How does PS fix a picture with serious missing details?

In the face of such a serious loss of details of the picture, how do we repair it?

Here are the results of our repairs:

So, how can we adjust to achieve this effect?

Here comes a very important term: gamma value.

In photography, we often encounter a similar set of concepts: exposure and gamma values.

Let's look at a picture first:

If we lower the exposure value:

The changes in the picture are:

If we lower the gamma value:

The effect of the screen is:

Obviously, you can see the difference between the two.

Let's look at a picture again:

If we increase the exposure value:

The effect of the screen is this:

If we increase the gamma value:

The resulting screen effect is:

You can also clearly find the difference between the two.

So, what is the difference between the two?

Let's start with the gamma value (the following is a boring paragraph that you don't want to know about):

At first, the gamma value was a proper noun, which was later widely used in the correction of image grayscale mapping. What we see on the macro level is that gamma has changed and the brightness of the image has changed, essentially the mapping relationship of information has changed mathematically. The following is the text:

Gamma is one of the least understood concepts in the field of computer graphics, and a large part of it is blamed for its presence everywhere, and it is extremely easy to confuse. The initial definition of gamma is as follows: "Gamma is used to describe the" non-linear "level of the display device's proper term. This is the first definition that Gamma has ever had in history. But it is clear that such a definition does not explain any of the problems.

So to explain Gamma, the first thing to explain is the "nonlinearity" of the display device.

All display devices, such as a computer display screen, a cell phone or tablet screen, an old-fashioned cathode-tube TV (CRT TV), or a new LCD-screen TV, are "non-linear." Take the most typical CRT (cathode ray tube cathode ray tube) to display the screen as an example (whether it's a computer or a TV), the so-called "non-linear" means that if the voltage intensity V of the cathode camera tube is increased by 1 time times, the intensity of the light output on the screen surface will not increase by one fold.

Conversely, if the light-emitting element used by the display is an ideal physical model, the ratio of V to I is a constant k in the input signal strength V (voltage, voltage) and the output light intensity I (intensity, strength) conversion process, Here we can say that the light-emitting element is a linear element, transformation is a linear transformation, is the "lossless transformation" of electrical signals.

To express a popular point, the above formula means: In the ideal linear display element, a voltage intensity can generate a corresponding light intensity, the voltage turns several times, the illumination intensity also can turn over several times accordingly.

Of course, no display in the real world would have such an ideal feature, fortunately, the corresponding relationship between the input voltage V of the lossy display elements and the output illumination intensity I is mathematically simple, and the input-output formula of the nonlinear display element is very similar to that of the linear display element. The only difference is that there is an exponential gamma in the upper-right corner of the input signal V, which reads gamma (the "gamma" in Chinese), which is the only parameter used to describe the "non-linear" degree of the input signal V of a nonlinear element in the process of converting to a light intensity I.

The expression of the popular point is that the non-linear display element presents such a rule: the input signal v size in the power of gamma and multiplied by the constant K, will get the corresponding output signal I, regardless of the input voltage size or not, a nonlinear element of the signal loss characteristics of gamma is fixed.

However, different display devices, gamma size is not the same, the common gamma size may be between 1.4 to 2.6, the Apple Computer monitor gamma in 1, 8, the traditional CRT TV's gamma in the 2.35-2.55. In conclusion, Gamma follows the hardware device and has nothing to do with voltage and light intensity.

So here we can finally understand why Gamma is defined as "a proper term used to describe the nonlinearity of a display device", an ideal display element that corresponds to a linear ratio of light intensity, gamma=1, and a non-linear display element in the real world, the larger the Gamma≠1,gamma, The greater the signal loss, the higher the "distortion" of the signal, so that gamma can be used to indicate the degree of distortion of a display component to the signal.

What must be emphasized here is: "Ideally, the signal lossless display element", does not say that in the voltage to the light intensity conversion process does not have the energy loss, but as long as the voltage converts to the corresponding light intensity by the k times, one voltage corresponds to a light strong can, equal proportion transforms, namely y=kx linear mapping, At this point, gamma is equal to 1.

and the nonlinear display element in the output light strong I need to put the input voltage V on the basis of K-fold, the upper right corner of a gamma again, then this time this gamma can be larger than 1, can also be smaller than 1.

In computer graphics, the input voltage signal V and the output brightness signal I are in 0-1 (this 0 corresponds to a voltage of 0, color is black, 1 corresponds to the maximum voltage, the highest light intensity, color white) interval floating, if this describes the problem, we can not even the middle of the formula of the constant k are negligible , the input-output formula is simplified to:

OK, the core meaning of this big paragraph above is (this summary can be looked at):

The so-called gamma value, originally used to describe the degree of hardware distortion, the greater the gamma value, the greater the degree of distortion.

Later applied to the field of computer graphics, it became a mathematical tool to correct the intensity of the picture.

OK, the above principle part we know about the line, do not need to delve into.

We mainly understand the difference and application of gamma value and exposure value in photography.

Let's open one more picture:

We reduce the exposure value:

The resulting screen effect is:

We lowered the gamma value:

Screen effect as shown:

What's the difference?

First, gamma value of the color changes more intense.

Second, gamma value details loss more.

Let's raise the exposure value again:

Get the picture effect:

We raise the Gamma value:

The resulting screen effect is:

You can see:

One, the exposure value retains more color information.

Second, the exposure value of the details of the loss of less.

Combined with the above effect, we can come to a simple conclusion:

Lower gamma value, the effect is equal to simultaneously reduce exposure value and improve contrast, therefore, its screen color changes more intense, the screen more transparent, but the details of the loss will be more.

To increase the gamma value, the effect is equivalent to simultaneously improve exposure and reduce the contrast, so the color will be more gray, but also will retain more details of the screen, the whole picture feels like a layer of white layer added.

Exposure value to the overall effect of the picture is softer, the effect of the adjustment generally will be more natural.

Classification:

• PS Image Processing