Full Frame vs Crop Frame Comparison

[Dr Bob]

I did some reading here on full frame and crop frame cameras and have a better understanding of frame size and pixel density for each. However, I am interested in understanding better what are the types of photography (portraits, landscape, macro, etc) that each best performs and why. I would appreciate any comments on this that the CIC very knowledgeable audience can provide. For reference, I currently shoot with a Canon 7D.

Thank you in advance.

Dr Bob

[RustBeltRaw]

Disclaimer: everything below is a generalism, and assumes constant resolution unless stated otherwise. The technical details of different cameras make the situation fairly complex. I am going to focus on what each sensor type is best at, and ignore the fact that good photos depend way more on the photographer and the situation than the camera.

First off, the cost of a sensor increases with the area (or faster). So if you maintain the same aspect ratio, doubling the width will quadruple (or more) the cost. Big sensors cost exponentially more than small ones, which is highly relevant to manufacturers, and highly annoying to buyers.

The smallest sensors, used in cell phones and low-end P&S cameras, have extremely large depths of field and generally poor high-ISO performance thanks to small pixels. This makes them good for small, light, highly portable devices, but it's difficult to control depth of field or get clean files from low light. People shoot everything with these cameras, with widely-varying degrees of success. Macro photographers love small sensors for their inherently-large depth of field, which is very useful to combat close focusing distances.

1.5x and 1.6x sensors are extremely common. They're large enough to provide good depth of field control, even though their DoF is about a stop deeper than a full-frame sensor, and their photosites are big enough for very good high-ISO performance. The extra reach makes them a very good option for action, wildlife, and other applications requiring very long focal lengths or great distances between the photographer and their subject. Nominally, the aforementioned subjects need slightly more light than with a full-frame sensor due to the inherently noisier nature of smaller photosites at a given exposure level.

1.3x crop sensors exist, but they're rather rare. I only mention them because I own one, and I don't want the poor thing to feel left out.

Full-frame sensors are the standard to which everything else is compared. Very good DoF control, the best-available high-ISO performance, and they're generally paired with other high-end features that make them attractive to professional photographers (such as high burst rates for action photographers). Full-frame sensors work well in the studio, but their advantages are clearest in very low light. You will find these everywhere professional photographers work.

Finally, the big mamas. Medium-format sensors. Huge pixels, huge sensor area, huge resolutions. DoF becomes so narrow that f3.5 produces a DoF comparable (depending on exact sensor size) to f1.8 on full-frame. Their high-ISO performance should be great thanks to large pixels, but these cameras are generally poor at higher ISOs digital SLRs don't blink at, since their amplifiers and noise reduction algorithms lag behind the state of the art. These limitations mean that MF cameras are best suited to bright environments, such as landscapes and flash portraiture.

But I encourage you to ignore this, shoot what you have, and defy conventional wisdom. Cell phone street photography can be fantastic. Action can be captured with MF cameras. Crop sensors still pull great shots out of low light environments. Knowing your tools makes sure you know how to get shots out of them, no matter where you are.

[rpcrowe]

I agree with the comments from Lex, above. When I mention "you" in the following comments; I am referring to "you" in general rather than "you" Dr. Bob...

I am a crop camera shooter, also using the 7D... IMO, a crop camera, with a top-line lens and used correctly can function well in portraits, landscape, and macro venues. They could even be considered a bit better for wildlife because of the crop factor. The following is simply my opinion regarding crop vs. full frame formats...

Many crop camera photographers seem to think that switching from a crop to a full frame camera will automatically transform their photos from the mundane to great! The lines from an old song come to mind, "It 'taint necessarily so!"

If you are using a top-line lens and crop camera and are not producing excellent imagery, it may very well be "The Indian, not the arrow at fault!" You may very well be better off reconsidering your techniques.

If you are using a crop camera and less than a top-line lens; IMO, you might very well see more of an improvement in image quality by switching to a better lens.

If you are using a crop camera with a top-line lens and are getting excellent imagery, you may, under some circumstances see an improvement by switching to full frame gear. This might be most noticeable when shooting with very high ISO levels. However, for most uses, IMO, the crop camera will be sufficient. I tend to often creatively use flash or other lighting instead of shooting available light at a high ISO.

