Every digital camera is composed of a sensor. It is responsible for the image registration. But what elements form a sensor?
It consists of a large number of photosensitive cells. Each of them consists of a photodiode, whose function is to convert the photons that reach it into electrons. On top of each of them we find a microlens. Its function is to direct the rays of light that affect it towards the photodiode.
Going back to the cells, each of them corresponds to what we commonly know as a pixel, therefore, if we talk about a camera having 24MPx, it will have 24 million cells. The cells are responsible for detecting the intensity of the light, but not the color. For color capture, normally the sensors have the Bayer mask - composed of filters responsible for breaking down the light, into the three primary light colors, such as red, green and blue. The photons converted into an electrical signal are equivalent to the amount of light obtained and it is this signal - which in its nature is analog - that is sent to the DAC (digital-to-analog converter), in order to convert it into a digital signal by binary code (0 and 1).
The most common sensors are the CCD (charge-coupled device) and the CMOS (complementary metal-oxide-semiconductor). The CCD converts the photons that reach the cells into energy and sends the analog signal to the DAC in order to convert it into a digital signal. But the load of each pixel is transmitted row by row - both vertically and horizontally. The CMOS acts in a totally opposite way: each cell is independent, that is, each photodiode has its own electronics. Along with this, all work is done on the sensor itself, including image processing. The CMOS sensor is much faster, but digital noise is less present in the CCD sensor instead.
Digital noise is produced due to equipment deficiency. In chemical photography, we would be talking about the shortcomings that the film would have and call it grain. Noise is cleaner than grain. As have been said, it depends on the quality of the equipment, - understanding the camera sensor as equipment -, the more quality it has, the finer the noise it produces, and therefore the better definition the captured image will have. The noise measurement is set to signal/noise.
The sensors have a fixed sensitivity, it is always ISO 100. When we increase the ISO, we are not increasing the sensitivity of the sensor cells, but what we are varying is the electrical signal. The "problem" is that it always goes along the background signal. The background signal is greater than the electrical signal, therefore, by increasing the sensitivity, we are increasing the noise level. The higher the signal/noise (lower ISO), the lower the image noise. On the contrary, the lower that signal is (higher ISO), the greater the noise will exist. CCDs are especially characterized by better image quality since their "signal / noise" is wider than in CMOS. There are two types of noise: the photonic is generated when the photons reach the photodiode, occurs in the middle shadows and in the lights. And the reading noise generated in the electronics during the process of capturing light dominates in the deep shadows. Within this there are different types such as dark current, amplification noise and thermal noise.