AF Microadjust for Sharper Images
By John Gerlach
Be sure to subscribe to my website to be automatically notified of new articles! I have a ton of new photo tips to share!
Update: I continue to AF microadjust all of my Canon cameras with the Canon lenses I use on them for autofocusing. What do I mean by that? Many of my Canon lenses I never use autofocus with, such as my 16-35mm and 24-70mm. I use them mostly for landscapes, use a tripod, and manually focus using a magnified live view image to see where the focus is best. I also focus stack a lot, so again I find manual focus crucial to find my starting point. I sometimes do use autofocus with these lenses, but I am using autofocus in the live view and that circumvents the AF microadjustment problem as focus is done on the sensor plane, not someplace else.
This is the menu to get to AF Microadjustment in my Canon cameras.
This choice lets you adjust all lenses by the same amount or individually. Since all my lenses require a different adjustment, I see no reason to adjust by the same amount.
When AF microadjusting a zoom lens, you can do it for both the short focal length and the long one on the zoom, but I only do it for the long one. Normally, when using shorter lenses, I focus manually use a magnified live view image, or focus automatically using the AF in the live view by touching the LCD. No Af microadjusting appears to be necessary when autofocusing in live view.
This is my AF microadjusting target. I set the camera and lens locked on a tripod head so a single AF point is on the hair of the left side of the head, and check that for sharpness. Remember I now use a flash to light the target - no matter how bright the ambient light - to use short flash durations to eliminate vibration from other sources. The test is to see how well the camera's autofocus hits sharp focus. None of my lenses have been perfect, but most fall in the range of plus or minus 5. If more than that, you want to know that and set the AF microadjustment to make the lens focus either a littler closer to you or further away.
I made one major change when I run the test now. Hummingbird photography requires super exact focus to at least have a sharp image of the hummingbird's face. Having your autofocus off at such a magnification seriously hurts the sharpness. Therefore, when I was in British Columbia photographing hummingbirds at my workshops, I ran the AF microadjustment test at about six feet - the distance I am photographing the hummingbirds at. I know "sources" say you should do this AF test at 25 to 50 times the focal length, but as I do so often, I ignore the prevailing advice and do things that make more sense to me so I ran the test at 6 feet since that is how far away my subject will be. Normally, subjects are at various distances and running the test at a greater distance makes since, but not when I might take 10,000 hummingbird photos at six feet. I was teaching my workshop clients how to run the test to help them get sharper images. On one week, the weather was challenging - sun, clouds, wind, no wind, more sun, and then clouds. How to run the test?
The AF microadjustment test should only have one variable - how accurate does the camera and lens work together to produce sharp focus. Wind shaking the camera a little or not at all should not affect the results. Same for mirror slap and other factors that cause an image to be less sharp. While I normally run the test in bright sun to achieve super fast shutter speeds at ISO 200 that eliminates various sources of camera vibrations, I could not that day. So............to teach the AF microadjustment method I use and to eliminate variables, I did it indoors out of the rain and wind, and lit my focus target with a single flash set to 1/32 power. The extremely short flash duration at 1/32 power eliminates vibration problems entirely! Doing this made the results more consistent by a wide margin. I will always light my focus target with flash for now on. Oh, I did use two LED lamps to light the focus target enough for my autofocus mechanism in the camera to lock on to the target, but the flash is much brighter and overpowers the relatively dim LEDs.
AF microadjusting helps make the focus on this male rufous hummingbird's head perfectly accurate.
This is all natural light with a Canon 600mm III lens and a Canon 1.4x teleconverter. My AF microadjustment for the 600mm lens by itself is only -1 (that isn't much), but with the converter it is +5 and that is significant. What to the values mean. With a -1 settings, the camera now focuses a tiny bit closer to the camera, and a +5 makes the lens focus a little further away than the default setting of 0. Negative adjustments correct backfocusing and positive adjustments correct frontfocusing. This is the "famous" calliope hummingbird that two hummingbird workshop groups photographed. He owned this twig, so they shot at least 50,000 shots of him - super adorable he was.
