When I transferred this web page from WordPress to Quarto I produced new versions of the photographs from the same RAW files of pasta. In November 2024 I expanded and edited this page.
Clicking on photographs opens a gallery view with them displayed at a larger size and at higher resolution. This view also allows easy navigation among multiple photographs from a given Figure.
Definition and explanation
I by explaining some terms that I will use in this page. Bracketing consists in acquiring a series of images with different camera settings. The word bracketing comes from the idea that we have a target value for the setting, say exposure, and we acquire images with this exact target setting and settings both at slightly large and slightly smaller values (bracketing both “sides” of the target). At least three images need to be acquired for the meaning to strictly apply, but in some cases even hundreds images are acquired.
Bracketing is an old term and an old idea, which of course can be used also with film cameras. Some automatic film cameras and advanced digital cameras can automate bracketing, at least for some parameters. Bracketing can be also achieved by manually changing the settings through the acquisition of a series of images. Exposure and focus distance are the two parameters for which bracketing is most frequently used. In the case of exposure, bracketing can be achieved by bracketing (diaphragm) aperture, shutter speed, or ISO setting.
The most basic approach is to just select the best image from the bracketed series, and use it discarding all others. What I also discuss below is the merging of a series bracketed images into a single composite image which is “better” than any of the individual images. In the nine years since I wrote the first version of this page the automation or bracketing, and more recently in-camera merging of bracketed images has improved greatly. One factor has been in mirrorless cameras the availability of an electronic shutter and improved sensor read-out speed. The much faster speed with which images are acquired (up to 50 frames per second in the Olympus OM-1 camera) combined with improved image stabilisation has made the acquisition of bracketed images for merging possible with a handheld camera. See my Notes on the OM-1 (digital) camera for an example of the current state of technology. Here, I will give an introduction to the bracketing and merging without focusing too much on the specific camera I use.
Exposure bracketing
In Figure 1 bracketing (AE BKT) was done using 2 EV steps to make the difference clearly visible. However, when bracketing is done to later select the best exposed photograph out of a bracketed series, smaller steps even as small as 1/3 EV are used. However, when the aim is to merge the frames into a single photograph with high dynamic range (HDR), three to seven exposures with 2 EV or 3 EV steps are the norm.
What setting is used to adjust the bracketed exposure (AE BKT) depends on which exposure mode the camera is set to. If the camera settings is aperture priority (A) it brackets shutter speed, in shutter priority (S) it brackets aperture, in program (P) it brackets both, and in manual (M) it brackets ISO. This is different from ISO bracketing which varies ISO keeping the exposure value constant.
Exposure bracketing followed by “HDR” image merging can be used to increase the dynamic range, i.e., increase the range between the brightest and darkest luminosity values in the scene that can be recorded reliably in a photograph. In the example in Figure 2 the scene was dark and I had to use a rather small aperture (f:5.6) to ensure that both the distant landscape and the room would remain in focus. The luminosity range was broad but not extreme, so to avoid use of an extremely high ISO value I bracketed by only \(\pm 1\) EV. The three images without edits clearly differ in the bright sky outdoors and darkest shade inside the room.
The raw files as saved by the camera when displayed as is do not reveal all the stored information. By editing a file this additional image information can be made visible. In Figure 3 A: there is no use of image merging, the RAW file of the image with default exposure was edited, brightening the shadows and darkening the sky. In B: the three images have been merged into an HDR photograph. Now the default luminosity adjustment creates an artificial-looking image with very little contrast. In C: the merged HDR image has been edited to obtain a more natural tone gradation. In this case the gain from image merging was rather small. In cases with higher differences in illumination the gain is more.
Even though modern mirrorless cameras use very good auto-exposure algorithms and the inherent dynamic range of sensors is frequently more than enough, there are situations where the dynamic range of the sensor introduces a constraint. In the less extreme cases, we need accurate exposure tailored to our interpretation of what is most important in the subject. In this case exposure bracketing can work either as a safety net or as a way of delaying the choice of settings to the editing stage. In more extreme cases, the difference in luminosity between the brightest and darkest parts of a scene is so large that one or the other falls outside the sensitivity range of the sensor irrespective of the exposure setting. In such a case, a suitable approach is to merge the bracketed images so that darkest and brightest parts of the scene come from images suitably exposed to record them. For merging to HDR or high dynamic range images usually 3 to 7 frames are more than enough.
Some cameras can merge the images immediately after they have been acquired and most advanced photo-editors are also capable of aligning and merging the images. Because of the alignment step, in some cases the algorithms tolerate slight movement of the camera position between the successive exposure bracketed images being merged.
