Is it my lenses or my camera?

I just purchased a new 7D and decided to see how my lenses work on it.

Specifically, I mounted the following lenses on it and shot from wide open to completely stopped down. 100 f2.8 (non IS) macro, 24-105 f/4 L, and a 70-200 f/4L. The results were somewhat surprising and not a little disappointing. Without exception they all did great up to about f/11 and after that it all went downhill, fast. For some reason the photo's I shot with these lenses got progressively worse from f/11 - f/32 In fact none of the pic's would be usable beyond f/11. I had my camera mounted on a Gitzo 2531 tripod and a RSS ball head, both locked down tightly. I only extended the first (thickest) legs on the tripod.

So my question is, is it my camera that needs adjusting or is there something else going on? Any help would be greatly appreciated.

Thanks
Mike

Grab a beverage and a snack, Google the word "Diffraction" and do some reading about the physics of bending light rays in excess.

There are actually limits to the IQ curve and that curve is made steeper by high MP sensors in APS-C format.

Did you use a remote release cable, USB cable or the camera's self timer to take the picture? If not then the effect the vibrations from pressing the shutter will get worse as the apertures get smaller and the shutter speeds gets slower.

Also did you set the ISO to 100 or use auto ISO? With auto ISO, the camera would be raising the ISO as the shutter speeds were getting slower and apertures smaller. This impacts image quality not only from an increase in noise but in the rather ugly noise reduction that smears fine details that is applied to jpgs in camera, or even automatically by raw converters if you take their defaults.

Diffraction is real but a minor effect that only is usually visible at very large magnifications. The 7D makes at least as nice of a print for a given print size as any other camera with the same size sensor at f/32.

Im gonna say it is likely Diffraction as well. Beyond that there may be 'other' circumstances and if the OP is pixel peeping to determine the 'quality' then ...

" The 7D makes at least as nice of a print for a given print size as any other camera with the same size sensor at f/32."

...this might not make much of a difference.

GL OP!

On top of the obvious diffraction effect, the other thing to consider is that each step in a narrower aperture will lower the light level, meaning you will need to compensate by either slowing the shutter speed or raising the ISO to get the same exposure/brightness.

That's all well and good excepting for two considerations: raising the ISO means that you are amplifying a low-light image, and in doing that you are amplifying noise, which can be detrimental to your image quality. This combined with diffraction could in fact be, well, not good when you are viewing at a fairly large image size.

Slowing the shutter will let more light in so that you don't need to increase the ISO, but the problem there is your shot will be vulnerable to any vibration of the camera/lens, which will show up as image blur. The vibration could come from any of several sources -- you pushing the shutter button, the mirror "slapping up" to take the picture, a car driving by, a breeze, an unsteady tripod, or a lens on a tripod but with the IS on, acting "funny".

So, again, add any vibration blur to the diffraction effect, and you will get worse additive results.

So, yeah, diffraction alone begins to show up as you stop down, noticeable when you view an image at 100%, less noticeable at a more "normal" view size. But like you observed, on a 7D you will start to notice it at f/11 or even wider if you are doing real technical analysis.

So, in this testing you take that into consideration, and then realize that it is up to you to do anything and everything possible to cut down on camera shake. I consider a sturdy tripod and a cable or wireless/remote release to be "basic equipment" and then knowing how to use things like mirror lockup and, in the absence of a cable/remote release the self-time are critical skills to have not just for this type of testing, but for field photography where optimal sharpness can make a difference, such as landscape and other scenic photography, these skills and the needed accessories are also critical.

Thanks for the comments. I used the timer to trip the shutter and I also had the mirror locked up and 100 ISO and my exposures were all automatically adjusted as I was shooting in aperture priority. So it looks like that pretty much leaves it to diffraction. Thanks for the input, I really appreciate it.

Diffraction it is then. Consider though that viewing at 100% with a monitor with 100 pixels per inch is the equivalent of printing at about 52x35" and then viewing from very close. This usually isn't the final use of the photo.

If you resize the picture down to whatever the final size will be and then sharpen it, you should find that you can use the smaller apertures. It is just a matter of scale.

Well, good for you that you were using good technique! The only question is was your self-timer set to a short (2 second) time or to a longer (10 second) time? For critical stuff, unless you are using a cable/remote release, vibration from pushing the shutter button can be a much more widespread problem than the mirror slap (which is only noticeable for a relatively small duration withing, say, half a second.

But, yeah, I imagine that with the 18MP 7D files, viewed at 100%, diffraction could set you back! My three DSLRs are older, so have much lower pixel densities, meaning as "dramatic" of an effect when viewing at 100%. But still, I don't push my aperture. With my full-frame 5DC, I occasionally shoot with f/22, but not often. Sometimes with macro photography I'll push things with small subjects, just to get the focus/depth of field as "even" as possible, but still I like such images to have good, crisp, sharp detail, so I tend to not push to the limit of my macro lens, which is f/32!

