How is the "noise" formed in photography? That is the mechanism of noise formation?

Concept: Noise

Noise mainly refers to the rough part of the image produced by CCD in the process of receiving and outputting light as a received signal. So how did the noise come from? First of all, we should understand that for a digital camera as an electronic product, the internal image sensor will inevitably suffer from different degrees of optical electromagnetic interference from peripheral circuits and its own pixels. In short, there must be some noise in the photos taken, which is inevitable. All we see is the degree.

Why is noise generated at high sensitivity?

The sensitivity of traditional film is to change the sensitivity to light by changing its chemical composition, while the photosensitive element of digital camera is unchanged. Generally speaking, digital cameras generally adopt the electronic signal amplification gain technology, and the electronic signal amplification gain value corresponds to the ISO value, for example, it provides the gain amplitude equivalent to ISO 100 when set at the standard value, and the gain values corresponding to ISO200 and ISO 400 can be realized by increasing the gain amplitude. So how is the gain amplitude achieved?

When providing high sensitivity, it is naturally necessary to provide corresponding gain amplitude, and the corresponding signal must be amplified before outputting the image signal, because the output level of CCD is low, especially when the ambient light is dark, in order to make the image change quantitatively, the amplifier will increase the gain amplitude according to the corresponding ISO value.

In addition, increasing the number of pixels in a given CCD area will make it difficult to maintain sensitivity. The area of unit pixel decreases and the intensity of incident light weakens. If a higher gain value is called to improve the ISO value, the image quality will deteriorate. Therefore, the manufacturer also provides a solution, that is, when reading the image signal, adding the original image signal stored by CCD can improve the sensitivity relatively. Adding two pixel signals and transmitting them can double the original sensitivity. However, this working mode also has its disadvantages. It is only suitable for smaller image modes, because the number of pixels is halved during processing.

The third solution for the image industry to provide high ISO value is to take a few pixels as 1 pixel for photosensitive, so as to improve the photosensitive speed. For example, the standard ISO 100 is sensitive to every pixel of the photosensitive element. To improve the sensitivity to ISO200, we only need to treat two pixels as 1 point to get the original photosensitive speed, and if we want to improve to the level of ISO400, we can get four times the photosensitive speed by analogy as long as we treat four pixels as 1 point. The difference from the second method is pixel merging, and the second method is the superposition of photosensitive signals, so this method has advantages for high pixel models, but the noise is more obvious than the first two methods.

(Reference: /view/ 142807.htm)