The more pixels the better? The smaller the pixel area, the better? Why are people crushed to death?
The pixel is: the number of the smallest image unit, for example 6000*4000, the number of pixels is 24*10^6.
If there are too few pixels, of course, the image will be invisible and not clear.
But the pixel counts of almost all mobile phones and cameras now far meet the requirements. Nowadays, the limitation of image clarity is not usually a pixel problem, but a lens and frame problem.
The clarity of the captured image is determined by the lens and sensor.
Please refer to my other blog on how to decide this together:
https://blog.csdn.net/cashmood/article/details/112799151
We often say that the first level is overwhelming. It is true that full-frame cameras are relatively expensive.
Since the full-frame camera has a large area, does it have more pixels?
I found Sony a6000 half-frame and Sony a7m3 full-frame on Taobao to compare:
We see that the two of them have a full frame and a half frame, but the number of pixels is similar.
According to the formula: area of a single pixel * number of pixels = total area of the sensor
Therefore, it can be seen that the essence of the biggest difference between full-frame and half-frame is: the size and area of a single pixel are different.
Obviously in the above case, the larger and more expensive the pixel area, the better the photo effect. In fact, the colors are more vivid.
why?
Because the size of a single pixel is very small: (the disadvantage of a small pixel)
- The smaller the light input, the weaker the signal received by a single pixel, and the lower the signal-to-noise ratio .
- Due to the lateral diffusion of minority carriers in the substrate, the signals between pixels will interfere with each other.
- Each pixel on the sensor cmos has a microlens to increase the ability to receive light. When external light is incident obliquely, the optical signal between the microlenses on the pixel crosstalks.
- From the understanding of wave optics and Fourier optics , the diffraction effect of light is more obvious, when the light is incident obliquely, the signal crosstalk caused by the incident of high frequency light.
Is the area of the pixel unit larger, the better? Of course not. If the entire full-frame sensor has only 10 pixels, it is big enough, but it must be invisible.
Therefore, the larger the pixel, the better the premise is that the number of pixels is enough to satisfy the Nyquist sampling theorem, or that the acceptable frequency component is greater than the frequency component obtained by the lens.
Then there are strange phenomena, why do mobile phone manufacturers still make pixels so big? For example, the original pixel of the main camera of Huawei P30Pro is 40MP, which is 40 million pixels, but the output is 10 million pixels.
This is another very important benefit of small pixels:
When a high-pixel photo is reduced to a low-pixel photo, the noise of the image becomes smaller and the signal-to-noise ratio increases.
Summarize the two benefits of a small pixel area:
1. Under the same sensor area, it is easy to make the number of pixels large, and then collect high-frequency components with many details. Nyquist sampling theorem.
2. When the high-pixel photos are reduced to low-pixel photos, the noise of the image becomes smaller and the signal-to-noise ratio increases.
So, the conclusion is:
On the premise that the number of pixels is sufficient, the larger the pixel area, the better, and the larger the frame, the better.
But the picture frame cannot be infinitely large. On the one hand, it is limited by the CMOS process, and on the other hand, it needs to be matched with a large lens.
This is my summary after various understandings. If you have different ideas, you can leave a message for discussion. Please indicate the source for reprinting.