The goal is almost always to reach the best possible balance between sharpness and noise. Sharpness, as we’re about to see in this chapter, is a huge area of importance and covers much more than most people might imagine. This is why I plan to devote many pages to it, because in low light photography more than most kinds, an essential skill is to be able to judge sharpness and unsharpness to a particularly fine degree.
In practice, setting out for a low light shoot requires a key decision—whether or not to use a tripod. On the face of it this may sound a little trivial, but it has far-reaching consequences. The two paths, handheld or tripod shooting, are each suited to different subjects and conditions, and they are widely divergent in style. They also depend on a decision before planning the photography, because unless you are fortunate enough to have an assistant or a willing companion to carry things, what you set out with in your camera bag and on your shoulder will determine the kind of photography you can do.
In this chapter, I want to look at the tripod-free approach, relying on hand-holding skills to get as close to normal, rapid-reaction, reportage photography as possible. Here, your own physical skills become important, especially how steadily and reliably you can hold and operate the camera. Professionals take some pride in doing this, but they are by no means immune to failure, precisely because they are pushing limits. One thing that low light photographers quickly get used to when shooting without rock-steady support is there is always a failure rate. Of x number of frames shot, y will be flawed because of blur of one kind or another. The aim is to bring the number y down as much as possible, and reduce the proportion of failures. Indeed, if every frame were technically perfect, this would be a sign that you were not pushing the limits as hard as you could.
TOWARD THE LIGHT
Making the most of what little light there is is essential when shooting handheld. Here shooting toward the light created dramatic silhouettes of these U.S. warplanes.
Low light photography is one area in which process dominates. By this I mean that all the various limits and difficulties that we looked at in the first chapter need to be thought about before you start shooting.
In practical terms, low light shooting is about pushing technical and mechanical limits as far as you can, and this in turn means that there are two competing ways of approaching it. In one—the subject of this chapter—the idea is to do everything necessary to retain the freedom of using a camera in much the same way as you would under normal light. The alternative is to abandon freedom in favor of a full choice of technical settings—sharp, clean images without noise, and at any aperture setting, and with no limit to the exposure time (except, of course, for short exposures).
It is not easy to switch from one to the other mid-shoot, so it is normal to decide at the start which route you are going to take for any one session. There are good arguments for both handheld and for using a tripod, and which to choose depends on the subject, your preferred style, the results you are looking for and the quality you absolutely need. The last factor is too often ignored, as we’ll come to in a moment, and can sometimes make the others redundant.
SUBJECT MOTION SETS LIMITS
Handheld shooting reaches its practical limit when the movement of subjects is too fast even for a combination of wide-open aperture and high ISO setting. This is a tripod shot at 1/10 sec, making deliberate use of the blur of a passerby.
LOW LIGHT REPORTAGE
Shooting handheld means being able to follow classic reportage methods, which include quick observation and photographing unnoticed. The difference is the lower light levels. Here, at a museum in Shanghai, bracing against a wall allowed a shutter speed of 1/4 sec with a 33mm EFL, and a low ISO setting of 100.
SIZE MATTERS
While noise is evident at 100% magnification on the computer screen, in print and at this size of reproduction, it becomes unnoticeable. Size of use is always an important consideration.
DETAIL
Remember to use the 100% magnification to check pixel-level detail.
The key factor for most photographers is the optimum shutter speed for subject movement. If the subject is active, for instance people in motion, a tripod is no help in freezing the movement. Some kinds of movement may well be acceptable, or even desirable, such as the streaked trails of vehicle lights on a highway, but if you need sharp capture, the threshold shutter speed for this determines everything. This in turn is affected by the relative movement in the image frame (see here), so a walking figure naturally calls for a faster shutter speed if it fills the frame than if it takes up just a small part of the picture. But basically, the rule is, if you are photographing people, then shoot handheld.
The obvious advantages are freedom and mobility, and the chance to shoot unobserved, without drawing attention to the camera as is inevitable with a tripod. You can strip down the amount of equipment drastically; one camera, one or two lenses, maybe a flash unit if the camera does not have one built in, and very little else. Gone are the days of two camera bodies, each for a different film (such as high-speed color, with high-speed black and white to be pushed to the maximum if necessary). Digitally, everything can be dialed up from the menu.
The disadvantages are in camera handling and image quality. As handheld shooting in these conditions almost never enjoys the luxury of fast shutter speeds, holding the camera (and yourself) still is always an issue, and may often—as we’ll see a few pages further on—call for over-shooting to lessen the risk of frames lost to camera shake. A fast lens, with a maximum aperture wider then f/2, is a distinct advantage, though costly. Image quality centers on noise, because the key strategy is setting the appropriate ISO. High sensitivity means high noise levels, and while there are ways of dealing with this it reduces quality. This completely depends on how the pictures are going to be used, in particular the size. It is very important to become familiar with just how bad different noise levels look when reproduced at particular scales. The positive side of this is that if you are going to use the picture small, the noise may be invisible, thus giving handheld almost no disadvantages at all.
