camera that simply converts the sensor per-pixel output voltage via the ADC to a per-pixel 16-bit number in the raw data segment of a "raw" file. Anybody?

Can you tell me why anyone should care? Most people are interesting in producing images and don't really care how we get there. If you have a look at some of the fancy algorithms companies like Apple and Samsung, among others, throw at the captures, the results are pretty darn amazing. Signal processing has been around for a very long time and is used because the results are what people want.

I know of no one (or perhaps you are the exception) who really cares what happens in the background.

Again, I claim that what you call "the real raw data" only exists briefly in camera's buffer memory and what ends up on the memory card has had some pre-processing applied. Anybody?

Again, can you tell me why anyone should care? Camera manufacturers are all about pleasing their customers so that they can produce good images (whatever that may be). What is wrong with pre-processing?

I still remember years ago when Sony was accused of doing so and the astrophotographers were all concerned about "pixel killing" star removal. The mFT camera makers have long done this to correct for optical issues with their lenses. In the early days of the Lumix line, Leica apparently had a contractual agreement with Panasonic that Leica branded lenses could not used these enhancements. I suspect that these days are long gone.

[DanK]

In my experience, when it comes to large prints, the most important factor is sensor size, closely followed by the number of MP on the sensor. In large prints, we end up enlarging (upsampling) the data, so a larger sensor needs less enlargement, hence improved quality.

Isn't this mixing two different things? If I'm not mistaken, upsampling isn't magnifiying; it's using various mathematical algorithms to interpolate more data points to deal with the situation where there are fewer pixels than the native resolution of the printer requires for a given print size. As such, it is a way of dealing with pixel count, not sensor size. Magnification matters for different reason: the more magnification, the more apparent the resolution limits of the lens. Magnification therefore is a function of sensor size. Am I wrong?

you raise an interesting point: when higher pixel density gives you more pixels on a smaller sensor, when is the advantage of that offset by the greater magnification required? It's taken as a given by many wildlife photographers that greater pixel density matters more because of the severity of many of the crops they have to make. Ditto, some field macro. But I have never seen real data.

Perhaps I overstated my earlier comments. I wasn't meaning to imply that neither sensor size nor pixel count matters. My point is that they matter a lot less than many people think. Going back to my Bergen photos, it's definitely not the case that prints from that 12 MP Lumix are on a par with those from my 30 MP 5D Mark IV. They also take a lot more processing to look good. I think both size and pixel count work against the Lumix. My point is that I can get quite serviceable 13 x 19 prints from the Lumix, and while I (and I'm sure you) can see that they aren't quite up to the standards of the FF prints, most people don't notice. And that's an extreme example. The Lumix LX-100 first version is a very big step down from something like the Fuji XT-4.

An interesting take on this is a video by Andy Mumford, a British photographer who makes his living entirely with landscape photography. He largely abandoned his FF Nikon gear 6 years ago in favor of APS-C Fuji gear, and he says that while he still uses the Nikon from time to time, he has no regrets about switching to the Fuji, and he claims that the camera used to take a photograph has never been a criterion for selecting images for clients or for other uses. You can see the video here.

I personally like shooting with FF--although with each passing year I'm more annoyed with the weight of the gear--and I'm certainly not advising against it. I think if a truck ran over my gear, the probability is better than 50-50 that I would by FF again. I just wouldn't rule out APS-C now, given technological progress.

Re a system: unless David has more gear than he noted, I wouldn't pay much attention to the fact that he has Canon now. His standard zoom is a decent kit lens but not what I would use if better large print quality were my goal, and it won't work on Canon FF cameras anyway. The 75-300 is unfortunately a weak lens, and it's the first thing I would give up if I were interested in top print quality. I personally like Canon gear, but I think he's in a position where he can cast his net broadly.

[xpatUSA]

Can you tell me why anyone should care? Most people are interesting in producing images and don't really care how we get there.
I know of no one (or perhaps you are the exception) who really cares what happens in the background.