As far as making large prints; I can make some very large prints from my crop camera images, especially since I purchased the OnOne Photo Suite which has Perfect Resize; a refinement of Genuine Fractals. Along that line, good post processing skills (especially sharpening) are necessary in producing any decent imagery...

And, if you are not using the latest edition of the Canon 5D (the 5Diii) you "may" even see a slight drop in image quality when following fast moving subjects since the auto-focus capability of your 7D is a bit better than that of a 5D classic or 5Dii.

I don't know where the 6D sits in relationship to other cameras. I have never really investigated the 6D.

The latest 1D full frame cameras will give you excellent auto focus as well as excellent image quality. The 5Diii will also provide excellent A/F. However, you pay for the advantages of the 1D series cameras (and there are a lot of great advantages) in the initial price of the camera and also in the weight of that camera. The 5Diii is also considerably more expensive than a 7D. Whether the advantages of the 1D series or 5Diii are worth the extra price is totally up to the individual photographer.

[kdoc856]

Lex gave an excellent overview. I recently added a full frame, in addition to my APS-C camera. I don't think of it as an upgrade, nor was I unhappy in any way with my APS-C, and many shots show no higher sharpness just on the basis of the body. They are just different tools in the tool box. I do a lot of low light shooting with natural light, and the ISO advantage is huge- that is to say the noise control allows me at least extra 2 f stops, and lets me handhold a lot of shots that previously I'd either need a tripod, or just couldn't shoot at all. I think nothing of shooting with an ISO of 1600, and with noise reduction software have gotten many usable shots at 6400. No way with the APC-s.

The 40% increase in sensor size surface area also lets me crop quite aggressively without paying a resolution price- very handy when I don't have the focal length or physical access to be more selective in excluding extraneous material in a scene.

I use the APS-C to get extra reach when I'm using a long lense for sports or potential wild-life shots.

Hope this helps a little.

[ajohnw]

Things change so in some ways it isn't easy to answer the question. I've used full frame, APS and micro 4/3. In my view there really isn't much difference between a typical APS crop fact or 1.6 to the m 4/3 one of 2. On the other hand I was rather surprised how good a shot my iPad took of part of the living room with a wall light in the shot - tough thing for any camera. Another. I recently found my m 4/3 camera has an iso setting of 25600. Much to my amazement with post processing it's usable. This fact probably explains why ipads and phones etc can take decent shots over a rather wide lighting ranges. I have no idea yet at what iso this camera can be used at without processing noise out. Shutter speed will have an influence on that as well.

I wouldn't be inclined to argue with the general trends already mentioned but changes in sensor technology have made something of a dent in them several times over the years. Cameras have historically been sold on the basis of ever increasing pixel counts for a long long time. A pixel is a bit like a bucket, light fills up smaller one more quickly = higher base iso's. Noise fills them up more quickly too so improvements have to be made in that direction as well. Actually I suspect my camera and some others now take many shorter exposures and add them up rather than taking a single long exposure. Another way of improving noise performance. They might also do the same thing to average noise.

As to M 4/3 I usually post a link to this shot. It's nothing special. I have taken lots of them. It gives an idea of max size

http://backup.cambridgeincolour.com/...98480/original

I do have a portrait of sorts too

http://backup.cambridgeincolour.com/...82543/original

Both expand to 100% resolution views when clicked on after they open. In real terms they show that given decent lenses that suit the format the max image size on a PC is determined by the pixel count on the sensor.

In terms of landscapes, portraits etc I don't think that there is much difference in any format either. The increased depth of field works to a maximum sized circle of confusion. For the same sized final image the size of this circle has to be scaled down according to the size of the sensor so the increase in depth of field is not as much as people expect from purely the focal length. It's interesting to tie these size of circle of confusion with max print size but as I see it if not viewed from the end of the nose either of the 2 shots linked to could be printed at that size. One of the problems with things relating to circles of confusion is that they go way back in time and people just continue to quote things based on them. JPG's type processing is probably the reason that these prints would be ok even though they are well over the size the ancient rules predict. PP and the subject matter will have an effect too. If some one uses a coarse pitch monitor they might disagree but then the image size would be a lot bigger.