Male rufous hummingbird at my natural light station during the workshop. He would perch on anything we put out for him. Thanks to Scott and Melanie for letting us all know about him.
This is huge! Every Canon lens/camera combo that I use produces sharper images for me when they are AF microadjusted. If you think your gear doesn't need it, most likely it does and you are not as critical as I am. Not every camera lets you do this adjustment, but if yours does, take advantage of it.
It was two winters ago when I tried to photograph the chickadees in my yard with my new Canon 100-400mm lens. I had the lens zoomed out to 400mm, used a high ISO to get about 1/500 second for shutter speed, had image-stabilization on, and shot handheld as I had no other choice - chickadees move too fast. I noticed my sharpness on the eyes were not as good as I had hoped. Did I get a bad lens? Then I ran my first AF microadjustment using the Canon 1DX Mark II and the lens at 400mm, and at the closest focus distance as that is where I was photographing the chickadees. The result - I needed to use an AF microadjustment setting of -9 to achieve sharp focus on the eye of the bird. I have had no problem hitting sharp focus with this camera/lens combo since.
Beginning in 2007, Canon began to offer a focus adjustment in their higher-end cameras to allow better control over the autofocus. It is called AF Microadjustment. What does this mean? First, the autofocus in your camera is made to strict specifications and does a remarkable job. But, manufacturing tolerances are allowed. Sometimes, the focus isn’t quite as precise as it could be for a camera/lens combination. Assume the autofocus is reasonably good to begin with, but if you notice the focus is a little in front of the spot where focus should be right on, or a little behind that spot, and either problem is consistent, then microadjusting the autofocus will improve sharpness overall. I find that all my telephoto lenses with a specific camera benefit to some degree from microadjusting the autofocus. Currently, I am using two cameras - the Canon 5D Mark IV and the 1DX Mark II. I must individually microadjust focus for each lens and camera combination I use with autofocus to achieve sharply focused images consistently.
I held a single active AF point on the head of this wood duck in Ohio and shot truly sharp images most of the time. Obviously, when birds move fast, or erratically, I do miss focus. My Canon 800mm lens needs a -3 AF microadjustment to achieve sharp focus on my Canon 5D Mark IV.
Since I teach many photography workshops where I show clients how to run an AF microadjustment test while using their gear, I seldom find a camera/lens combo that does not benefit from setting a certain microadjustment. In other words, if you think you don’t need it, you probably aren’t checking closely enough. Though, I am sure it happens, not once have I found a camera and lens to focus precisely on the default values set in the camera.
What does Microadjusting do?
The factory sets the camera to focus on the target. But, sometimes the focus is a little past the focus target or too close, instead of right on the spot that should have sharpest focus, like a bird’s face. If the focus is slightly behind the target, then the camera is said to be back focusing. Conversely, if focus is slightly in front of the target, then the camera and lens together are front focusing. Microadjusting the camera and lens combo enables the lens to focus on the target consistently most of the time. Bear in mind that autofocus does vary a little from shot to shot, even when all conditions remain the same. You could lock your camera and lens up tightly on a tripod, not vary the composition at all, use superb shooting technique with high shutter speed, and still the focus might vary a little in a set of otherwise identical images. Therefore, to run the tests, I always shoot at least three images at each adjustment value and hope a majority (two of them) are focused where they should be.
This is squaw root in the Great Smoky Mountains. As for focus, I never use any autofocus for close-ups. For objects that hold still, I am way better off to manually focus the lens using the magnified live image. It is also important to shoot with live view active to eliminate both mirror slap and shutter shock!
Because autofocus does vary a little, even with the best technique, I now always use manual focus with a magnified live view on my camera’s LCD whenever possible. I find my focus is more likely to be right on using manual focus with a magnified live view image. But, of course, photographing wildlife and even landscapes from a moving boat generally work best with autofocus. In that case, I focus on the exact spot where I want the sharpest focus to be, shoot several images, then quickly refocus on something at a different distance, and then return to the subject to make the camera focus the lens on the target again and shoot more images. Focusing on the same spot more than once increases the odds that some images will be sharply focused where they should be.