The usual slap-down. It seems that responding to technically inaccurate statements is frowned upon by Moderation here.

Again, can you tell me why anyone should care? Camera manufacturers are all about pleasing their customers so that they can produce good images (whatever that may be). What is wrong with pre-processing?

I did NOT say that there is anything wrong with pre-processing!! My response was to David, who appears to think that what's in the raw file is exactly what the sensor saw.

I still remember years ago when Sony was accused of doing so and the astrophotographers were all concerned about "pixel killing" star removal. The [µ4/3] camera makers have long done this to correct for optical issues with their lenses. In the early days of the Lumix line, Leica apparently had a contractual agreement with Panasonic that Leica branded lenses could not used these enhancements. I suspect that these days are long gone.

Manfred, it seems that, by the tone of your responses, you are deliberately discouraging technical comment to the extent that I am beginning to feel unwelcome in these fora. If that is the case, so be it.

Should I ignore technical errors in future and never comment on any issue involving "The Print" other than in terms of Standard Photographic Vague (SPV)?

David, feel free to ignore earlier comments about what constitutes raw data and I apologize for wasting your time. After all it apparently does not matter and nobody should care so long as The Print is OK.

I'll leave you with some "pure" raw composite data:



No demosaicing, no post-processing, scaled to 16-bit with 2.2 gamma.

[William W]

David - I think it all depends on what you look at as a "large print". . . .

Unless you are shooting in near-ideal conditions (i.e. a heavy duty tripod), any reasonably modern lens produces excellent results. I try to use a heavy-duty tripod whenever I can, to minimize camera movement. While image stabilization works well, I find that it is not as good as shooting from a very stable platform. Most of my work is landscape, street photography, portraiture and still-life work (the latter two are primarily studio based and even when I do location shooting, I use studio lights). I don't do any macro work or a lot of flowers, so I defer to Dan on those topics.

I don't understand the sentence: and if I do understand it I disagree.

"Unless you are shooting in near-ideal conditions (i.e. a heavy duty tripod), any reasonably modern lens produces excellent results." means "any reasonably modern lens produces excellent results provided that you are not shooting in near-ideal conditions (i.e. a heavy duty tripod)"

Manfred, is that what you meant to write?

WW

[ajax]

Several participants have commented about the importance of lenses in relation to sensors. I'd have to say that my intuitive thinking would be that lenses are the starting point when seeking accuracy. In that, it is the optical capability of the lens that projects light on the sensor. If that is NOT done well the sensor can't make up for it after the fact. At least that would be my way of seeing things.

That has triggered another thought on my part. Apparently in the case of Canon they have EF mounts for DSLR cameras and RF mounts for mirrorless cameras. The lens that presently appeals to me is the xx 24-105mm f/4L IS II USM. Apparently xx can be either EF or RF. Implication being they are the same other than the mount which pertains to the type of camera they will be mounted on. There is an adapter for EF lenses that normally mount on a DSLR camera that allows them to be mounted on a mirrorless camera. However, I don't think it works the other way around (i.e., RF lens cannot be mounted on DSLR camera).

As a result, one could obtain the EF version and use it on either type of camera. Might it make sense for me to do that which would both allow me to use it on my present cameras as well as a Canon mirrorless camera (with the adapter). This would also reserve the option of upgrading to a better DSLR camera if desired. Thinking being that at least in the near future there are lots of DSLR models that sell for less money than the mirrorless camera that could come into play. It also puts off the need to decide on FF or crop sensor.

Question being what are the negatives associated with using the adapter to mount the lens on a mirrorless camera? I would like to reserve the option to go mirrorless but putting that off might be a better approach.

This approach allows to more gradually tip toe into the swamp of things that I've NOT yet experienced.

[Stagecoach]

Question being what are the negatives associated with using the adapter to mount the lens on a mirrorless camera? I would like to reserve the option to go mirrorless but putting that off might be a better approach.