HANDHELD MOVEMENT
Handheld shooting allows for a number of loose, ad hoc effects, such as a slow speed while walking, as in this image, taken at 1/3 sec.
DEPTH OF DETAIL
Images that rely heavily on fine resolution of detail, and which are likely to be used large, such as this formal architectural view of Angkor Wat, call for maximum image quality. Noise artifacts are especially unwelcome, so a tripod is the natural choice.
Handheld
Free, quick, and mobile
Unobtrusive, allows reportage-style shooting
Lightweight, needs little equipment
First choice for subjects in motion
Camera shake always an issue
Fast lenses are a priority
Needs high ISO settings, resulting in noise
Depth of field always shallow because wide apertures used
Locked down
Static, takes time to change viewpoint
Conspicuous, may attract unwanted attention
Tripod can be an inconvenient burden
First choice for static subjects
If tripod used well, no risk of camera shake
Any lens can be used, even very long slow telephotos
Can use lowest ISO setting. Good depth of field practical by stopping down
The ability to hold a camera steady tends to be assumed, and the general opinion is that you either can or cannot. In fact, it can always be improved by following certain techniques and by training yourself in specific ways.
In normal lighting and with standard camera settings, this is rarely an issue, as shutter speeds of at least 1/125 sec are usually possible. When the shutter speed needs to be longer, as in 1/30 sec, 1/15 sec and slower, camera shake becomes a pressing problem. At worst, you lose shots, while at best you waste time checking the image after each shot to confirm that it is sharp. Low light photography, with its typically slow shutter speeds, is when you most need to be able to hold a camera steadily enough to remain on the subject.
See here we’ll look in detail at what happens to the image under the influence of camera shake, but the issue is whether the image moves sufficiently in the time allowed for the exposure to be noticeable. This boils down to the speed of the twitch or jitter, which determines what distance the image jerks across the frame. An important secondary factor is the focal length of the lens. A longer focal length magnifies the image more, and this magnifies any movements, including camera shake. A widely accepted standard is that camera shake is unlikely at a shutter speed that is the reciprocal of the focal length, because most people can hold steady at such a speed. Thus 1/100 sec with a 100mm lens should carry no risk.
These are reasonable and slightly conservative estimates, but any serious photographer should aim to improve on them, provided you do not suffer from any condition which makes it impossible. When seeking that improvement, it makes sense to turn to activities that rely totally on holding a piece of equipment steady against a distant object. There is much to learn from rifle shooting and from archery, where there is a rich history of training and practice in aiming and hitting targets. In photography, locking onto the “target” is only one step in the complex process of making an image, but in low light it becomes especially important.
The first step is to know what you are capable of, and the way to do this, as well as to improve and train, is a kind of target practice. Photographing meaningless subjects may be anathema to many people, but frankly it is better to test your hand-holding abilities when the results do not matter. As always, it is best to avoid on-the-job training. Take a stationary subject that is convenient, at an easily focusable distance, and which has good sharp detail. Printed type is good, as in an optician’s. Make this a test that is practical for your own way of shooting, so assemble the lenses you normally use. If a zoom, decide on certain focal lengths, such as both ends of the scale and two or three points between, depending on the zoom range. Set the camera to shutter priority. Take a series of identically framed shots at one focal length. Begin a shutter speed high enough to guarantee sharpness—for instance, twice that of the reciprocal of the focal length, as 1/200 sec with a 100mm lens. This will be the reference. Then set the shutter speed to half as fast and take three shots. Adjust the shutter speed to half as fast again and take three more shots. Follow the steps shown in the Test Procedure table (see here). The reason for shooting three frames for each setting is that camera shake is unpredictable.
The amount of movement revealed by camera shake depends on the magnification, and so is proportionate to the focal length. These are the relative movements of the same amount of shake at three different focal lengths. Wide-angle shots suffer the least of all.
