Is Full Frame sharper than APS-C? My answer here

[NormanBates]

I have to decide between APS-C and full frame, and one of the things to look at is sharpness.

With APS-C, as opposed to full frame, you're only using the central portion of the image circle, the sweet spot: you're not using those blurry corners, and this should lead to sharper pictures. On the other hand, because the sensor is smaller but final image resolution is the same, you're blowing up the analog image that the lens projects on the sensor; this is bad, because every defect will be magnified. Which force is bigger?

I ran some tests trying to find out, shooting stills with a 5D2.

My conclusion: as a general rule, a lens will always be sharper on full frame than on APS-C, both in the center and in the corners.

You can see the tests and read a bit more about them here:
http://www.similaar.com/foto/lenstestsff/lenstestsff.html

 

[Sporgon]

Errr..... you're not "blowing up the analogue image bigger" if the APS or FF or any other sensor size for that matter have the same number of digital info capture points - pixels. MP is MP of digital info - the computer program enlarges the image.

The issue on smaller sensor regarding "enlarging" is can the smaller pixels capture the info accurately. In that respect they are certainly improving.

 

[NormanBates]

What I mean is that you've got an image circle, a bunch of photons that are projected back by the lens. This is anlogue. You put the same ~20 million photosites behind that in both cases, but on APS-C each of them is smaller.

If you want to think it like this, it's as if, instead of a ~20 Mpix full frame sensor, you had a ~50 Mpix full frame sensor, and then you used only the central portion, and enlarged it so that it is now the full image (and this is where my "magnyfy issues" argument comes in: you are enlarging the image, and this magnifies any defects in the lens).

So, yes, as you say, the thing is that the pixels are smaller, and this reduces quality.

In any case, the proof is in the pudding: fullframe gives a lot more detail.

 

[Loswr]

No surprise there. But just want to point out, this is a DxO-like conclusion. You used a 5DII for the test, I can say from experience that in many situations, the APS-C sensor in my 7D delivered sharper pictures than my 5DII, by virtue of its better AF system.

 

[East Wind Photography]

I don't disagree with your conclusion. The only reason I might upgrade to an aps-c from my 5d3 would be if it was also around 22mp. However your test results would be more compelling if the target distance was also compared when at the same distance. Effective comparison of sensor sharpness should keep all other factors the same. Distance plays an important factor in a lenses resolving ability.

Still I do agree with your results when applied to the field. I shoot primarily wildlife now with my 5d3 and resolving ability isn't even comparible to my 7d.

 

[tortilla]

Nice work, but have a look at center crop APS-C 24 mm, how come that f9 is sharper than f5.6?

You already mentioned DoF on your site:

Add to that the fact that for a given depth of field you're closing down the iris further on your full frame lens (and this helps sharpness too) and I think the result will be pretty much general.

When you want large DoF than we're talking about f/8 (APS-C) and f/11 (FF) or so. IMO closing down even further makes the image actually softer, due to diffraction and lens design.

 

[Sporgon]

I don't disagree with your conclusion. The only reason I might upgrade to an aps-c from my 5d3 would be if it was also around 22mp. However your test results would be more compelling if the target distance was also compared when at the same distance. Effective comparison of sensor sharpness should keep all other factors the same. Distance plays an important factor in a lenses resolving ability.

Still I do agree with your results when applied to the field. I shoot primarily wildlife now with my 5d3 and resolving ability isn't even comparible to my 7d.

+1 to that in practical application.

 

[aznable]

No surprise there. But just want to point out, this is a DxO-like conclusion. You used a 5DII for the test, I can say from experience that in many situations, the APS-C sensor in my 7D delivered sharper pictures than my 5DII, by virtue of its better AF system.

a lot of people forgot that a camera it's not just a sensor

 

[NormanBates]

No surprise there. But just want to point out, this is a DxO-like conclusion. You used a 5DII for the test, I can say from experience that in many situations, the APS-C sensor in my 7D delivered sharper pictures than my 5DII, by virtue of its better AF system.

Indeed. You may also notice that I listed my best 3 lenses there, and none of them has AF.

Nearly everything on that site was created because I needed it, including these tests. They're made for me, and designed for my very special needs.
Everybody should test their equipment to see how well it fits their needs, what can be done and what's not a good idea, and where there's room for improvement. I'm just sharing my results. They shouldn't stop anybody else from running their own tests.

 

[TrumpetPower!]

All else being equal, especially including the final print size (a factor which often gets ignored and confused with respect to pixel peeping), the larger the format, the better the image quality in every aspect. That includes sharpness as well as noise / grain as well as dynamic range as well as anything else you care to mention.

Of course, all else is never actually equal, which is why it's important to compare complete systems. Neoruanatomist made that point; better theoretical maximum image quality is useless if you can't properly operate (in the specific case, focus) the equipment.

There's another factor to consider. Image quality with even entry-level DSLRs and kit lenses is superlative. Anything going to the Web, 4x6 prints, and even 8x10 prints...there's no practical real-world image quality difference between a Rebel and medium format (with the exception that you can more easily get a greater amount of background separation with the larger format).

Once you're printing on a machine that won't sit on your desktop, though, the difference between the different formats starts to become apparent. You can make a great roadside billboard with a Rebel and a kit lens, but walk up to a pair of door-sized fine art prints, one made with said Rebel and the other with an 8x10 view camera, and even your half-blind great uncle will be able to tell the difference.