As a Nikon mirrorless user using an adapter with some of my older lenses I would say,

a) It adds extra length which is most noticeable with a heavy lens such as a 70-200 f/2.8 as it affects balance. E.g small lighter camera with heavy mass extending further.

b) Additional 'weak' point of flange, may or may not be significant.

[xpatUSA]

Several participants have commented about the importance of lenses in relation to sensors. I'd have to say that my intuitive thinking would be that lenses are the starting point when seeking accuracy. In that, it is the optical capability of the lens that projects light on the sensor. If that is NOT done well the sensor can't make up for it after the fact. <>

Yes, the sensor itself can not "make up for it after the fact".

However, I quote "Every lens has its own specific distortion and optical defects. If you’re shooting to RAW, you can suppress these automatically using the corresponding LCP."

https://learn.zoner.com/how-lens-pro...r-life-easier/

Lens correction "after the fact" is common in the world of digital photography. My Panasonic cameras know what lens is connected to them and they apply lens correction automatically when creating the JPEG output, as has already been mentioned by our worthy Moderator.

My editor has dot-lcp files for many lenses and can correct for those lenses' aberrations with a click of the mouse.

[ajax]

This thread has meandered quite a bit, and while there is nothing wrong with that, I think it's important to keep the OP in mind, and this would in my opinion be very bad advice for David. David is wondering whether the next step up for him from a beginner's kit should be full frame. I and others have suggested that it might or might not be, depending on what he is going to shoot. However, I'm willing to walk the plank and say that jumping to medium format would be a big mistake for him. It makes sense for some people, e.g., people who shoot landscapes and print very large, but it would be a bad choice for David, from what he has written. It would be a bad choice for me as well, given what I shoot and how much I lug my gear around.

Dan

Yes it has but it has also been very informative, at least to me, for which I’m most grateful.

…

David--If you go to my website and look at the "built environment" gallery, you will see photos taken with a cheap Lumix micro four thirds camera, a 50D (an old crop camera with an older sensor than yours), a 5D Mark III (I think 22 MP), and a 5D Mark IV (30 MP). I'm guessing that if you don't click on the little information links, you won't be able to tell which images were taken with which camera. You can see differences once you print at about 13 x 19, but I have a 13 x 19 of one of the Lumix images that is both in my wall and in a gallery. Once you get past 13 x 19, especially if you crop, one can sometimes see a small difference between the crop and FF, but it's usually not big. Keep in mind that all but one of these cameras is old, with sensors that aren't up to current standards.

...
Dan

The website is very nice but I do get your point without having to do that.

Display devices just aren’t good enough to show all the detail that sensors are able to detect, at least in lots of (most?) cases. This is also a problem for those of us developing photos on our computer. Even the good quality displays that you can put on your computer don’t have the resolution and depth of color that might be needed to make such distinctions. Of course, this is based on my own experience and a bit more intuition since I’m NOT using the very best displays.

Also, the lossy compression technique used for jpg files washes away some detail.

Good prints just look better. To my way of seeing it, I’d like to avoid loosing quality, that my display can’t show, when printing.

I also have to admit that the prints I’m producing now look quite good to me. This is why I’m so interested in going bigger than the 13x19 limitation of my Canon Pro 100. Of course such prints would be limited to selected pictures thought to worthy of the cost.

Based on suggestions found herein, I like the idea of using a service provider as way to get started making larger prints. If any of you have such experience and can suggest providers you'd recommend I'd be most grateful for such input.

...

So I'll walk a little further out the plank. Unless you are printing very large or shooting a lot in low light, I think the choice between FF and crop won't make a large difference for you. What will make a difference if you shoot wildlife or macro is pixel density, and for that, crop sensors are usually superior. All that aside, the biggest factor apart from the photographer influencing how sharp images look is usually lenses. A few years ago, I would have suggested that the first step might be replacing the 75-300, which is the weakest link in your kit. However, it's a different world now because DSLRs are disappearing, and it may not make sense to buy another lens designed for them if you are going to end up with a mirrorless body. YOu can add an adapter to use them with mirrorless bodies, but it might work out best to have a plan for what system you want to end up with.

