Not long ago, I was reading some forum comments regarding the announcement of a new filter and I was struck by how many people were of the opinion that filters, and in particular neutral density graduated filters (GNDs or ‘grads’), were an irrelevance as far as modern digital photography was concerned. Around the same time, someone commented on one of my Facebook posts, asking whether I was going to stop using GNDs now that the HDR function in Lightroom is so good.
The answer to the question is that I continue to use grads, as I think that they offer a number of advantages over the alternatives.
Before we look at those, though, I need to tackle one common misconception, that the graduated filter tool in Lightroom provides an alternative to using physical grads. While Lightroom’s grad tool is excellent for enhancing skies – giving you a huge amount of control over a number of adjustments and not limited to just exposure – it will only work if your camera’s sensor has captured all of the information in the sky in the first place. In other words, if the tonal range of the scene you are shooting exceeds the dynamic range of your camera’s sensor and the sky is overexposed, no amount of fiddling around with graduated filters in Lightroom will help – you can’t reveal detail that isn’t there. So, in this situation, you will have to either compress the tonal range of the scene by using a grad, or bracket your exposures and blend them during processing. In post, you can then use the graduated filter tool to fine tune skies.
Exposure blending (where two or more bracketed images are combined to create a single image containing the full tonal range) is a popular technique and does offer some advantages over traditional, in-camera filtration. Firstly, overall image quality should theoretically be better, as you are not placing a piece of glass or plastic in front of your lens. Secondly, it offers finer control than using grads and, finally, the results can be less obvious. One criticism of grads is that they can darken parts of the scene which you don’t want them to (e.g. tops of mountains or buildings), making it obvious that a filter has been used. Advocates of blending argue that it gives a more natural result.
Starting with the last point, if the use of filtration is obvious, then this indicates poor technique on the part of the user rather than a problem with grads per se. By choosing the correct strength filter and the correct type – hard, soft, medium, very hard, reverse grad, etc. – and placing it carefully in the frame so that the transition line blends in unobtrusively, the use of filtration should be more or less invisible. On the rare occasions that it is not possible to prevent the transition line showing in the image, then a little dodging and burning at the processing stage will take care of the problem. Equally, badly executed exposure blending can also be obvious, leaving ‘halos’ around the edges of objects – but this doesn’t invalidate the technique.
It is probably true to say that you get greater control if blending exposures, but remember that skies can be fine-tuned quickly and efficiently in post. Regarding the impact of filters on image quality – provided you use good quality filters and keep them clean and free of scratches, they should have no real-world impact.
There are situations when each approach has its distinct advantages. If you bracket exposures, there is always the chance that something can happen between one frame and the next: the light can change (especially during long exposures), making it difficult to achieve a natural-looking blend; the branches of trees can move, or you might find yourself shooting your exposure for the foreground at the moment when the sky is at its peak and thus fail to record the sky at its best. There have been occasions when I have just not had the time to bracket exposures. For example ...
... I have been lucky enough to shoot a number of times at the ‘ice beach’ at Jokulsarlon in Iceland. For the best images there, I needed to work quickly, getting in close to the ice in between waves. With only a few seconds to compose, focus, set exposure, get the picture and then run before the next big wave wiped me out. Staying an extra few seconds in order to bracket exposures was not an option. Obviously, in these situations, filters such as the B+W series are by far the best way to deal with excessive contrast.
On the other hand, grads are not always the best option. Sometimes, the contrast is so great that you can’t control it with filtration, or sometimes the horizon is so uneven that, even with good technique, it is not possible to use grads unobtrusively. In these situations, I prefer to blend exposures.
Mastering FiltersRead more free articles on filters and improve your landscape photography
The main reason I generally prefer to control tonal range in-camera rather than in-software, though, is that I like to keep my time in front of the computer to a minimum. As a full-time photographer, I already spend far too much time in the office and I like to maximise my time behind the camera. If I had to blend the majority of images, no matter how efficient I became at doing so, it would add significantly to my computer time.
There is also the fact that, to be honest, I just enjoy using filters. For me, it is part of the whole ritual of setting up the camera and preparing for the capture. This ritual is important to me – I enjoy the process of photography as much as I do the end result – and using filters is an important part of this process.
The one thing that might make me leave the grads at home, however, is if dynamic range continues to improve. For the average dawn or dusk picture, we would still need several stops more dynamic range than the best sensors already have, but this is not impossible to imagine. With digital sensors having five or so stops more useable dynamic range than the transparency film I used to shoot with, I already use grads less often than I did a decade ago. Who knows, I can’t see it happening in the immediate future, but there may come a day when they become redundant – but not yet!