Here’s how I run my test!
Make sure the exposure is ideal. It is hard to check for sharpness if the image is overexposed.
Turn image-stabilization off
You don’t want the IS to activate on a tripod and causes image softness because the image-stabilization system is active. IS generally works best if you are shooting in the wind on a tripod or shooting hand-held where real camera vibration is likely to be the case.
Use the maximum aperture (smallest f/stop value)
I was so pleased to see how sharp this gray squirrel turned out when I used my Canon 5D Mark IV with the 800mm lens set to -3 AF microadjustment.
Even though you may wish to stop down more, and f/8 is likely a sharper aperture than wide open, say f/4, for test purposes you want the shallowest depth of field. This makes it easier to see the change in sharpness from one image to the next. Stopping down increases the depth of field that obscures the small autofocus errors that you are trying to detect, so shoot at the biggest aperture for test purposes.
Single Servo AF
For test purposes, do not use continuous AF. You want the autofocus to lock in on the target and not hunt for sharpest focus after initially finding focus.
Use a solid tripod with head to lock your camera and lens completely still during the test.
Run the test outdoors in bright light, but avoid breezy conditions.
Bright light allows a fast shutter speed to minimize any camera vibration that might be occurring due to the breeze. It also helps the autofocus system “see” better. Indeed, I am considering running the test using flash to light the test target to use the short flash duration to eliminate “camera shake” vibration factors. Remember, this test is to determine how well the autofocus focuses on the target. It is important to eliminate all other factors as much as possible that can contribute to a loss of sharpness.
Use a perfectly flat target with lots of detail
The camera’s autofocus works best if the target has fine detail with contrast. I normally use a $100 bill because I find the sketch of Ben Franklin’s face is an easy target to evaluate. The fine markings in his hair on the left side of his face work well for detecting image sharpness. If you just paid your bills, you can get by with a $1 bill as George Washington’s hair has fine detail too.
The camera’s sensor must be perfectly parallel to the flat target. That means the camera should be at the same height as the bill, so the focus nicely covers the surface of the flat bill. I attach the $100-dollar bill with two large paper clips to flat cardboard and stand that up against a wall.
Use a Higher ISO, such as ISO 400 to use a faster shutter speed – typically I use 1/1000 second or even faster if the light is bright enough.
The target should be about 25 to 50 times the focal length of the lens.
A single active AF point on the eye of this singing Northern mockingbird in Arizona with a microadjusted lens did the trick for sharpness. Naturally, I had the lens firmly attached to a Wimberley gimbal head.
I always adjust by lens as every one of my Canon lenses has required a different AF Microadjustment value. In my case, all have been a negative value that varied from -3 to -9. That means each lens back-focused a little - focused too far away. The difference between -9 and 0 isn't really that much, but it does make a difference in how sharp your images will be. If your lens need -5 adjustment, then the default zero will definitely be a little less sharp if you check closely. I enlarge Jefferson's hair on the left side to 100%, and it is rather easy to tell the difference.
With Canon cameras, notice you can AF Microadjust a zoom lens to both the short and long focal lengths on the lens.
Use a two-second delay to fire the camera or a remote release.
Shooting the Test
Find the Micro adjustment controls in your camera’s menu. On my Canon 5D Mark IV, the control is Under AF and tab #5. Look at the bottom of the list to find AF Microadjustment. Naturally, the location will vary with different camera models. And another camera brand may call this something different. I know Nikon calls this AF Fine-tune.
My camera gives me two options. These include:
All by same amount
Adjust by lens
I already know that my Canon lenses require different adjustments, so I use Adjust by lens. And with a zoom lens, two settings can be adjusted. One is for the shorter focal length on the zoom lens, and the other for the telephoto range. My camera can remember the focus adjustment I have set for up to twenty Canon lenses! That is handy.
I align the camera’s sensor perfectly parallel to the flat $100 bill target – at least as close as possible. Then I shoot a series of images to find out if the camera/lens combo is front or back focusing. Let’s adjust my 800mm lens for the Canon 5D Mark IV. My adjustment range provided is +/- 20 increments in focus distance. These increments are tiny, and if the autofocus requires more adjustment than this, the manufacturer must adjust the cameras autofocus. But, so far, I have never had to do that.