Dan

Your mention of low light is interesting. I’ve become a fan of Elle Stone and her website called Nine Degress Below Photography which is mostly about post processing and a little bit about photographing. As a result, I’ve become a strong believer in lower, if NOT low, light.

This makes me think that an example might be of some help. A recent attempt of mine has been to photograph sunrise. This turned out to be quite difficult and probably a good learning experience, at least for me. However, after many attempts on different occasions I finally got something I like. The opportunity for this photo came about as result of some picturesque cloud formations along with very calm water which provides for good reflections. Here is the camera developed version. Then I have 2 versions that I developed on my computer. This one using Rawtherapee and another one using the Cannon supplied software called Digital Photo Professional (DPP4).

Which is best is a matter of opinion. Each of you are allowed to have your own. This happens to be one of those cases where I sought the opinions of others about which they liked best. Your most welcome to weigh in on that.

However, the real point I’d like to make is that what you can get from the camera is quite limited compared to what possibilities exist. It is worth pointing out that this does NOT require raw files. I started learning to do such when I only had point and shoot cameras. While the camera produced jpg files can be similarly adjusted this does take time and effort. My idea is why NOT start with the best quality file if you are going to expend the time and effort. Even when the difference is undetectable on displays and small enough prints, the master file (actually file/s) is better if it was done using the more detailed data obtained from a raw file. Even if better means for some future application that may or may NOT happen.

[ajax]

...

However, I quote "Every lens has its own specific distortion and optical defects. If you’re shooting to RAW, you can suppress these automatically using the corresponding LCP."

https://learn.zoner.com/how-lens-pro...r-life-easier/

...

Yes, lens correction is a standard feature of the software I use for post processing. I do use it even though as with other aspects of quality discussed herein, it is NOT something where I've been able to see the improvement on a display device. I'd also be surprised that it can substitute for a better lens. In that, my expectation is that it can adjust for known defects in the specified lens.

[DanK]

No, the RF and EF lenses with the same focal length are not the same lenses apart from the mount. I don't know how different the 24-105s are, but the 70-200s are entirely different in design.

Yes, you can use EF lenses on R-mount cameras. There is no appreciable loss of function if you use Canon's own adapter.

No, you can't go in the other direction. Mirrorless cameras have shorter flange distances, and the lenses designed for them make use of this. The back of lenses designed for DSLRs have to be held farther away from the sensor, which is all the adapter does.

I have quite a few expensive EF lenses, and I would plan to keep them and use them with an adapter if I were to buy a Canon mirrorless camera. In fact, when Canon announced that they were stopping production of some EF lenses and that one of them was the new and superb 70-200 f/4 L II, I snapped one up.

However, if I were starting from scratch, I would probably get RF lenses unless I had reason to use a DSLR also.

A few things were added to the thread while I was typing:

Lens correction deals with things like barrel and moustache distortion and vignetting. it does not address softness, that is, lack of resolution.

It is worth pointing out that this does NOT require raw files. I started learning to do such when I only had point and shoot cameras. While the camera produced jpg files can be similarly adjusted this does take time and effort.

If I understand this, it's wrong. Creating a JPEG file loses information. It also bakes in various editing decisions, depending on the picture style you use--e.g., contrast. In many cases, the reduced-data JPEG file is fine. In many cases, it's not. If you want real control, shoot raw.

[xpatUSA]

Display devices just aren’t good enough to show all the detail that sensors are able to detect, at least in lots of (most?) cases.

Is it possible to clarify the above statement? I ask because I have never had a problem showing "detail" on my humble 24-inch 1920x1200px sRGB monitor. Every app that I have can zoom in as much I want. One image pixel can be for example 40 pixels on my screen.

