Micro Four Thirds

A development of the Four Thirds format originally developed by Olympus and then Panasonic, which keeps the same sensor size but changes the lens mount and lenses to allow smaller, hybrid cameras.

The sensor size is 17.3 x 13mm, which is smaller than the APS-C format used for most digital SLRs, but still many times larger than the sensors used in compact cameras. Technically, Micro Four Thirds cameras should not be able to match APS-C cameras at higher ISO settings, but in practice the differences are not always easy to spot.

A combination of good quality lenses and compact, efficient camera designs means that Micro Four Thirds hybrid cameras are seen to offer pretty much the same quality as digital SLRs.

Focal factor

The focal length of a lens doesn't really tell you its actual angle of view because this depends on the sensor size too. But you can use a 'focal factor' to work out what the equivalent focal length will be.

For example, most digital SLRs have an APS-C sized sensor measuring around 24 x 16mm. This is smaller than a full-frame sensor, or a 35mm negative size, which is what most people go by. The difference is 1.6x (a full-frame sensor is 1.6x wider).

So if you use a 50mm lens on an APS-C camera, you multiply this focal factor by 1.6x, which gives you an equivalent focal length of 80mm.

In other words, because the sensor is smaller, a 50mm lens fitted to an APS-C camera actually looks like an 80mm lens.

Noise reduction

Image processing to reduce the noise produced by small sensors or high ISOs.

Noise is caused by random variations in the light captured by each photosite (pixel) on the sensor. Some cameras have sensors which apply 'on-chip' noise reduction, where electronic circuitry adjusts the signal values captured by the sensor's photosites. This is probably the best sort.

Otherwise, noise reduction is applied by the camera to the raw image data as it's processed to produce the photo. This reduces noise, but can result in smudging and loss of detail.

The same applies to noise reduction processes applied using software on a computer, though with more processing power available and more time to find the best settings, this may produce better results.

In any event, noise reduction is something you'd rather wasn't necessary at all. Using a small sensor and applying noise reduction is not as good as using a larger sensor which doesn't need it.

Megapixels

'Megapixels' refers to the number of pixels (in millions) captured by the camera's sensor. It's still used as a principal selling point for digital cameras, even though it's no longer anywhere near as relevant as it used to be and, in some cases, more megapixels do more harm than good.

In the early days of digital photography, cameras didn't have many megapixels. This limited the size of the pictures you could print before the blocky pixel pattern became visible. It also meant that cameras couldn't resolve much fine detail.

So higher megapixels meant bigger, smoother prints and finer detail, but only up to a point, because the other limiting factor is the physical size of the sensor.

Compact digital cameras use very small sensors. There's a limit to how much lenses can resolve on tiny sensors, no matter how good they are. And the smaller the sensor, the more you have to blow up the image to get a same-sized print. We'll call this the 'enlargement factor'. With enlargement factors over 30x, you can expect to see the image quality deteriorating, no matter how many megapixels you've got.

The table below should make this clearer. It shows how many megapixels you need for different-sized prints, but it also shows the enlargement factor required for these print sizes with different types of sensor.

As you can see, the small sensors in compact cameras are hitting very high enlargement factors very early on. This is what limits their picture quality in big print sizes, regardless of how many megapixels they have.

In fact, higher megapixel ratings bring serious technical difficulties which do affect the picture quality. If the megapixel count goes up but the sensor stays the same size, it means the individual pixels ('photosites', to be exact) on the sensor have to get smaller. This makes them less sensitive and more prone to random noise. As a result, makers have to use strong noise-reduction processes to make the picture quality acceptable, and today's high-resolution compact cameras often display weak definition and a 'smoothed-over' look to subtle textured details. The pictures you get are no better than those from cameras of five years ago with half the megapixels.

So why do makers keep increasing the megapixels? It's because megapixels sell. Most buyers will be unaware of the technical implications and just see a bigger number as being 'better', and as long as that keeps happening, the makers will keep on increasing megapixels regardless of whether it's a good thing or not.

To sum up, increasing megapixels in compact cameras are achieving nothing except camera sales. In fact, in many ways it's driving picture quality downwards not upwards.

Hybrid cameras and SLRs are different because the sensors are much larger and they haven't quite reached the same level of  'megapixel saturation'. Even here, though, sensor size counts for more than megapixels, and you shouldn't necessarily expect a proportional increase in resolution just because one camera has 18 megapixels, for example, and another has 12.

Sensor size

Different types of digital camera use different sized sensors. The larger the sensor, the better the picture quality, though the cost increases too. This diagram shows the common sizes:

Apart from specialised studio cameras, the largest are in full-frame digital SLRs, the next largest are in the more common 'APS-C' format SLRs and Micro Four Thirds hybrid cameras, while the smallest are in compact cameras.

The physical size of a sensor is now more important for picture quality than its megapixel rating. With a small sensor, the image has to be enlarged by much more to produce a same-sized print etc. It's like the difference between small negatives and big negatives in the days of film.

There are also technical problems with small sensors that have high megapixel ratings. The tiny pixels on these sensors are nowhere near as sensitive to light and they produce much more random noise than the larger pixels on bigger sensors. This means the makers have to build in strong noise reduction processes, which gives photos a soft and hazy look.

In fact while the resolution of compact cameras has climbed steadily over the past few years, it's probably fair to say that the actual picture quality has stayed the same or even gone backwards.

The bigger the sensor, the better the picture quality, and the gain in quality is pretty much proportional to the sensor size. The table below shows that you don't need that many megapixels to produce good-quality prints, but that the enlargement factor with small sensors is many times higher. Enlargements of up to 30x are usually fine (indicated in green), but when you go beyond this, the definition starts to suffer. There is a limit to how much any lens can resolve, and how much you can blow up the image it produces.

So megapixels are no longer particularly relevant for compact cameras because they've already reached the point where more isn't making any difference, and in fact may be making things worse.

Because the sensors in digital SLRs are much larger, there is still something to be gained from higher megapixel counts, though even here there are signs that diminishing returns are setting in.