Shulou(Shulou.com)11/24 Report--

We will use high sensitivity when shooting dark-light portraits, galaxies, and ecological sports in dark-light environments, so what factors affect the quality of the pictures with high sensitivity, how we choose machines, and how to shoot, is to explain to you this time.

First make clear what the "pixel" is: the photosensitive element cmos is responsible for sensing light, which is what we often call full-frame, APS-C-frame photosensitive elements. The photosensitive element is covered with photodiodes, and there are as many photodiode units as there are pixels, so each photodiode unit is a pixel. Each photodiode is attached to the "microlens". The photodiode is responsible for processing the brightness information of the light, and the microlens is responsible for processing the color information, so that the photosensitive element converts the light into an electrical signal and gives the camera the next step of processing.

The most basic imaging principle is divided into the following four steps: 1. Light enters the lens and shines on the photosensitive element.

2. The photodiodes on the photosensitive elements will be converted from optical signals to electrical signals.

3. The electrical signal is converted into digital signal by A / D analog-to-digital converter.

4. The digital signal is converted into image and displayed by CPU and controller.

Which step does the noise come from? When the photodiode of the photosensitive element converts the electrical signal, it will produce electromagnetic interference and form noise. In fact, in the process of converting optical signals into electrical signals, there are both normal and abnormal electrical signals, and abnormal electrical signals are noise points.

Why does the noise increase when increasing the sensitivity? Shooting under the condition of sufficient light, there is no need to improve the sensitivity, and the abnormal signal is very few, so the anti-jamming function of the camera can easily filter out a small amount of interference signal, so the noise can hardly be seen in the final photo.

However, under the condition of low light, it is necessary to increase the sensitivity in order to obtain a suitable exposure, and to increase the sensitivity is essentially to increase the magnification of the optical signal into electrical conversion, so the intensity of both normal and abnormal signals will be improved. as a result, some abnormal signals can not be recognized by the processing circuit and can not be removed, so the higher the sensitivity setting of the camera, the more obvious the noise of the photo.

To suppress and reduce noise, there are the following solutions: the solution is related to the four steps of camera imaging.

1. In step 2 of selecting a low-pixel machine, when the photodiode is sensitive to light, the larger the area of each unit pixel, the more sufficient the sensitivity will be, and the better the quality will be when the light signal is converted into an electrical signal. So how do you make the area per pixel larger? Under the premise that the area of the photosensitive element is the same, choose a low-pixel machine! So that's why Nikon Z6's high-sensitivity picture quality is better than Z7's, Sony a7s3's high-sense picture quality is better than a7r4's, and sports machines are all low-pixel. Therefore, the choice of camera is not the higher the pixel, the better, non-commercial large-scale printing, no more than 40 million pixels, shooting Milky way, dark portraits, video and other themes, choose more than 20 million pixel machine is very good.

2. In the process of selecting a machine with dual primary sensitivity from step 2 to step 3, the ordinary photosensitive element has only one processing circuit (single primary circuit), so no matter how high the luminosity is, they all carry out the same treatment and have no advantage.

But for a machine with dual primary sensitivity, in this step, there is more than one processing circuit, as shown in the following figure, above is a single primary circuit, and below is a double primary circuit.

The photosensitive element with double primary photosensitivity adds more than one analog signal amplification path. after the optical signal is converted into electrical signal, it enters the processing stage of double primary photosensitivity. When using high sensitivity, enter another processing circuit, specifically for high sensitivity signal processing, stronger noise reduction ability.

The above figure is a schematic diagram of the conversion circuit, the ordinate is the noise level, the Abscissa is the sensitivity, this figure is in iso5000, another circuit will be enabled, thus controlling the noise to a level similar to that of iso1000.

For example, Panasonic S5 full-frame micro-single uses dual primary sensitivity, in the iso640 and iso4000 two nodes, will switch to a more noise reduction circuit, the image quality will be improved by about 4 times. The price is only 11000, our platform has more discounts, and the picture quality is obviously better than that of Sony a7m3.

The following picture is a dim portrait I shot with Panasonic S5, with a sensitivity of 2000. The legs of the characters are almost black in the original image.

After being brightened by ACR, the dark leg skin is brightened with little effect on skin tone and no too much noise. This is very difficult for ordinary cameras with single primary sensitivity to do.

3. The choice of a new type of image processor or dual-processor image processor is the last step to save the image quality. with the update of the processor, the high-sensitivity image quality of the camera is generally improved, and the machine with dual processor will be better than a single processor. For example, Nikon Z6 and Z6 second generation, the difference is that the second generation uses dual processors, so the high-sensitivity image quality and focus speed are improved, but the improvement is not obvious. Because this is not the step to suppress the noise, but the processing after the noise is generated.

4. when shooting, the principle of exposure to the right as much as possible is that you would rather have a high sense of normal exposure than a low sensitivity and underexposure and then brighten it. Let me explain:

Take a look at this work, the sensitivity is 5000, but the effect has no effect at all. The aperture is 1.8 and the shutter is 1max 60s.

With a sensitivity of 5000, I can't see that there is something wrong with the quality of the picture. On the contrary, it has a good texture. Actually, the reason is very simple.

In a dim light environment, the principle of shooting a single sheet is that it is better to have a high sense of normal exposure than to brighten up the underexposure of the sensitivity that is never too low.

How do you understand this sentence? For example, when the sensitivity is 5000, it can be exposed normally, so do not use 3200 underexposure and then brighten it in the later stage.

Why are you doing this? Because, when the sensitivity increases, the picture tolerance will also become worse, but the extent of lighting will be greatly reduced in the later stage.

When the high sensitivity is normal exposure, the only problem you need to solve is noise reduction.

However, when you are underexposed in order to reduce the sensitivity a little bit. In the later stage, we should not only solve the problem of sensitivity noise, but also produce new dark noise and the problem of picture quality caused by insufficient tolerance.

So beginners, do not think that the sensitivity can be a little lower is a little bit, would rather be underexposed, anyway, it will be brightened later. This is a wrong idea. In this way, you upgrade the problem of dealing with high-sensitive noise to the problem of noise + picture quality.

Looking at the picture above, the exposure of the picture is basically normal, there is no obvious dark part, the sensitivity is 5000, and the noise is completely within the controllable range. And this picture is without skin, it is all clothes, so it is less difficult in the later stage.

And look at the firefly figure with skin below.

When shooting, the sensitivity is 3200, so that the character part is exposed normally. Dim portrait theme, must not be underexposed and brightened, which has a great impact on the texture of the skin! So, find a way to make the skin exposure normal.

Of course, you should pay attention to the available sensitivity of your machine, which is the maximum sensitivity that you can successfully reduce noise in the later stage. Within this range of available sensitivity, try to expose normally. If you exceed the available sensitivity and cannot expose normally, you should consider other shooting methods.

This kind of scenery theme is easier to handle, because it is static, we can use the later stack noise reduction, shoot multiple scenes, and then average stack noise reduction, and then overlay the part of the Milky way.

This article is from the official account of Wechat: light and Color World of Shadow Leaf (ID:wangyikai-2016). Author: shadow Leaf.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.