How digital affects the image

Digital gets as close as it can to reproducing an exact image of the subject. But the image is analog, the camera sensors are analog, and the camera digital output is an average based on several factors. There really isn't anything true about the image, just as film images are affected by a number of factors besides rolling shutter and moiré.

The sensors in a digital camera cannot recognize color; they only respond to light intensity. A filter is used to channel red, green, or blue light to different sensor sensors in an area. The sensor responds to the intensity of light hitting it, and creates a corresponding analog output for the color of light hitting it. Surrounding sensors are sent a different color of light by the filter.

Everything is averaged. The electronics takes an average of a sensor that sees one color, and its surrounding sensors that see other colors, and from that average determines a color value. By sensing each sensor with its other close sensors, an accurate color representation is found.

Digital technology is affected by a wide variety of factors. The ones that are really important for your consideration are explained in following sections. These include the size of the sensor chip, the frame rate, the number of pixels, the output format (compression used), and the number of encoding code bits.

Sensor chip size

Sensor chips, which are filled with millions of tiny sensors (photo sites), each representing a pixel or a group representing a pixel, come in many different sizes. The physical size of the sensor doesn't necessarily indicate the number of photo sites on it. A 1/3" sensor may have as many photo sites as a full 35 mm sensor. But in general, a larger sensor chip may have more photo sites. Advantages, or disadvantages, come from the size.

Light sensitivity increases with a larger photo site, which can receive more light. Light sensitivity can be very important when operating in low light conditions, and at higher ISOs. It gives you less noise, and a higher quality (truer) result. So in general, larger sensors offer a higher quality image.

Larger sensors actually narrow (crop) the diagonal measure of the area being shot. A wide city scape may be narrowed to only a few buildings when using a larger sensor. This is counterintuitive, but when using the same lens, the focal width at the sensor is the same, but the larger sensor is less covered. So a smaller sensor will give you a wider view. This can be confusing because, for example, Canon® came out with a new 7D camera but the sensor was smaller than in their earlier 5D.

With the diagonal measure being shot being smaller with a larger sensor, this means you may have to be farther away from the subject to get the wider shot. On a set, or for a city scape, you may not be able to back up far enough to get the shot.

Larger sensors have the subject in focus for a greater distance (deeper depth of field), so the subject, someone farther away, and the background may all be in focus at the same time. This can be good, but provides less opportunity for using rack focus and bokeh in your shots.

A smaller and less sensitive sensor may make your shots very subject to lens type and quality. Your zoom lens may not have as good a quality glass (or plastic), as your prime (fixed) lens. If this is so, the larger sensor may give you better light sensitivity in low light conditions (or allow you to use a higher ISO.

A small sensor may give you a picture that is much closer to 35 mm in diagonal size, but not give as good a quality picture as a larger sensor, particularly in low light conditions. This can vary by product by increasing the number of photo sites on a sensor, and by improvements in CMOS sensor technology. A small sensor may give you greater control over focus and bokeh.

Number of pixels

The larger the number of sensors, the more accurate an image can be represented, in color, in resolution, and..., oops, covering up for bad sensors.

When you have a device with millions of sensors, one or more may go bad. But because several sensors are averaged, the effect isn't noticeable.

The more sensors you have, the less the final image can be affected by a bad sensor. To create an HD 1920 x 1080 image, over 2 million sensors are required. More sensors, more accuracy. For example, the Canon 5d Mark III has a 22.3 Megapixel sensor, but its output is 1920 x 1080.

The NIKON® D800 has a 36.3Mp sensor. The Blackmagic® Cinema Camera EF has a sensor resolution of 3.3 Megapixels. The Canon 7D has a 18 Megapixel sensor. The Panasonic® AJ-HPX3000 (HD 1080p broadcast camera) has a 2.2 Megapixel sensor.

Bits: 8 - 32

The number of bits your camera stores information in directly affects the color quality. Cameras with an 8-bit output are only able to store 256 different colors, which is a very low color depth, and very noticeable. For example, when you look at the gradation of colors and shadows on a face using 8-bit color, the face will look splotchy and unreal. You can store a larger variety of colors by buying a camera with higher number of bits used to store the colors.

For example 64-bit computers are able to store and display 16 million colors. But that comes at a price. It takes a lot more bandwidth to send, and a lot more computing power to display. 12-bit cameras store over 14,000 colors, and 14 bit cameras store over 16,000 colors. It is very difficult to do color correction using only 8 bits. High compression often reduces the number of bits, and so reduces the number of colors that get stored and displayed.

The Nikon D800 creates a 14-bit file. The Blackmagic Cinema Camera EF creates a 12-bit file. The Canon 5D and 7D create a 14-bit file. The Panasonic AJ-HPX3000 (HD 1080p broadcast camera) creates a 10-bit file.

Bitrate and bandwidth

Bitrate is the number of computer code bits sent in one second. This affects the bandwidth required to stream or download a video file. The higher the number of bits, the less compression used, and the higher the resolution, the more bits per second are sent, and the more bandwidth required. This is becoming less of a factor for many users. The average broadband user today has a 4 Mbps connection speed (bandwidth). This is the minimum required for 1080p movies.

Conversely, the fewer number of bits, the less compression used, and the lower the resolution, results in a poorer quality image.

Next page: The digital output.

2   3   4

Or click a Page selection on the top menu.