I'm wondering if it's normal for a Canon 7D with its crop sensor to get unsharp pictures with a 70-200mm 2.8? I just bought a used 70-200 and it seems a little bit blurry. Maybe it needs calibration? IS seems to work fine, AF seems to be working fine in the viewfinder too. But when I open images up on computer, not as great as I expected.
-
\$\begingroup\$ How do these images compare to those taken with other lenses? (In general, this other question might help.) \$\endgroup\$– mattdmCommented Oct 19, 2012 at 17:32
-
1\$\begingroup\$ Do you mean Blurry as in motion blur (lens/camera movement, IE smear) or blurry as in out of focus? \$\endgroup\$– Digital LightcraftCommented Oct 19, 2012 at 20:34
-
1\$\begingroup\$ Following on to the above comments: please, if possible, provide some examples, ideally including relevant EXIF information (ISO, shutter speed, aperture, etc.). There are lots of possible reasons for blurriness but seeing images will narrow the actual reasons down very quickly. \$\endgroup\$– djangodudeCommented Oct 19, 2012 at 21:25
-
1\$\begingroup\$ When you mean blurry do you mean the picture is a bit soft? You might have some subtle shake, but we won't really know until we see some examples with a breakdown of the shot. \$\endgroup\$– Peng Tuck KwokCommented Nov 13, 2012 at 7:44
-
1\$\begingroup\$ It would be extremely helpful in finding out what's wrong if you would let us look at one of the blurry photo. \$\endgroup\$– GaptonCommented Nov 13, 2012 at 8:20
3 Answers
The Demanding Sensor
The Canon 7D is a very demanding camera, with a particularly demanding sensor. When compared on a megapixels basis, 18mp doesn't sound like much...the 1D X has 18mp, the 5D III has 22mp, the D800 has 36.3mp.
Megapixels and Line Pairs
A simple scalar number doesn't tell you the whole story though. The 7D's 18 megapixels differs a lot from, say, the 1D X's 18 megapixels...in terms of spatial resolution. The 7D's luminance spatial resolution is just shy of 116 lp/mm, where as the 1D X, despite having the same number of pixels, has a spatial resolution of 72 lp/mm. In terms of line pairs per millimeter (lp/mm), assuming the same focal length, same sufficiently wide aperture (say f/4), and subject distance...the 7D is capable of resolving details 61% smaller than the 1D X is capable of.
The smaller the pixel pitch, the more magnified the effects of camera shake, poor lens quality, or improper AF micro focus adjustment will be. The large pixels of the 1D X are more forgiving of optical aberrations and a small amount of camera shake, as the effects can be largely contained within a single pixel. With a 7D, you need to address IQ-diminishing factors more aggressively.
Fine lens...
Your 70-200mm f/2.8 IS lens, even if it is a first generation, is not necessarily "bad". Older lenses may not be quite up to snuff when it comes to the spatial resolution of what they project on the sensor. Lens designs that are decades old may need to be replaced with something much newer to really provide the 7D's sensor what it needs. Lenses released within the last five years or so should be fine in most cases, unless you really have a need to extract every last ounce of resolution your 7D is capable of. Any of Canon's Mark II L-series lenses are more than capable of resolving enough detail for a high density sensor. Canon Mark II telephoto lenses released in 2011 and 2012 (the 300mm & 400mm f/2.8 and 500mm & 600mm f/4 lenes) are some of the sharpest lenses on earth, and probably set to handle increasing sensor resolutions for the next decade or so (i.e. 30mp APS-C/80mp FF...possibly beyond?)
In the case of your lens, it probably just needs to be tuned for your particular 7D. Professional grade cameras usually include some kind of AFMA, or Autofocus Micro Adjustment feature. This allows you to calibrate your camera to account for any misalignment a given lens has relative to your exact copy of the body. The adjustment range is from -20 to +20, or a latitude of 40 steps. Outside of the very rare case where you actually do received a bad lens that needs professional tuning by a Canon service tech, a little AFMA tuning will correct any focus problems.
...so Self-Calibrate!
There are two tools that you can purchase to help you address AF micro adjustment needs. The simplest is Reikan FoCal. This is a software tool that runs on Windows (and soon MacOS, or so they say) that can automatically identify the necessary AFMA setting and configure the camera for you. I own this myself, and it is quite handy, although not 100% accurate every time it is run. It takes a little care to make sure the camera is in the right state before running an adjustment cycle, however with a properly printed test target the system works great. It is not a terribly expensive solution to the problem, and definitely better than a purely manual trial and error approach.