Focus bracketing
The change in the focus plane is more difficult to see than for exposure above, as the subject has a depth of less than one centimetre. By using a rather large aperture and barely overlapping focus steps the effect can be seen (Figure 4). For focus stacking the steps have to be small enough to ensure partial overlap between the in-focus regions of successive bracketed images. Improvements in autofocus performance combined with subject recognition has made focus bracketing with the aim of choosing the sharpest photograph less common than in the past, possibly except in macro photography. On the other hand, merging a series of focus-bracketed images into a single photograph with a deeper in-focus region has become more common than in the past. Focus stacking is computationally expensive and only rarely available in-camera. Specialised software such as Helicon Focus is usually needed, especially for deep stacks.
There are two different approaches to focus bracketing: 1) changing the focus setting of the lens or 2) changing the camera to subject distance. For distant subjects only the first approach can be used. However, for close-up and macro photography both approaches are in use. The higher the magnification and smaller the subject the more difficult it becomes to bracket by manually moving the focusing ring of the lens. In the case of low-magnification close up photography displacing the camera using a manual focusing rail is slow but feasible. The best approaches for high magnification close-up and macro photography are to use either a motorised focusing rail or a lens with motor-driven focusing (i.e., an auto-focus lens) controlled remotely. In my experience, best of all approaches is automatic focus bracketing built into a camera. The main advantage is speed, as a series of hundreds of images can at it fastest be acquired in a few seconds. Speed not only saves our time, it helps avoid changes in the subject between frames. This speed combined with image stabilization makes it possible in some the newest mirrorless cameras to acquire series of focus-bracketed images with the camera handheld and have them merged immediately in camera for their inspection. This can be made to work both at low and relatively high magnification in the field (Figure 5).
I use focus bracketing frequently, both outdoors and indoors. I rely on the camera, even though have access to a motorised macro focusing rail (Stackshot from Cognisys) that I have not used it in a long time. I regularly use a manual focusing rail (Sunwayfoto MFR-150 Macro Focusing Rail) to adjust the position of the camera and changing magnification, but use the camera controls for automatic bracketing. A motorised rail is necessary if one uses a manual focus lens, or a microscope objective adapted as a camera lens. Good motorised macro focusing rails like Stackshot have a minimum movement step on the order of \(2\,\mathrm{\mu m}\).
Focus bracketing has been traditionally used for static subjects with a camera on a tripod or copy stand. Nowadays, it has become usable also with the camera handheld. It is useful when the depth of field is too shallow to include the whole subject, which is a very frequent difficulty in close-up and macro photography (Figure 6).
To some extend closing the diaphragm helps, but diffraction puts a limit as it hampers resolution. Closing the diaphragm may require the use of a higher ISO setting which if taken too far compromises the quality of the photographs. Nowadays, I tend to use focus stacking very frequently for close-up and macro photography, except for completely flat subjects. One thing to keep in mind is that when merging bracketed images in post we can chose which section of the bracketed series to merge, allowing a better control and a sharper transition between in-focus and out-of-focus areas. Figure 7 gives two examples, in each pair, one photograph is a single image from the focus-bracketed series and one image is the merged focus-stack.
I have described the use of focus bracketing and other settings in macro photography with the Olympus OM-1 camera and M,Zuiko lenses in the page The nitty-gritty details of macro photography.
Bracketing of other settings
I normally use raw files (.ORF in Olympus and OM System cameras) and find frequent use for exposure and focus bracketing. Sensor sensitivity (ISO BKT) bracketing can be, at least in principle, useful also when saving to raw files, as a compromise is involved between avoiding excessive “image noise” and using a fast enough shutter speed to stop movement or small enough aperture to have a deep enough in-focus region.
Some cameras have additional settings for which bracketing is possible, but these are mostly relevant to cases when the raw data conversion into JPEG compressed files takes place in camera. These include bracketing of white balance (WB BKT) and of the “art mode” (ART BKT).
In the OM-1 camera the bracketing of different settings cannot be used simultaneously or combined. A more detailed account is available in the page Notes on the OM-1 (digital) camera at this site.
There are situations where merging of multiple images taken with identical settings, i.e., without bracketing, can result in an improved final image. A typical case is when a very high ISO setting is used, as averaging multiple image frames is very effective in controlling “sensor noise”. I have used this technique occasionally with some success (Figure 8).
In some cameras, like recent ones from Olympus and OM System, this type of merging can also be done in camera using the Live ND feature. This feature is intended to allow longer exposure times in strong light, but works.
Conclusion
Bracketing makes it possible to take photographs with optimal settings when we are uncertain of the optimal values for these settings. It is also the starting point for different techniques that rely on merging multiple image frames into a single photograph with enhanced qualities.