I shoot three images at each of these adjustment settings: -20, -15, -10, -5, 0, 5, 10, 15, 20. As there are nine settings, that means a total of 27 images. For each image, I begin out of focus, press the shutter button to initiate ONE Shot autofocus, and then press the shutter button all the way down to start the two-second self-timer, and immediately remove my hands from the camera to avoid vibrations. And don’t touch the tripod legs either. (I see a lot of folks do this in workshops.) After two seconds, the camera fires the shot.
My focus was right between the eyes of this barred owl in TN. I feel so much more confident about achieving sharp focus by using excellent technique and having my camera/lens AF microadjusted!!!!!! It is important!!!!
I start the procedure at -20.For the second and third shot at – 20, I manually turn the focus ring until the lens is significantly out of focus, then press the shutter button again half way to make the camera refocus on the target and shoot another image.It is important to always start the AF with the target not in focus at all.
I follow this procedure for all nine focus adjustments listed above, and then view my test images on my computer using Canon’s DPP4 at 100%. Of course, you can use whatever software you normally use. Just view the images at 100%. Don’t go more than 100%. If you do, all images may begin to look too soft at extremely high magnification greater than 100%.
Hopefully, image sharpness will clearly be better at one of these settings. I ran my test to find +5 and 0 were the best. The larger AF adjustments, both positive and negative were clearly less sharp. That is good news as I know the Canon 5D Mark IV with my Canon 800mm f/5.6 lens is close to being right on with the factory default at zero.
I ran a second test to fine-tune my AF. I shot another series of three images each at -6, -5, -4, -3, -2, -1, 0, +1, and +2. Then I looked at the images on my computer monitor with Canon DPP4 and used 100% magnification. All of them looked quite similar in sharpness, but by carefully observing Benjamin Franklin’s hair on the left side of his head on the $100 bill, I marked the images that were clearly less sharp that others. Eventually I cut the images down to eleven. Then using the Info button in Canon’s DPP4 software, I looked at the EXIF data for each image that survived my cut hoping to find an AF microadjustment value that dominates. Here are the values I got from the eleven remaining images: -6, -4, -4, -3, -3, -2, -2, -2, -1, 0, and +2. Remember that autofocus does vary a little, so that accounts for -6 and +2 to be sharp in one image. Notice the trend, though. Most of my surviving images (7 total) fell in the range of -4 to -2. So, I set my AF microadjustment to -3, went outside and photographed some pine grosbeaks and got excellent sharpness.
A grizzly bear in West Yellowstone at the Discovery Center. (captive naturally)
From these results, I know two things. First, the camera and lens do need microadjustment to give the sharpest possible image. Second, because the AF microadjustment value selected is -3, I know the lens is back focusing a little. That means the focus is a little beyond where it should be. The -3 setting makes the camera focus slightly closer to me. This is a tiny correction, but it does improve my image sharpness! In another test with my Canon 5D Mark IV and Canon 100-400mm lens, I needed a -9 AF microadjustment to achieve sharp focus. No wonder all my autofocus images were not as sharp as expected when I first tried it without doing the AF microadjustment.
I realize many photographers AF microadjust their camera/lens combo by using tools made for the purpose, such as the highly regarded Lens Align focusing tool. I own this tool, and it does work well, and helps you to know if your camera and lens is back focusing or front focusing, but I can easily do that myself without carrying more stuff with me. What if I need to microadjust my camera when I am on the road? I really don’t want to carry anything else when I lead my safaris to Kenya or return to Iceland. By doing it the way I do, I can always set up a flat target with detail, and microadjust my gear quite successfully, and so can you. Many photographers swear by commercial AF microadjustment tools, but then I notice they sell the tools!
Although I have not seen it yet, I hear that the microadjustment settings for a particular camera/lens combo can change over time, so watch out for that. And I suspect, it is especially important to retest the equipment when shooting temperatures are drastically different.
