Google the term ‘converging verticals’ and you’ll be convinced the effect is the work of the devil, with countless articles and tutorials dedicated to rectifying the issue. It’s certainly true that the effect can be undesirable, particularly in architectural photography where it can make buildings and other structures appear as though they are falling over (an effect known as keystoning).
And yet, embracing it can pay off creatively too.
In the following article, I’ll run over how to maximise this effect in your images and the key things to bear in mind for successful results. But first, it’s useful to understand why it occurs.
Why it happens
Converging verticals are typically caused by a camera's angle relative to the subject, and the difference in magnification between different parts of the scene that come as a result of this. We usually see this effect when photographing buildings, so what follows will use this as the example.
Normally, you will be far closer to the base of the building than to its top, but to fit it into the frame you have to have your camera angled to a certain level. Doing so can make linear details and the sides of a building appear slightly skewed at best, and potentially seem like they're disappearing into the distance at worst, depending on how and what it is you're photographing.
Tilt-shift lenses are an obvious solution to this. By shifting the lens, you move the optical axis so that you can get the top of a building in with your camera perpendicular to the subject. Yet, the expense of these lenses largely confines them to professional use.
Tilt-shift, or Perspective Control, lenses are often used to correct converging verticals
You can correct minor converging verticals with image-editing software, but anything more serious introduces a handful of issues. One is that you need to ensure you have enough redundant space around your subject that can be safely cropped away once you straighten out your image. Otherwise, you risk cropping away valuable details.
The other issue is that this stretching compromises details in the image. Typically, the base of the image will not be stretched too much as it will be relatively close to how you want it to end up, but the top will be – and the results aren't always pretty.
There's not enough detail around this building to be able to correct the building without cropping it into a different composition (right). Furthermore, the required amount of correction would severely distort the hotel's sign.
Assuming you don’t have a tilt-shift lens, the best approach to this while shooting is to have your camera as perpendicular to the subject as possible, which will usually require you to step back until you can fit it in.
This will reduce the difference in magnification ratio between different parts of the subject, although you may have to compose with a wider lens than usual to get yourself as head on to the subject as possible. Peripheral details can then be cropped away later to your chosen composition.
Clearly, then, doing the opposite to this – getting as close to the subject as possible and angling your camera sharply upwards – will help you to exaggerate the effect.
Using a wideangle lens to get as much detail into the frame at the closer end of things will also help to place emphasis here, at the expense of details further into the scene. Following these two principles, however, is no guarantee of success; there are a few other things to take into consideration for good results.
Choice of subjects
The first of these is your choice of subject. One thing to try is to seek out subjects that already have some kind of converging verticals, as we already expect them to appear this way to some degree.
In the image below, we can appreciate that the buildings in the foreground have straight sides, and capturing them from this angle makes them appear to be toppling backwards.
Here, however, the very different form of these spires means we can tolerate the effect to a greater degree, even though the camera was positioned much further to the subject than in the previous image. The fact that there are more details here also makes the image more interesting.
Subjects that are positioned all around the frame in a relatively even manner – tall trees captured from underneath with a wideangle lens, for example – also work well. In such images, we can appreciate that the photographer is not trying to record this without converging verticals – otherwise, they would not have chosen to capture it from that position. A well-balanced composition and use of lines that lead through the frame can often be the best foundation for successful images captured in this way. Appropriate depth of field, however, is also a key consideration.
Depth of field
Depth of field decreases as you get closer to the subject. So, if you’re focusing on something very close to you it may be difficult to get a more distant part of the subject in focus at the same time, even with a very small aperture.
The ability to adjust the plane of focus is one of the key advantages of the tilt functionality of tilt-shift lenses, and this saves you from having to use a very small aperture to get the depth of field you require. If you don't have one to hand, though, you can try a couple of alternative techniques.
Depth of field is a key concern when attempting to exploit converging verticals for creative effect
Hyperfocal focusing is one of these. This is the distance at which everything from half of this point to infinity will be rendered as acceptably sharp. Depth-of-field calculators, such as mobile apps or physical charts, are a useful way to check this distance.
Focus stacking, where you blend a number of images with different parts of the scene in focus, can also be very effective, although this is only really practical when using a tripod. Traditionally such images are assembled using software, although some cameras now are beginning to offer the function too.
It’s also worth remembering that you don't get the same ratio of depth of field behind and in front of the point of focus at all subject distances. It's often stated that two thirds of this zone extends behind the subject and a third extends in front of it, and this is the case at a specific distance, but you actually get more equal proportions as you focus closer.
Distortion
Even very expensive wideangle lenses can suffer from curvilinear distortion to some degree, and subjects with linear details highlight the degree to which this is the case.
Turning on your in-camera corrections may help here, although you may not be able to appreciate how this will appear until you look at a much larger version of the image on your computer.
Left: The uncorrected image, which clearly shows curvilinear distortion from the lens. Right: The corrected image, which has also been processed for greater effect
Treatment
For images that do not have much detail (modern architecture, for example) increasing contrast to highlight form and converting to black and white can also work well. The application of vignetting can also mask any more problematic areas (such as the joins in the top-left corner in the images above).
One way to get around issues with focus is to go the other way and produce something completely out of focus and/or abstract. The image below, for example, is of a handful of coloured crayons that were captured beyond the minimum focusing distance of the macro lens used, in order to keep their form but render them out of focus. Obviously you need a fairly strong subject to begin with and take care not to make it too obscure.
This abstract, defocused composition gets around any issues concerning depth of field
. digitalrev.com2017-4-20 03:00