Really if you upgrade your camera it's best to look for distinct improvements in noise, usable dynamic range and functionality. In my view of late pixels counts are of limited interest. Dpreview is probably the best site to use for this but there is a need to spend some time understanding why they produce the results as they do and what they mean in practice.

I'm not trying to push you to m 4/3 either. It's a rather big jump from a typical canon camera. My reason is weight of the kit and that's about it.

John
-

[mahfoudhhi]

Very nice replies here. I learned more information here.
Thank you very much everyone.

[PhotomanJohn]

Bob - I have a D7000 and am about to add a D800 to my collection of gear and am not doing it because I expect my photography to noticeably improve. It is more that I like the feel of the D800 which fits better in my hands and some of the additional controls and features that the camera has which I will enjoy using. To me the only down side of the new camera is the cost. Some talk about the extra reach of the crop sensor cameras and the increased DOF. In the case of comparing the D800 and the D7000 neither is really valid. The D800 has the same pixel pitch as the D7000 with similar sensor performance so with the same lens at the same distance, one can put the D800 in the DX mode or simply crop the resulting file in post and have the identical image in all respects to what would be taken with the D7000.

To me the bottom line is I like the feel and features of the D800 enough more than I do those of the D7000 that I have convinced myself to buy the camera along with upgrading a couple of my lenses. And in some situations the additional performance of the camera may actually improve an occasional image. My current situation is much like yours in that we both have excellent APS-C cameras. I personally would not buy a new camera just because it has a bigger sensor.

John

[DanK]

Bob,

I have a 50D and recently bought a 5DMark III after many months of dithering, so I have given this a lot of thought.

First, like John, I bought the new camera in part for features other than the sensor size. For example, the AF on my 50D is fairly rudimentary. The AF system on the 5D Mark III is simply amazing. If the long-rumored 7D mark II had actually appeared, it might have had most of these features, and if so, I am not entirely sure I would have bought the 5D. So here is my list:

--Cost. Winner: crop sensors
--Weight and size of both body and lens. Winner: crop sensors
--DOF: neither is better; they are simply different. The best explanation I have seen is here. For people doing posed portrait work, the shallower DOF FF gives you at any given aperture is an advantage. I don't do that; I do candids of people, and f/4 even on a crop is often fine. For macro work, assuming a fixed aperture, the shallower DOF of FF is a disadvantage, but you can compensate by closing down the aperture of the FF camera). for a great deal of work, it makes no difference, as you can adjust the aperture to get what you want. I don't see it as a reason to pick one of my cameras over the other, although I do need to remember to close down further on the FF to get the result I was used to with the crop.
--Low light performance/ noise. No contest, at least as of now. FF is better. That is one of the primary advantages for me.
--Printing large. Advantage goes to FF. For web presentation, I doubt it makes any difference.
--Chasing animals. No question: crop is superior because of the greater reach. FF folks often say you can crop to compensate, but that ignores pixel density. If you crop most FF cameras to produce reach comparable to a crop, you will have a lot fewer pixels on the subject. So, with a crop you can make do with shorter, lighter, and cheaper lenses.
--macro. A complicated question. The better low-light performance of the FF is an advantage. The longer reach of the crop is an advantage. At minimum working distance, the crop will put more pixels on the subject. You can compensate by adding tubes, but you need quite a bit of extension to make FF comparable to crop. I'm still uncertain about this. I have a hunch I may end up using my crop for a lot of bug chasing, to have less weight and less need for extension, but it's the wrong season, so I don't know yet. For table-top macro, I suspect I will end up using the 5D.
--Diffraction. This is a hornet's nest, since people argue about how serious an issue this is. My experience (not formal testing) suggests that at 8 x 10, diffraction is not a practical issue until you stop down severely. However, whatever that point is for you, for a given print and viewing distance, it comes at a smaller aperture with FF.

For me, I think the biggest tradeoff specific to sensor size was low-light performance vs. reach. For others, the tradeoff would be different. I greatly enjoy the 5D Mark III -- it's by far the best camera I have ever owned, and I bought my first SLR in 1968 -- but if I had the same features on a crop body with moderately better noise characteristics than the current 7D, I am not sure it would have been worth the splurge for me. But that is water over the dam: there is no turning back now.

Dan