空影

Today, after reading these pages of the book, I finally understand that beauty has standards that can be followed. It seems that my previous understanding of photos was extremely shallow!

Concept of Histogram#

First, let me give my own understanding: the so-called histogram is a statistical chart drawn based on the brightness and quantity of different samples taken from different positions in the image.

• Explanation of histogram by the official Photoshop tutorial:
  The histogram¹ represents the number of pixels in each brightness level of the image, showing the distribution of pixels in the image. The histogram displays details in the shadows (shown in the left part of the histogram), midtones (shown in the middle), and highlights (shown in the right part).

• Explanation on Wikipedia:
  Image histogram² is a histogram used to represent the brightness distribution in a digital image, plotting the number of pixels for each brightness value in the image. By observing this histogram, you can understand how to adjust the brightness distribution. In this histogram, the left side of the horizontal axis represents the pure black and darker areas, while the right side represents the brighter and pure white areas. Therefore, in the histogram of a dark image, the data is mostly concentrated on the left and middle parts; while in an overall bright image with only a small amount of shadow, it is the opposite.

Function#

Histogram plays a crucial role in judging the exposure and contrast of photos³. Why is that? Because whether it is the human eye or a $\color{purple}4^{4^{4^4}}$K monitor, the interpretation of the same photo is different. The simplest example is to develop a photo, and you will easily notice the color differences between the screen and the developed photo. Seeing without understanding—this is the limitation and sadness of all living beings and technologies. But don't be discouraged, there are ways to "see without understanding", and the histogram can give us the ability to overcome visual limitations: explaining with data.

Evaluation Criteria#

The brightness⁴ of an image can be divided into 0-255 levels. A perfectly exposed image with a full tonal range should have distribution in all brightness ranges. In other words, the histogram should not hit the walls—consider the far left and far right of the histogram as "walls", and the pixels should be distributed between the two "walls". They should touch the edges of the "walls" without overflowing. If the pixels hit the left "wall" and overflow, it means that the details in the shadows are lost; if the pixels hit the right "wall" and overflow, it means that the details in the highlights are lost.

If we adjust the shadows or highlights of the histogram to "hit the wall and rise", it means that the dark or bright details of the photo are severely lost, and it does not belong to a full tonal range photo and does not meet the requirements of high quality. Pursuing the transparency of the photo too much, without mastering the correct methods and without referring to the histogram, adjusting a series of parameters such as "curves" and "levels" will lead to this phenomenon. Therefore, any photo adjustment should be based on the histogram, make reasonable evaluations and judgments guided by the histogram, and control the histogram properly using reasonable techniques.

Unfortunately, according to this criterion, I looked at many of my previous photos and did not find one that met the requirements of perfect exposure. Either there was no distribution in the shadows or highlights, or it hit the walls. I am really a little trash.

Another point to mention: the height or low of the peaks in the histogram is not an important reference. The height or low of a certain area of the peaks can only represent the abundance or scarcity of brightness pixel distribution in that area. If the peak in the highlight area of a photo is low, it means that there are not many bright areas in the photo, and the proportion of highlight areas in the photo is not large, so the numerical value in the highlight area of the histogram is not high, but it does not affect the details and levels in the photo.

Failed Tone Cases#

Now let's "criticize" myself. The software used in this article is the open-source DigiKam image management software (similar to LightRoom) to view the image histogram.

Insufficient Contrast#

This is a photo taken in foggy weather. From the histogram, we can see that there is no distribution of pixels in the shadows and highlights, which means it lacks contrast.
Scenarios: Hazy weather, cloudy weather, low contrast environments
Solution: Can only be adjusted in post-processing

Excessive Contrast#

In this photo, the shadows are fine, but the highlights hit the wall and rise, which means that the brightness between the leaves on the left is too high.

Because I wanted to capture the silhouette effect of the house and signal tower during the sunset, I lowered the exposure of the camera. As you can see, both the highlights and shadows hit the wall and rise, which means that the details in both the highlights and shadows are partially lost. From the photo, the details of the dark areas of the house are gone (which was my intention), and the roof clouds are slightly "overexposed" and some details are lost.

Scenarios: Indoor to outdoor, shooting in strong light
Solution: Shoot in RAW format, take multiple photos with different exposure levels for HDR synthesis
Advice: Pay attention to the histogram during shooting. Digital photography exposure follows the principle of "better overexposure than underexposure" for easier adjustment in post-processing without losing too much detail.

Special Tone Cases#

For common tone photos, "hitting the wall without rising" is a good criterion for judgment, but for some photos with special effects, their histograms are different and cannot be generalized.

Low-key Photos (mostly dark tones)#

High-key Photos (mostly light tones)#