Why do moving objects in photos appear blurry?

Question

When taking a photo on most cameras, if you take a picture of a moving object, the object appears blurry. Why does this happen exactly?

Answer 1

First, I'll talk about what cameras do normally, then about how motion affects this operation.

In order for an image to be sharp and in focus, all light coming from a single point on the object being photographed must fall on a single point on the film or sensor. If you take a picture of a face, you want all of the light reflecting off the left eye fall on one part of the image sensor and all the light reflecting off the nose fall on a different part. If the picture is out of focus, the light from different parts of the face can hit the same bit of sensor, and light from the same part of the face can be spread all over other parts. This results in a picture where every part of the face is mixed up with the other parts. This is called a blurry image.

If the subject is moving, a similar blurring occurs because the shutter of a camera is open for a span of time. Imagine you're taking a picture of a person, and that person moves their hand. When the shutter first opens, the camera directs the light from the person's had to a certain part of the image sensor. However, because the hand is moving, light from the hand's new position will be directed by the camera to a different part of the sensor. So, the camera will receive light from all of the positions of the hand while the shutter is open. That light from different hand positions will end up on different parts of the sensor. This results in what looks like a smeared image of a hand tracing the path of motion.

Answer 2

It happens because your subject is moving relative to the camera frame while the exposure is being made AND the shutter speed isn’t fast enough to freeze it.

Going into the details:

Shutter speed or exposure time is the length of time when the film or digital sensor inside the camera is exposed to light,when a camera's shutter is open when taking a photograph.The amount of light that reaches the film or image sensor is proportional to the exposure time. For example: 1/500th of a second will let half as much light in as 1/250th. When the shutter speed is slow (i.e., below 1/60th of a second), even relatively slow motions show up in photos. To brief up, fast shutter speeds have the effect of freezing motion in the scene you are photographing and conversely, slow shutter speeds will blur motion in a scene.

The chart below shows how different shutter speeds would effect the sense of motion if you were photographing a person running. Fast shutter speeds will freeze the motion.The slower the shutter speed becomes, the more blurred the person running becomes in the photograph.

And slow shutter speeds are usually caused by insufficient light. That’s why you rarely see motion blur problems outdoors on sunny days.

The solution:

The solution is to increase your shutter speed. And often the only way to do that is to add more light. One obvious way to do that is to use your flash. If you’re inside during the daytime, you could also just go outdoors. You can also increase shutter speed by decreasing (widening) your aperture. A wider aperture lets in more light giving you faster shutter speeds. If you are at your widest aperture and you still aren’t getting enough speed, you can try getting “fast” lenses(A lens with a larger maximum aperture,that is, a smaller minimum f-number): glass with f/stops 2.8 or wider.

You can also try a faster ISO setting.

Answer 3

The same thing happens with your own eyes, though your brain tries its darnedest to hide it. The fundamental problem has to do with how the image is created in the first place.

Sight is the interpretation of visible light reflected (usually; we can ignore active glow for now) from objects. To see something, it must be lit, and reflect that light differently than its surroundings. Light is formed out of tiny mass-less particles called photons - the carriers of the electro-magnetic charge. When a photon enters the retina in your eye (or the film in a camera, or the chip in a digital camera), it deposits some of its energy in some kind of photo-sensitive material, causing a change that can be measured and interpreted. By measuring the photo-sensitive material's response in many individual points, the brain (or chip) reconstructs an image of your surroundings.

The photon has three important properties - energy, position and direction. With a bit of geometry and optical correction, sight exploits the direction of the photon and the place where it interacts with the photo-sensitive surface to find out where the photon came from - roughly, which 3D point corresponds to a given 2D point on the image. The energy determines the colour of a particular photon. The idea is that the light coming from the object you're seeing comes roughly parallel, which makes the 3D->2D projection trivial. You get static blurring in a photograph when the optical correction is insufficient to compensate for the scattering of the photons in air - the bigger the distance to an object, the more dispersed the reflected photons are on average, and you need more correction to bring them back to be parallel. When the photons do not travel on paths that are parallel, the same point in the 3D space will correspond to multiple points in the 2D image - parts of the image mix with other neighbouring parts of the image.

But images usually aren't pure black and white. There's two other things that matter to humans - colour and intensity. Colour corresponds to the energy of the photons, while intensity corresponds to the amount of photons. And this is where things get interesting - to get any useful image, you need to absorb huge amounts of individual photons - a single photon doesn't really tell you much. So what actually happens is that you take (roughly) an average of photons that reached your sensor over a given amount of time - this gives you the relative brightness of things in the image, along with a good idea about the colour of the objects.

