In use: Panasonic Post Focus after-capture focus selection feature
As announced on Friday, Panasonic has just released new firmware that enables a post-capture 'refocusing' feature in some of its cameras that run 4K video on a Micro Four Thirds sensor. I got hold of a pre-release copy of the firmware and have been able to try it out on the Lumix DMC-GX8 body.
The new feature, which Panasonic calls Post Focus, is essentially a focus bracketing function that stacks a burst of images together into a single file and allows you to choose which focus point you prefer after the image has been taken by touching the subject on the camera's rear screen.
What makes this interesting is that the camera uses its 4K video mode to record the bracket, so it can shoot 30 frames in a second providing 30 different focus points in just one second of burst/footage. During the 'exposure' the camera scans the scene from front to back and shoots a frame to cover every focus distance required, so the depth of the scene has an influence on how long the process takes. The resultant images are grabs from 4K video footage, and come out the other end as 8MP JPEG stills.
The firmware adds a new item to the main shooting menu called Post Focus. In the menu you activate the mode and then shoot as you would normally. The new mode doesn’t use the whole area of the sensor to make the image, applying a 1.45x crop to the frame that adds apparent length to the lens in use.
When the shutter release is pressed the camera quickly scans the scene to determine the closest and furthest objects and then records a segment of 4K video while the focus passes over the pre-determined range. After a moment of processing the image is presented on the rear screen, and the user can touch the different objects in the scene to find the frame in which that object is in focus.
When an object is identified as the desired focus target, the user can magnify the view and use a scroll bar across the base of the screen to perfect the focus, with the option to use peaking for assistance. Once satisfied that the sharpest frame has been selected for that point, pressing the 'set' button will save it as an individual JPEG file on the memory card.
You can take as many frames from the footage as you like, capturing stills that show the same scene but with different objects in focus. There are 30 frames for each second of footage, so the total number of frames will be determined by how long the camera took to complete the file.
The basic idea is that focus point can be chosen after the picture is taken – much like the idea behind the Lytro cameras. That might seem a little odd to most photographers, as we tend to put focusing beforehand quite a long way up our priority list. However, the feature is aimed at more than the serious enthusiast and there are some other advantages to using this system that more advanced photographers might be interested in.
The Post Focus clip can provide a rich source of material for a focus stack, which effectively increases the depth of field in a given scene at a given aperture setting. Shooting a sequence of differently focused images at F5.6, which can then be stacked and blended, will produce a final result that has the sharpness of F5.6 but the depth of field characteristics of an aperture in the range of F32. This is one way to avoid diffraction-induced softness while retaining depth-of-field.
|Post Focus used to create a focus stacked landscape image. Panasonic Lumix DMC-GX8 and Leica DG Summilux 25/F1.4 @ ISO 200, F1.4, 1/8000sec|
The same principle can be used to maintain the defocused effect of F1.4 in the background of a portrait while the focus range across the face might be more typical of F4. By shooting a sequence at F1.4, using focus points from one eye to the other, the background can be kept extremely unsharp, while the face is all perfectly in focus. This gives something of the effect of a swing movement in a technical camera or with a tilt and shift lens.
|Composite of 7 frames blended. Panasonic Lumix DMC-GX8 and Leica DG Summilux 25/F1.4 @ ISO 200, F1.4, 1/2500sec|
The benefit of using the Panasonic system for both of these techniques is that multiple images can be collected in a very short space of time, and a tripod isn't needed, unless your hands are very unsteady. The images are collected automatically and all that's required is to review the clip to extract the images you want to work with. The user can specify aperture, ISO, filter effects, picture modes, exposure modes and aspect ratio in Post Focus mode.
The disadvantage is that the final image will never be more than eight megapixels. The image bursts are saved on the card as an MP4 file, so the clips can be played in movie software and the stills extracted in Lightroom if you find it too hard to view the camera's rear screen.
The Post Focus process is designed to be used with scenes that are primarily static, a fairly severe limitation that Lytro cameras do not have. A person walking across the frame will possibly never be in focus, and panning the camera during the exposure just doesn’t work. When the camera records it needs to be looking at the same content it looked at when it did the pre-exposure scan. You can have running water from a fountain, for example, as the stream of water doesn’t move, but anything shifting about in the frame upsets the apple cart.
This is a pretty impressive piece of technology, and it certainly does exactly what Panasonic says it will. But in reality, Panasonic is one of the least in need of a don’t-miss-the-subject focus bracketing feature, as it already offers a highly accurate autofocus system with very capable AF tracking. The Micro Four Thirds format also tends to have more forgiving depth-of-field than larger sensor cameras.
The feature is certainly very easy to use, and although it takes a good deal more time than a standard single frame, it also delivers a lot more options. It is also a lot of fun to use, amazing to see in action, and I think that there are plenty of people who will get value from its use. The attention Lytro cameras got clearly demonstrated the interest in this kind of feature, but in the Lumix cameras the images are bigger and easier to view – and the bodies also function as normal high resolution cameras. On the other hand, light field cameras offer benefits far beyond simply Post Focus, like reducing depth-of-field, directly selecting depth-of-field via a virtual aperture control, and depth-based image editing. And, of course, the ability to shoot non-static scenes.
It is true that many of us are keen to have pictures with more than 8MP in them, but on holiday and snapshot occasions eight is often more than we really need. For occasions that require larger images we’ll just have to wait for the arrival of 8K Raw video in this camera line. I'm not sure that the Post Focus feature will be something people buy Lumix cameras especially to get, but alongside the 4K Photo 30fps stills mode, it is another clever side benefit of having high resolution video.