Multiple jokes

The noise of digital camera mainly refers to the rough part of the image produced by CCD(CMOS) in the process of receiving and outputting light as the receiving signal, and also refers to the foreign pixels that should not appear in the image, usually caused by electronic interference. It looks like the image has been dyed and covered with some tiny rough spots. The digital photos we usually take may not be noticed if we use a personal computer to reduce the high-quality images and then look at them. However, if the original image is enlarged, there will be a color (pseudo-color) that did not exist originally, and this pseudo-color is image noise.

In addition to noise, there is another phenomenon that is easily confused with noise, which is a bad point. Under the same digital camera setting conditions, if the clutter always appears in the same position in the captured image, it shows that this digital camera has bad points. General manufacturers have regulations on the number of bad spots. If the number of bad spots exceeds the specified number, you can replace the camera with dealers and manufacturers. If the noise points do not appear in the same position, it means that these noise points are due to the noise formed during use.

Noise: both CCD and CMOS photosensitive elements have the problem of hot pixels, which is related to image quality and temperature. If the camera temperature rises and the noise signal is too strong, it will form mottled spots where it should not be. These points are what we call dry points. Various brands and models of cameras have different ability to control dry spots. Cameras of the same model also have certain individual differences, and some cameras have the function of reducing dry spots. But the problem of dry spot is a problem that all DC can't completely overcome now (improving the sensitivity (ISO), especially when the exposure time is long or the camera temperature rises).

The cause of noise

Both CCD and CMOS photosensitive elements have thermal pixel problems, which are related to imaging quality and temperature. If the temperature of the machine rises and the noise signal is too strong, it will form mottled spots where it should not be. These points are what we call noise. Various brands and models of cameras have different noise control capabilities. Cameras of the same model also have certain individual differences, and some cameras have noise reduction function. However, the noise problem is a problem that all DC's can't completely overcome now (improving the sensitivity (ISO), especially when the exposure time is long or the camera temperature rises). How different is the sensor structure of noise from the processor?

1, noise generated by long exposure

This phenomenon mainly occurs when shooting low ISO night scenes, and there will be some messy bright spots in the dark night sky. It can be said that the reason is that the processor can't handle the huge workload caused by the slow shutter speed, which leads to some specific pixels out of control. In order to prevent this image noise, some digital cameras are equipped with a function called "noise reduction".

If the noise reduction function is used, the image noise will be eliminated by digital processing before recording, so it will take a little more time before saving. However, with the opening of noise reduction function, the details of the picture will be lost.

Noise generated by 2.JPEG format image compression.

Because the JPEG format image still looks natural after the image size is reduced, special methods can be used to reduce the image data. At this time, processing will be performed in units of 8×8 pixels. Therefore, especially at the edge of 8×8 pixels, it will be unnaturally combined with the next 8×8 pixel unit.

The image noise produced by JPEG format compression is also called block noise. The higher the compression ratio, the more obvious the image noise.

Although the noise will become invisible after reducing the image, it will be very obvious as soon as the color compensation is carried out after enlarging and printing. This image noise can be solved by recording images with the highest possible quality or by methods other than JPEG format.

3. Noise caused by fuzzy filtering

The image noise produced by fuzzy filtering is the same as that produced by JPEG when processing images. Sometimes it is generated during the internal processing of digital cameras, and sometimes it is generated when using image retouching software for processing. For small-size images, when color edges are emphasized to make the image appear clearer, image noise will occur.

The so-called clear processing refers to the function of digital camera to emphasize the color edge of image and the "fuzzy mask" function of image editing software. In different styles of digital cameras, some cameras also emphasize the color edge of the whole image. After processing, other color lines will appear outside the original edge.

If you reduce the image and use it on the internet, don't you always feel that the image will become blurred? At this time, if the image is clearly processed by using the "fuzzy filtering" function, the image will look better. However, due to image noise, this image noise is very troublesome in the second or third processing. Don't make the image look too rough because of over-processing.

4, the photosensitive element area is too small to cause noise.

Noise comparison between SLR digital camera and ordinary consumer digital camera In general, the noise of SLR digital camera is obviously better than that of ordinary consumer camera, which is determined by the area of photosensitive chip. More than 90% of the noise visible to ordinary people is caused by this reason.

The conversion efficiency of the photodiode of a pixel is inversely proportional to its area. For example, if the conversion efficiency of a photodiode with an area of 1 cm2 is about 35% (this set of data is hypothetical data for intuitive explanation only), when the area is reduced to 7 square microns (pixel area is 1 DSM K3), the photoelectric conversion efficiency will drop sharply to 1%.

The low conversion efficiency makes the signal read from CCD/CMOS have to be amplified before it can be used. The ISO we adjust is actually to adjust the amplification factor of this signal, but it is impossible to amplify only separated level signals in the signal amplification process, which is bound to be accompanied by synchronous amplification of noise signals. In addition, the integration of CCD/CMOS is higher, and the heat generation is bound to be higher, and the thermal noise will be greater. This requires various noise reduction algorithms to contribute, but if the input signal quality is low, the noise reduction algorithm itself can play a limited role. The graininess we often see on portable DC is heavy, or the color boundary is blurred, which is the effect brought by different noise reduction algorithms.

If it helps, please LZ adopt it in time. Thank you~