In-Camera Focus Stacking

Plus, cropping versus smaller sensors and file size considerations
flower image taken with focus stacking technique

Super Sharp. Lepp used the Focus Bracketing function of the Canon EOS RP mirrorless camera to capture 80 focus-stacked images of a single primrose blossom, achieving extreme depth of field at 2x magnification. Canon EF 100mm f/2.8L Macro IS USM with and without a Canon Extender EF 2X III. Exposure: 1/125 sec., ƒ/8, ISO 400.

In-Camera Focus Stacking

Years ago, I was asked by the folks at Canon what features I’d like to see in future cameras, and one of my suggestions was an automated focus-stacking capability—dramatically increasing an image’s depth of field by combining multiple captures taken at different focus points. I’ve been using and writing about the manual technique, for both landscape and macro subjects as small as snowflakes, for several years. Now the industry has incorporated in-camera focus stacking features in a number of ingenious ways; it’s offered cameras including Panasonic LUMIX, Olympus OM-D, Nikon’s Z 6, Z 7 and D850, and Canon’s new entry-level, mirrorless full-frame EOS RP camera.

I got my hands on a Canon EOS RP and started experimenting with its In-Camera Focus Bracketing feature. The method incorporated into the EOS RP moves the focus elements inside the lens at incremental distances based on the focal length, aperture and camera settings. In the camera’s menu for focus stacking, the photographer can control three variables: Number of Shots, Focus Increment (1 to 10 scale, with larger numbers indicating larger focus increments) and Exposure Smoothing.

The number of shots and focus increment you choose will vary depending upon the lens being used, focal length, magnification and positioning of the depth of field within the image; you have to experiment with the many possible combinations. For example, adding depth of field to a landscape at 50mm and ƒ/11 might require three shots at the upper end of the increment scale, while at high magnification, where depth of field is extremely shallow, the photographer might select 40 or more shots at the lower end of the scale.

Exposure Smoothing, when enabled, automatically compensates for changes in image brightness from one capture to the next. Note that focus stacking requires a consistent frame and motionless subject; a tripod is needed. As with manual focus-stacking techniques, the in-camera capture sequence is later transferred to computer-based software for compositing into a single high-definition image.

In my initial test of the in-camera technique, I used the EOS RP to take macro shots of some flowering plants. I used a Canon EF 100mm f/2.8L USM macro lens set to just a bit less than 1x. At high magnification, depth of field is extremely shallow, so in the Focus Bracketing Menu, I selected 40 shots, with the focus increment set to number 4 on the scale. I disabled Exposure Smoothing as recommended in the manual for the Canon 100mm and 180mm macro lenses. I set the aperture to ƒ/8 because that is an optimum setting for sharpness with this macro lens, and I set the camera mode to Aperture Priority. Lighting came from a small LED light panel, as flash is not possible using this function.

A factor to consider when selecting the focus increment is that to achieve sharpness, the in-focus areas of the stacked images must overlap. If you choose too large an increment, the image may be softened. If you choose a smaller increment, however, be aware that an increased number of shots will be needed to cover the target area. It is my experience that you can never have too many shots in focus stacking, but you surely can have too few, which affects sharpness and possibly causes out-of-focus banding in an image.

The output images from the EOS RP can be in RAW or JPEG format; considering the number of images often required to complete the task, I usually work with large JPEG files when focus stacking.

Having completed the test at 1x with excellent results, I took it one step further and added a Canon EF 2x III extender to the 100mm macro for just a bit less than 2x magnification. I adjusted the number of images to 140 and the increment scale to 3, and the Focus Bracketing in the EOS RP still worked perfectly. Keep in mind that the depth of field at 2x is only about 0.4mm per image. The results were excellent and significantly easier and more precise to accomplish than with my earlier methods of incrementally moving the focus, the camera or the subject to capture the necessary range of images for maximum depth of field.

To expand my knowledge beyond Canon, I went to a friend who has both a Nikon Z 6 and Z 7, both cameras featuring what Nikon calls Focus Shift Shooting. The Nikon D850 also has this capability. We set up the Z 6 and accomplished the same series of images that I had done with the Canon EOS RP, also with great results. Nikon has added a few additional capabilities to its Focus Shift Shooting. You can use most any Nikon lens that has autofocus (Canon has a select group that work with Focus Bracketing), there is the possibility of changing the interval between shots in the series so flash can be used, and you can lock the exposure on the first image. But keep in mind that the Nikon Z 6 is almost twice the price of the Canon EOS RP, and the Z 7 is three times the cost.

With either brand, once the series is captured, the images must be composited in the computer with stacking software. For Canon users, Digital Photo Professional 4 will do the compositing (this software comes with the camera at no additional cost), while photographers who have already been working with this technique may prefer specialty software such as Photoshop CS, Zerene Stacker and HeliconFocus. I would recommend the latter two for the best result.