50mm
100mm
200mm
SLIGHT MOTION BLUR
SIGNIFICANT MOTION BLUR
EXTREME MOTION BLUR
EXTREMELY SHARP
ACCEPTABLY SHARP
Examine the results in an image browser or image database, ideally at 100 percent magnification and in small groups side by side. The results shown here are my own, created just as described. As I would have hoped after all these years, I can hold a camera steady at the reciprocal of the focal length, and that’s practice and experience more than anything else. It could still do with some improvement, and it occurred to me that I haven’t been following my own advice as much as I could. Nevertheless, that’s my benchmark, the speed at which I would expect to get every frame sharp. Slower than this, however, the figures become more complicated. Here, it’s useful to grade the degree of sharpness. Exact measurements, such as pixel smear, are irrelevant, as sharpness is ultimately perceptual. You do need to use your own judgment. I find the following categories, which are entirely personal, to be useful:
Extremely sharp
Significant motion blur
Acceptably sharp
Extreme motion blur
Slight motion blur
The results are shown in Table 1. On the face of it, there seems to be a reasonably clear separation, with the results from the three slowest shutter speeds quite useless. However, the interesting speed is the one immediately below my benchmark speed, where one of the three was acceptable.
So, on to the next stage of the test; to get a better sampling by shooting more frames in quick succession I took 20 consecutive frames at my benchmark speed, 1/focal length, or in the case of this 85mm lens, 1/80 sec. The result was that only a third of the frames were perfectly sharp, just over a half were acceptable, and a few were blurred. I followed this with another 20, at 2/focal length, or 1/40 sec, and finally 20 at 4/focal length, or 1/20 sec. The results in Table 2 show that there is no sharp dividing line between the lowest reliable shutter speed and the ones afflicted by shake. Instead, at the speeds below the benchmark, there is a percentage success rate. Moreover, even 1/20 sec, which would be foolhardy to try with any expectations, actually did deliver two acceptable frames out of 20. What we can learn from all this is first that over-shooting at certain speeds increases the probability of a sharp, usable frame; and second that the goal of training and practice is not just to improve your slowest reliable speed, but to improve the percentage success rate at all speeds.
How to improve these figures? First you need to know exactly what you are trying to correct and improve. Although this is a personal interpretation, four things cause camera shake, and each calls for different corrective techniques. These are drift, shaking, twitching, and jerking the shutter release. The first is a relatively slow movement around the subject caused mainly by the difficulty of concentrating absolutely on one thing for a period of time, and is entirely natural. However, it becomes important in photography only when some precise alignment needs to be made, such as the edge of a subject very close to the frame edge. Shaking is connected to heart rate, breathlessness, and the general level of nervous energy. This is promoted by adrenalin, so being excited about a shot you are about to take is unhelpful, even though difficult to control. Twitching is also related to nervous energy, but is a much more distinct and jerking movement caused by muscle contraction. In the context of aiming at a target it is more likely to be semi-voluntary than involuntary. Semi-voluntary means that you are conscious that it is about to happen because your brain is about to trigger the response, but that the tension caused by thinking about it, and by thinking about holding steady in general, seems to make it happen anyway. Finally, pressing down too hard on the shutter release does what you might expect, when all that is needed is a gentle squeeze.
People vary in how much they are prone to each of these, for physiological as well as psychological reasons. Drift is the least problematic, and can usually be solved by full concentration and by not trying to hold the subject in the viewfinder for too long. As for the minor tremors of shaking, some people are by nature calmer and steadier than others, but ways of improving this include any number of ways of relaxing and slowing the heart rate, such as by deep breathing. Taking a deep breath, then slowly and steadily exhaling, and squeezing the shutter release at the end of this cycle is one reliable technique. Another is to move onto a subject by either raising or lowering the camera, in order to take advantage of the steadying effect of motion. Twitches perversely tend to increase or are encouraged by the tension of trying to hold steady. One answer is to time the shot between two twitches, another is to raise the camera and shoot with as little delay as possible, trading accuracy in composition for steadiness. As you can see, there are a number of different approaches, and people tend to choose the one that suits their personality best.
Test procedure (100mm focal length)
This table indicates the camera settings you should shoot test frames at.
Reference frame
Basic standard
1 stop over standard
2 stops over standard
3 stops over standard
4 stops over standard
1/2 focal length
1/focal length
2/focal length
4/focal length
8/focal length
16/focal length
e.g. 1/200 sec
e.g. 1/100 sec
e.g. 1/50 sec
e.g. 1/25 sec
e.g. 1/12 sec
e.g. 1/6 sec
3 frames
3 frames
3 frames
3 frames
3 frames
3 frames
Table 1 (85mm lens handheld, 3 frames each speed)
1/160 sec
1/80 sec
1/40 sec
1/20 sec
1/10 sec
1/5 sec
Extremely sharp
3
3
Acceptably sharp
1
Slight motion blur
2
Significant motion blur
1
1
Extreme motion blur
2
2
3
Table 2 (85mm lens handheld, 20 frames each speed)
1/80 sec
1/40 sec
1/20 sec
Extremely sharp
6
2
Acceptably sharp
11
3
2
Slight motion blur
3
7
4
Extreme motion blur