Cheers,

b&

 

[tortilla]

There are certain situations where a crop delivers higher resolution than a FF with the same lens (even disregarding DoF, AF and the corners). Roger M. Clark has nice shots of the moon with the 5D2 and the 7D that show this:

www.clarkvision.com/articles/digital.sensor.performance.summary

 

[NormanBates]

Indeed, if you keep lens and distance constant, APS-C will give you lots more detail. Because it's just as if you had a 50 Mpix full frame body and cropped the center of the image.

But that's not my usual scenario. For me, the right comparison is "Elmarit-R 35mm f/2.8 on APS-C set at f/2.8" vs "Summilux-R 50mm f/1.4 on full frame set at f/4.5". Similar FoV, similar DoF --> full frame is clearly sharper.

As I said, if this is not the way you shoot, run your own tests designed for your usual scenario (actually, you should run them anyway!).

 

[Loswr]

Indeed, if you keep lens and distance constant, APS-C will give you lots more detail....

Yes, at low ISO. But once you raise the ISO a bit, the FF delivers more detail in focal length-limited situations.

 

[AlanF]

My conclusion: as a general rule, a lens will always be sharper on full frame than on APS-C, both in the center and in the corners.

There is no general rule. Whether FF beats APS-C depends upon distance, size of the detail on the subject and whether it will span the length of two pixels, how good the light is etc as well as on sensor and and lens. It is remarkable how often in these forums someone does a restricted series of experiments and then claims a general rule.

 

[East Wind Photography]

I never did buy those as well. The atmospheric distortion can change from 1 second to the next affecting any perceived increase or decrease in resolution.

Best tests are definately terrestrial in a controlled environment.

There are certain situations where a crop delivers higher resolution than a FF with the same lens (even disregarding DoF, AF and the corners). Roger M. Clark has nice shots of the moon with the 5D2 and the 7D that show this:

www.clarkvision.com/articles/digital.sensor.performance.summary

 

[NormanBates]

Wow, you're a tough crowd to please!

I'm very happy with the results I got. Given how I use my camera, I don't need further testing.

 

[TrumpetPower!]

There are certain situations where a crop delivers higher resolution than a FF with the same lens (even disregarding DoF, AF and the corners). Roger M. Clark has nice shots of the moon with the 5D2 and the 7D that show this:

www.clarkvision.com/articles/digital.sensor.performance.summary

When I wrote, "All else being equal, the larger the format, the better the image quality," there was obviously a bit of shorthand going on. Particularly, to maintain the same field of view with a larger format, you need a longer focal length. Just as you wouldn't compare a P&S with an 8mm lens to a 5DIII with an 8mm lens, you really can't compare a 7D with a 300mm lens to a 5DIII with a 300mm lens -- even if you can physically mount the same lens to both cameras. Rather, you'd compare that P&S with an 8mm lens to a 5DIII with a 50mm lens, and you should compare the 7D with a 300mm lens to a 5DIII with a 500mm lens -- native, in all cases, not with extenders.

And you will, indeed, get substantially better images (with comparable framing) from a 5DIII with a 500mm lens than a 7D with a 300mm lens. That an uncropped image from a 7D with a 300mm lens is superior to a cropped image with a 300mm lens is no more surprising than that slicing away all of an 8x10 negative shot with a 300mm lens but for a 1" x 1.5" rectangle results in a worse image than you'd get from a 135 frame with a 35mm lens.

Once again, all else is clearly not always equal. If all you've got is a 300mm lens and you want to shoot the full disc of the moon, yes, of course, you're going to get better results from that lens with a 7D than a 5DIII. But the guy next to you with a 5DIII and a 1200mm f/5.6 is going to get an image of the moon that puts yours with your 7D to shame. And the guy the next hill over, the one with a 32" Dobsonian? Well....

Cheers,

b&

 

[Robert Welch]

I didn't look at the tests done, as I've been using cropped sensor cameras since 2003 and full frame (& APS-H) cameras since 2007 and can say that generally, the bigger the sensor the better the overall quality of the image. It's not to say that a given combination of APS-C camera with a given lens in a certain situation won't out perform a full frame with a particular lens in the same situation, but in more situations than not you'll probably get the best results from the larger sensor camera. That's why they are still making large sensor cameras, if that wasn't true then the 4/3rds format would have taken over the industry years ago.

 

[Robert Welch]

This reminds me a little of the comparisons between medium format and 35mm film. Back in the day I shot both, had a Rolleicord that may have had a focusing problem, always gave me just slightly soft images. I could get better images with my Canon AE-1 using Technical Pan film, ASA (same as ISO) 25, which had super tight grain that could be enlarged up to 16x20 without seeing the grain for the most part. Did that mean that 35mm film cameras were better than medium format? Of course not.

 

[Sporgon]

At Building Panoramics we recently purchased a 650D as a family camera for my partner. When it arrived we thought it would be interesting to test it against our 5D mk1 and mk2s. Our test was wholly empirical - we shot both cameras at ISO 100, from a sturdy tripod, and used a 50mm 1.4 on the 650 and an 85mm 1.8 on the MK2, both at f8 as this is the aperture we tend to use for buildings. We expected to see a difference.

The first two pictures attached are the whole images, reduced for the web. The third and fourth are the 100% crops.

We could not tell any difference, apart from the DoF. This surprised us as the mk2 was about £2000 when we bought it in 2008, and the 650 cost about a third of that.

So we took the 650D out on assignment with the FFs, and I can say that there is something about the images from the 650 that does not quite have the quality of the 5Ds. Just what this is is difficult to quantify: maybe it's the subtlety of the tonal graduation in difficult areas, I don't know. I don't think there is any difference in the actual sharpness.

Most tests are done at close distances, and as East Wind Photography has suggested, the goal posts move in practical applications where greater distances are involved.