What Lenses to AF Microadjust
Keep in mind that any AF microadjustment setting is good for the camera and lens tested. You cannot assume that adjustment is valid for any other camera or lens, even if the camera or lens is the same model. Think in terms of camera/lens combinations! And many lenses I don’t need to AF microadjust because I never use them with autofocus. When I am making landscape images, for example, with my 16-35mm or 24-70mm lenses, I always focus them manually while using the magnified live view image. Remember, AF microadjustment only matters if you are using autofocus, not manual focus!
Since I am not using autofocus to sharply focus on the trees emerging from the fog in Smoky Mountain National Park, I do not have to use a microadjusted camera/lens combo. Instead, I manually focused on the trees using a magnified live view image.
Good luck and may the sharpness of your images improve!!
Question: I have had a few questions about how far the camera should be from the focus target. I don't have the definitive answers to that - not sure there is one - but I know Canon USA suggests using fifty times the focal length. So I tend to do that except in a case where I know I will be shooting at near the closest focus distance the lens offers. For example, when photographing chickadees at six feet, I noticed my Canon 100-400mm lens wasn't hitting sharp focus with the factory set default, so I tested the autofocus at six feet - my intended shooting distance and discovered my Canon 100-400mm lens required a -9 AF microadjustment as it was back focusing a little. That does not prove that if I was photographing objects much further away that a -9 is ideal, so I have to test for longer distances. And by the way, I did test autofocus in live view and found no AF microadjustment is necessary with the two lenses I tried it with. That isn't surprising as AF is determined right on the sensor.
Let's use 50x the focal length as a starting point, as I see that figured published quite often. How far is that? Use this chart, and remember, this distance is the distance between the camera and the focus target.
100mm = 16.4 feet
200mm = 32.8 feet
300mm = 49.2 feet
400mm = 65.6 feet
500mm = 82 feet
600mm = 98.4 feet
800mm = 131 feet
And one more observation. I have noticed on photography sites that many discount the need to AF micro adjust their lenses. Some say none of their Canon lenses, for example, need AF microadjustment as they are all sharp. And in my case, all of my Canon lenses benefit from it. Who is right? I suppose it is possible in theory to have lenses that don't need it, but that is highly unlikely. What happens is the less critical photographers among us shoot and think there images are sharp. And indeed, they do look reasonably sharp, so they think their lens is fine. But, if you test a lens properly, you will find there is sharp, and then REALLY SHARP. Unless you test the autofocus and compare images shot with and without AF micro adjustment, you won't see what sharp really looks like! And to be honest, the sharpness difference between shooting with it AF microadjusted isn't great, as opposed to the default AF setting, but it is there. And if you have other flaws in your shooting technique - use a filter when not needed, shoot handheld, don't keep the shutter speed up, no lens hood, tripod not stable enough, banging the shutter button with your finger - all these factors that cause a slight loss of sharpness add up to a serious sharpness loss. So, why not eliminate a slight loss of sharpness wherever it might exist?
And some say it isn't that necessary to have your lens focus precisely as depth of field by stopping down takes care of the problem. And to a large extent, especially in landscape photography, where you might stop down to f/11 or more, that is true. But, in wildlife photography often we must shoot wide open or nearly so in the range of f/4 to f/5.6. Sure, we would like to stop down more, but don't due to the long lenses being used, animal motion, low light, and trying to avoid too high of an ISO. Therefore, we have little choice but to use shallow depth of field to keep the shutter speed up, and that is where AF microadjustment is absolutely crucial for razor-sharp images. Here are some ducks shot around f/5.6 with my Canon 800mm lens that were on the zoo pond in Cincinnati, OH. There weren't a lot of birds there, and even less ambient light on the cloudy day I was there, but plenty enough to have fun with. And one fellow said the pond is full of waterfowl in late March and early April when they are migrating. I haven't been there yet, but I will!
Male Hooded Merganser
Female Hooded Merganser
Hen Hooded Merganser
Male Hooded Merganser - my AF point is right on the eye
Male Hooded Merganser
Male Hooded Merganser
Male Wood Duck