The beauty of this camera function is that it offers precise focus-stacking capabilities right in the camera for landscapes and macro, even up to 2x. There’s no moving the camera, manually changing focus, or repositioning the subject. It’s ready and able all the time. And regardless of your preferences for manufacturer or post-capture assembly, the resulting images will be sharper, have the precise amount and positioning of depth of field you want, and will greatly increase your technical and creative options.

Crop Now Or Later?

I have two cameras, a full frame and a crop sensor (APS-C). The crop sensor will give me a magnified view (1.6x) but contains less digital information. If images were taken with the two cameras using the same lens (my 400mm), and I crop the full-frame image to match the image that I attained with the smaller sensor, which image would be sharper?

–J. Terkel
Via the Internet

This is an interesting question that goes to the heart of choosing and using digital cameras optimally in various photographic situations. But first, let’s look at all the variables that can affect the outcomes, beginning with differences in pixel counts. Among the current models, most cropped sensors (APS-C size) are around 20 megapixels, and the full-frame could range from 20 to 30MP or more.

Assuming that the APS-C sensor (with a 1.5x to 1.6x crop factor) has 20MP resolution and the full-frame sensor has 30 MP, when the resulting full-frame image is cropped to match that obtained from the APS-C sensor, the two files will contain approximately the same amount of data and comparable sharpness. If the full-frame camera has a higher resolution, such as 46 or even 50MP, then the cropped image from the full-frame camera will contain more detail—an important advantage when the files are enlarged.

If the full-frame camera has a lower resolution, similar to that of the APS-C camera being compared, the full-frame camera may still have the advantage because its pixels are larger, offering better low-light capability and less noise at higher ISOs. Considering these factors, it’s likely that the full-frame crop will still be as good, or superior to, the smaller sensor’s file.

Of course, when it comes to getting the shot, the APS-C camera will likely have a few advantages of its own. Because the camera is processing smaller files, it may capture more frames per second. Some have great auto-focus capabilities, and the bodies may be smaller in size, and thus more easily and quickly positioned.

So how do you choose the best option? For me, the decision is always made on image quality. For better quality at higher ISOs, whether needed for faster shutter speeds, more depth of field (a smaller aperture) or low-light conditions, the full-frame camera usually wins. But read on for another perspective.

How Big An Image File Do You Really Need?

Is a larger image file worth the cost of large sensors, higher resolution, extra sensitivity and lots of technical calculations? The most practical (if not most aesthetic) answer is, it depends on how the images will be used.

Grub Grab. Lepp captured this image of a bald eagle feeding its chick on 4K video from a distance of more than 200 feet. The photograph is a frame grab from the video clip, with a file size equivalent to that produced by a 6MP camera (18 MB). The clarity is excellent, sufficient for publication, social media or a print up to 13×19 inches. Canon EOS R, Canon EF 600mm f/4L IS III USM, Canon Extender EF 2X III. Exposure: 1/350 sec., ƒ/16, ISO 1600.

As faithful readers know, I’ve been working for years on a project to document a bald eagle nesting site not far from where I live in Central Oregon. I pride myself on being able to accomplish full-frame images of the nesting activities from a distance of more than 200 feet. This means I use long lenses (500mm, 600mm and even 800mm) in conjunction with 1.4x and 2x extenders. Not many folks have access to, or are willing to haul around, this kind of photographic gear.

Yet my photographer and bird-watching friends come out to photograph the nest with 100-400mm lenses (sometimes with a 1.4x attached) and send me excellent close-up images that rival my megabuck, mega-millimeter images. It’s because sharing images through email or social media requires only a very small file, which allows the photographer to crop extensively…if the original capture is well-exposed and sharp.

So, depending upon your ultimate use, you may have a lot of leeway in cropping beyond the actual reach of your equipment. Don’t be afraid to reduce the image size to enlarge the subject or improve composition for web-based sharing. What if you have a home printer or send your images to a professional printer for enlargements? You will be surprised at what is possible even in these situations.

How small can you go? I enjoy demonstrating the capability of using a program such as Photoshop or Lightroom to extract single frames from 4K video clips, then creating beautiful 13×19-inch prints from the frame grab. Those individual frames are equivalent in file size to a single photograph captured with an 8MP camera. While we may insist upon 20 to 50MP frames from our DSLRs, the frame-grab exercise demonstrates that you can crop a 20MP camera image to 50 percent of the original file with great results for display as moderate enlargements and on social media.

So why do I employ the large full-frame cameras and monster lenses? It’s about big prints: In order to make really impressive prints for display, I still need sharp, well-exposed, large files, and while I can rarely predict the future use of any capture, I never want to limit my options.

One of North America’s best-known contemporary outdoor and nature photographers and a leader in the field of digital imaging and photographic education, Lepp is the author of many books and the field editor of Outdoor Photographer magazine. One of Canon’s original Explorers of Light, Lepp finds inspiration in advancing technology that fuels creative innovation and expression of his life-long fascination with the natural world.