Of course, in our world, "detail" means whatever Chuck Norris says it means.

[DanK]

Is it possible to clarify the above statement? I ask because I have never had a problem showing "detail" on my humble 24-inch 1920x1200px sRGB monitor. Every app that I have can zoom in as much I want. One image pixel can be for example 40 pixels on my screen.

Of course, in our world, "detail" means whatever Chuck Norris says it means.

I think the statement makes sense if you refer to the image as a whole, not a portion blown up. For example, the 27" NEC monitor I use for photo editing has roughly 3.7 MP. On my Canon printer, a 13 x 19 print is 22.2 MP, and I can get that with no upsampling. A 17 x 22 print is 33.7 MP, although with my camera, that requires a little upsampling even without cropping--that is, some of the detail is created by software.

[Tronhard]

Personally, I think that the final image is dependent upon a set of elements, some of which are to do with the technology and some to do with the photographer: in no particular order:

Output:
How big the image is going to be
The quality of the printer, paper and ink and the lighting in which it will be seen
Input:
The quality of the sensor makes a difference - a medium format camera should potentially render a bigger image, however...
Some depends on the sensor size relative to the distance to the subject. So, shooting a rutting stag at some distance with a long lens will have different results depending upon how much of the frame is used up by the subject (and how much you want it to occupy). For example:

George has a Nikon FF camera with an FX lens (which fits FF and APS-C bodies) that has a FL of 300mm. He also has an APS-C camera with the same number of sensor pixels. He has to decide which body he should take to go bird shooting. If he takes the FF camera the lens + sensor combination will give him a FoV of 300mm, but if he puts the same lens on his DX body (with a crop factor of 1.5) his FoV will be 300 x 1.5 = 450mm! The sensor has pre-cropped the image and that makes the bird fill a lot more of the image area, or he doesn’t have to get so close to the bird.

Not only that, but because the camera sensors have the same number of pixels (let’s say 30MP), and because the APS-C sensor is smaller, the pixel density of his cropped image will be much higher, giving him more resolution! If he took the FF camera, he could crop the FF image down in post-production to the same FoV as would have been achieved by the Crop sensor, but in doing so he will lose pixels and that is quite significant. If he cropped the FF image down to match that of the APS-C image, the resultant MP count would be 20MP / (crop factor)², so that would come out at: 30 / (1.5 x 1.5) = 13MP, which is a MUCH lower resolution image than the original 30MP that is achieved by the crop sensor. That is why most bird photographers will choose high-end APS-C camera bodies as the crop factor actually renders a benefit to them.

The quality of the lens, it's suitability for the type of image. For example, using a normal lens reversed for close-up work, shooting with a modest focal length and cropping significantly. Using poor optics in general

The human element. No gear will make up for poor technique. Camera shake, poor focusing, inappropriate DoF, poor exposure all will have a detrimental impact on the image.

It is easy to look at these issues as if it was all about the tech; but, speaking for myself, I have taken images with non current state of the art gear that looked and printed well at the time, so why would they not do so now?
Again, we come back to the way in which we intend to evaluate an image.

[William W]

David,

Please refer to Dan’s Post (#115).

In a nutshell I concur with all his points: my situation is similar to Dans’ –

My DSLR kit is Canon. It is extensive. It was born from commercial camera kits; we owned W&P Studios and changed our gear from Nikon, Minolta and Mamiya (Film) to Canon (Digital) in 2004.

I also have Canon Mirrorless Kit, and I used the Canon adapter so I can use my EF Lenses – the salient point being that I already own a bag full of EF Lenses and I see at this time only wasting my money buying dedicated Mirrorless Lenses when the adapter works fine.

Grahame made a point about adapters and balance. I am excessively tactile and my cameras work as a tool being an extension of my hands. Big lenses with an adapter on a (smaller than designed for) Mirrorless camera using an adapter can take some getting used to.