Human eyes add a few extra complications, so let's follow with an old-style film camera instead. The film is made of a material that changes permanently when exposed to light (think about what happens to a paper left out in the sun for a few months - but much faster). For simplicity, let's assume that the original material is perfectly black, while the changed material is perfectly white. Each individual photon causes a single molecule to change, but our eyes can't see the colours of individual molecules - they average the information from a certain area. So the more photons arrive in a certain area of the film, the brighter it will appear, corresponding to brighter light coming from that particular direction in space (and thus, the given volume of space, corresponding to, say, your bright red T-shirt). However, at some point, there's so many photons that all the molecules in a given area of the film are changed - illuminating it further cannot make it brighter anymore. Detail is lost, because as the surrounding areas get brighter, the saturated areas cannot. On the other side of the scale, if there's too little light, there will be too few photons to form a decent image - everything will be way too dark, with random-ish bright spots.

So to get a good image, you need to balance the time you expose the film to light. Too long, and your image is too bright and loses contrast. Too short, and there's not enough data to average out to a good image. As a side-note, this is the physical (as opposed to biological) reason why night sight is mono-chromatic - if there's too few photons coming in, their colour distribution results in a (random-looking) colour noise that makes it harder to see. Using only the intensity while disregarding colour results in a clearer, brighter image.

So let's imagine that you expose a bit of film to a 3D scene for a second. The brighter parts of the scene will result in more light interacting with the corresponding areas in the 2D image. But now imagine that at the 0.5s point, the guy in the scene moves his arm. The first half of the exposure has his arm in the original position, while the second half no longer receives the photons from the original position, and instead receives it from the new position. The total amount of photons reflected from the hand is the same, but they are now spread out over two distinct places in the 2D image; and averaged with the photons that came from the background when the hand wasn't there. If your hand moves at a constant speed, the corresponding photons will be evenly spread out over the path the hand takes between the start of the exposure and the end. You get the average of all the individual "images", just as if you took a hundred photos of people with slightly different postures and averaged them together.

How can you combat this? If there's enough light, you can keep the exposure short - this means that to get visible blurring, the object must be moving faster relative to a longer exposure. If there's not enough light, this will result in noise (the individual photons you measure are rather random - they just have a predictable distribution over time; there's a lot more red photons reflecting from a red shirt than green photons, for example). If you want to photograph a single moving object, you can try to eliminate any relative motion between the camera and the object - track the object. Humans do this automatically - you move your eyes and head to follow a moving object you want to examine, which gives you a clear picture of the moving object, while everything else is a blur (which the brain usually conveniently compensates for, but the camera doesn't).

Answer 4

The lenses of a camera carefully generate an image (usually upside down) of what you aim the camera at on a set of sensors.

These sensors add up the light that shines on them. Then they can be asked "how much light did you see?" and reset.

Typically, we only expose those sensors for a short period of time. The light coming from a particular direction over that short period of time ends up being the amount of light the particular sensor picks up.

The sensors are then mapped to pixels on an image.

When the object is moving fast relative to the amount of time we expose the sensors, the sensors at the edge of the moving object first picks up "no object here", then later "oh there is an object here". The amount of "object" vs "no object" is a function of how close to the object's edge you are and how fast it is moving.

If the object is a solid block of color, and the background a different color, this results in a smooth gradient going from background to object color at the object's edge along the direction of movement. We interpret this as "motion blur".

For the most part, objects and background are sufficiently different that we can spot it even if they aren't uniform in color.

We only see this sometimes because cameras vary how long they "stay open" depending on how much light there is. The less light, the longer they stay open, the stronger motion blur is going to be. Similarly, the faster the object, the more it will blur for a given fixed "stay open" time.

Modern computer science has actually reduced this problem; first, by making sensors more sensitive to light, and second by post processing. Many cameras will detect a uniform motion blur (caused by your hand moving) and invert it after the image is captured. In theory, this can even be done for a single moving object in a scene, but determining what is object and what is not is harder here. I'm unaware of a camera that does this automatically.

Answer 5

When the shutter button is depressed, an image of the outside world is fleetingly projected onto the image sensor (or film). This action is called the “exposure”. To answer your question, you need to know that during the exposure, the projected image is being recorded. The key point is, the image sensor (or film) is accumulating light energy over time. Should the image change in any way during the exposure, the image recorded will likely show this as fuzziness. We try to keep the camera as still as possible to avoid this fuzz. Additionally, we try and choose a shutter speed that is super quick. In this way our images are moments frozen in time.

Answer 6

There are two primary kinds of blur in photos (well, three, but I’ll assume you keep your camera fairly clean): focus blur and motion blur.

Focus blur happens when the subject of your photo is simply out of focus. The solution to that is to make sure your autofocus is on and try again. If it’s out of focus, re-focus and shoot again. Pretty straightforward. On point and shoot cameras, the most likely reason you’re out of focus is because the subject moved or the smart focusing system wasn’t so smart and focused on the wrong object.

Motion blur, on the other hand, doesn’t happen because your subject is out of focus. It happens because your subject is moving relative to the camera frame while the exposure is being made AND the shutter speed isn’t fast enough to freeze it. Let’s tackle those two aspects separately.

So, the solution is to increase your shutter speed. And often the only way to do that is to add more light. One obvious way to do that is to use your flash. If you’re inside during the daytime, you could also just go outdoors. Sometimes the difference between shade and sun is all the extra light you need.