I'll also toss in that such considerations that can affect image sharpness tend to mean more to me than they do to others because for some of the photography I do, such as macro and wildlife, often result in the need to closely crop an image because the subject was either too small or too distant to fully frame. In those cases things like diffraction and image noise, as well as "tack sharp" focus, can make a real difference as well as a "technical" difference.

And, I am also of the mindset that a "preferred" image needs to be good for printing/displaying "large", or else it takes a lower "rung" in how I would consider using it. Sure for the Web or printing up to, say, an 8x10, I can be happy keeping it, but ones that retain the qualities that would suit say a gallery print are ones that I want to keep a close eye on!

When I first got my 7D, I thought that it showed more diffraction, softness, image noise etc than my older camera at the same settings because I was viewing at 100% on both.

When I resized the 7D image down the the size of the other camera's full size and then viewed both at 100% so that they both have the same image size, then I found that the 7D actually had a bit less noise and showed a bit more detail regardless of the lens used or aperture setting. The 7D has such huge files that it is possible to zoom in so much more and see the effects of diffraction, lens aberrations, vibrations etc so much more even though they aren't worse as a percentage of the picture height.

I usually get prints made at 11x14 for matting and occasionally 12x18. Many of my files might not work well for a huge gallery print but for web use or small prints, I usually prefer the picture having deeper depth of field instead the slight increase in sharpness in the plane of best focus but starting to lose focus to the back or front if a larger aperture is chosen.

macroimage wrote in post #13123239
When I first got my 7D, I thought that it showed more diffraction, softness, image noise etc than my older camera at the same settings because I was viewing at 100% on both.

When I resized the 7D image down the the size of the other camera's full size and then viewed both at 100% so that they both have the same image size, then I found that the 7D actually had a bit less noise and showed a bit more detail regardless of the lens used or aperture setting. The 7D has such huge files that it is possible to zoom in so much more and see the effects of diffraction, lens aberrations, vibrations etc so much more even though they aren't worse as a percentage of the picture height.

I usually get prints made at 11x14 for matting and occasionally 12x18. Many of my files might not work well for a huge gallery print but for web use or small prints, I usually prefer the picture having deeper depth of field instead the slight increase in sharpness in the plane of best focus but starting to lose focus to the back or front if a larger aperture is chosen.

The diffraction effect is measured in um distance on the sensor, not in number of pixels on the sensor.

So a denser sensor will get more pixels capturing the soft transition than a low-density sensor. So it isn't hurtful to have a high-density sensor. It's just that it has a narrower range of apertures where it can take full advantage of the extra pixels.

This is a common misconception. People think the 7D is bad because it so quickly gets diffraction-limited, while totally missing that it will never be worse than any other camera. There is a range of apertures where the 7D is diffraction-limited but a camera with less pixel count isn't - but the 7D will capture more information. Then when reaching the aperture where the less dense camera gets limited, both cameras will follow each other - each being limited to the same amount of subject features being projected onto the sensor.

Great information everyone, thanks for the education.

I agree with tonylong's observation about macro and/or wildlife photography as well. I too tend to take a lot of wildlife and macro shots where cropping is often an issue.

That said, few if any of my shots will be printed larger than 11x14 so perhaps I'm worrying too much about something that might not even become an issue.

Thanks again.

macroimage wrote in post #13123239
When I first got my 7D, I thought that it showed more diffraction, softness, image noise etc than my older camera at the same settings because I was viewing at 100% on both.

When I resized the 7D image down the the size of the other camera's full size and then viewed both at 100% so that they both have the same image size, then I found that the 7D actually had a bit less noise and showed a bit more detail regardless of the lens used or aperture setting. The 7D has such huge files that it is possible to zoom in so much more and see the effects of diffraction, lens aberrations, vibrations etc so much more even though they aren't worse as a percentage of the picture height.

This is a very important point that, for some reason, most people refuse to understand.

When someone views a 7D image at 100% the magnification is greater than it is with a 40D at 100%, for example. It's that simple - at greater magnification you'll see more noise and less sharpness.

When the 50D first came out, with its 15 megapickles, lots of people were moaning that it was noisier and softer than the 10 MP 40D. It isn't. When a 50D image is resampled to the size of a 40D it's got less noise and more sharpness. Same for the 7D.

-js

A great summary and oh so true.

For all those saying diffraction, that doesn't sound right to me.

The OP said:

As I understand it diffraction doesn't kick in until f18 - f22 so how is it that all images above f11 are useless?
He should be able to take very usable images at f13, f14 and even f16, especially on a tripod with the mirror locked up etc.