Alternatively, you can also try LensAlign. If you want to have total control and prefer to calibrate your lens and camera body manually, this is by far the best tool. LensAlign is a professionally designed and built focus tuning device that makes it pretty easy to identify how far off your AF may be. This is a more expensive solution, but if you absolutely require perfect focus, I don't think there is any better way to achieve it. Handily, despite being fairly expensive, it is available to rent at LensRentals.com for cheap.
With any AF tuning tool, you will need to make sure you put the AF/test target at an appropriate distance. For wide and normal lenses, that may only be twenty feet or so. For telephoto lenses, it can be much farther than that. You will also need adequate light to maximize contrast, which is essential for the AF system "seeing" the target correctly.
-
\$\begingroup\$ Kind of an older post but I am researching the same issue. The Autofocus Micro Adjustment described by jrista has instructions here from and for Canon. \$\endgroup\$– user26785Commented Mar 14, 2014 at 19:31
-
\$\begingroup\$ @jrista: Would it be better to use CDAF in Live View against a stationary target to test whether the lens indeed needs AFMA at all? From everything I've read on this site, CDAF is close to 100% accurate and doesn't suffer from front/back focus issues. \$\endgroup\$ Commented Mar 14, 2014 at 20:12
-
\$\begingroup\$ AFMA isn't about the lens, it's about the lens+camera combo, and it specifically adjusts PDAF behavior. Testing CDAF won't really tell you anything about whether you need to AFMA, because it's testing the wrong AF system. All that AFMA does is set a digital shift value that the cameras AF drive firmware uses to direct the lens when focusing. If you were off by -10 AFMA steps, when the camera tells the lens to focus, it tells it to focus by the amount the PDAF sensor detected, minus the AFMA adjustment. \$\endgroup\$– jristaCommented Mar 14, 2014 at 21:09
-
\$\begingroup\$ Yes, but it would tell you if the lens was faulty... \$\endgroup\$ Commented Mar 15, 2014 at 0:25
-
\$\begingroup\$ Maybe. What you see on live view is a form of JPEG encoding...and depending on how large the focus error might be, you may or may not be able to tell with just live view. I still think that using something like Reikan FoCal or LensAlign would still be best to determine if there is an actual error in the lens or if it just needed microadjustment. It is practically impossible to tell a difference, visually, with AFMA between 10-12 steps with live view. That's more than half the range of AFMA, but focus error is usually very, very slight. \$\endgroup\$– jristaCommented Mar 15, 2014 at 2:31
I have that combo, and it's really nice and sharp. No complaints.
First thing I'd do is adjust the micro-focus for the combo. It's possible it's enough off to be noticable.
other common things to check out. Make sure the AF is focusing on where you want it to be sharp, and not on something else. Try it out on a tripod if you haven't, take camera movement out of the equation (and use a cable release). Try it at F8 and see how it looks. Try it with a higher ISO and faster shutter to isolate camera shake out of the issue.
And swap around a couple of lenses and bodies. See if the problem moves with a specific piece of equipment. Is the lens soft across two bodies? Is the body soft with multiple lenses? If everything seems to work fine EXCEPT that lens on that body, that pushes the problem back towards the micro-focus.
One thing to consider if you are new to the 70-200 range is that shutters speeds need to be much higher to prevent motion blur than what you may be used to in zoom ranges. Motion blur seems especially pronounced on a physically longer and heavier lens like the 70-200 (compared to a 50-250 for example).
-
\$\begingroup\$ Lets assume a camera is as heavy as a dinning table, would you be able to shake it much? I doubt it. I just tend to think that heavier body and lens are less likely to be affected by motion blur, not more. Landscape photographer actually use heavy tripod and even attach weights to the tripod to help stabilize the camera. Long focal length, however, will make motion blur more pronounced. \$\endgroup\$– GaptonCommented Nov 13, 2012 at 16:26
-
\$\begingroup\$ Possibly so, but I'm only speaking from personal experience. I know I had a lot more camera shake than I expected when I first got my 70-200 (even with IS) and had to adjust my shutter speed accordingly. \$\endgroup\$ Commented Nov 13, 2012 at 16:33
-
\$\begingroup\$ A heavily unbalanced heavy dining table is prone to cause quite a mess. \$\endgroup\$ Commented Sep 28, 2019 at 20:36