Please refer to Your Post (#110)

You’re in a different position.

I very much like the fact at you’re thinking about the ‘progression’ of your gear.

It is difficult to know exactly where the path of new gear takes you however, I have found that at the least, having a plan which has open opportunities and limiting dead ends is pretty good thinking.

I have the EF 24 to 105/4 L IS – it is a good lens. I have used the MkII and it is superior in IQ – I saw no advantage in ‘upgrading’ to the MkII when it came onto the market.

I also have the EF24 to 70/2.8L. (N.B. I specifically chose not to the MkII version, even though, reportedly the MkII is very slightly sharper wide open).

If I were buying now, I would buy the 24 to 105/4L IS MkII and the (original) EF 24 to 70/2.8L.

I mention these two lenses because either one is the best suited Standard Zoom Lens for any Canon FF Camera for 99% of most Photographer’s requirements.

In a nutshell, the choice of one of these lenses over the other comes down to: the value for you of one faster stop vs. the extra FL and/or the IS.

The (professional) choice, for me, was simple – we initially had the three F/2.8L zooms, that’s a standard press kit: it’s that simple. The value of the 24 to 105/4L IS was never for my professional use and I only bought it when I sold the studios and took more holidays. I think that the 24 to 105/4 choice is one which better suits 99% of General Photography, especially in 2022 (and this is an important point), where the newer Cameras accommodate higher ISO at reasonable quality – which, in most situations, cancels out many of the arguments/requirement for F/2.8 over F/4.

You may take all that I’ve written above as simply saying I think your thinking process is on a good track, I've simply given you a potted history of my journey for points of comparison and contrast.

***

I have re-read all your posts in this thread ad perhaps I have missed it, but nowhere can I find that you mention your typical Photography Genre and/or Subject matter. I think this is important because it will impact, to some degree though never usually to any great degree on your choices.

Though often nuanced, and often irrelevant in many cases, the fact remains that some Camera types (and Lenses) are better suited than others, to some few specific Genres.

WW

[dje]

That has triggered another thought on my part. Apparently in the case of Canon they have EF mounts for DSLR cameras and RF mounts for mirrorless cameras. The lens that presently appeals to me is the xx 24-105mm f/4L IS II USM. Apparently xx can be either EF or RF. Implication being they are the same other than the mount which pertains to the type of camera they will be mounted on. There is an adapter for EF lenses that normally mount on a DSLR camera that allows them to be mounted on a mirrorless camera. However, I don't think it works the other way around (i.e., RF lens cannot be mounted on DSLR camera).

As a result, one could obtain the EF version and use it on either type of camera. Might it make sense for me to do that which would both allow me to use it on my present cameras as well as a Canon mirrorless camera (with the adapter). This would also reserve the option of upgrading to a better DSLR camera if desired. Thinking being that at least in the near future there are lots of DSLR models that sell for less money than the mirrorless camera that could come into play. It also puts off the need to decide on FF or crop sensor.

Question being what are the negatives associated with using the adapter to mount the lens on a mirrorless camera? I would like to reserve the option to go mirrorless but putting that off might be a better approach.

Hi David

I'm very late in jumping in but I must say that if I were in your position, I wouldn't invest in an EF-24-105mm lens, not because the lens is no good but I think your strategy is flawed.

24mm is not quite wide enough for some landscapes on an APS-C camera and secondly the lens is optimised across the full "image circle" of a full frame camera. You could get a cheaper EF-S lens that's been optimised for APS-C with a wide end of 16mm for landscapes.

If you do go Full Frame at some stage in the future, it's likely to be mirrorless (at least it would be for me). I made the switch from APS-C DSLR to Sony mirrorless (both APS-C and FF) some years ago and there is no way I would go back to DSLR now. If you do go FF mirrorless, you would be far better to start with a zoom lens designed specifically for mirrorless. They are generally smaller and lighter and the use of an EF adaptor adds cost and extra length again.

Just my 2c

Dave