4 . Masking and texture mixing channels

Masks are also called "masks," and can be said to be the best representation of the "channeling" channel.

Add a black and white grayscale image to an image (or a layer). The black part of the image will be hidden (not deleted) and become transparent; the white part will be fully visible; and the gray part will be in Translucent state. The mask has an irreplaceable function both in image synthesis and in special effects production. Masks can also be applied to textures in 3D models. Metal stains on the rust, glass decals, these irregularly shaped graphics are often treated with rectangular maps and masks. Masks of this type are often viewed as a special form of texture, called "transparency maps" because they need to adjust their coordinate position on the three-dimensional surface.

Masks can be used not only in simple textures but also in complex multidimensional materials. When two materials are staggered and mixed on the same surface, people also need to use channels to handle their distribution. Different from ordinary masks, such “mixed channels” are directly applied to two images: a black part shows an A image; a white part shows a B image; and a gray part has both. It can be seen that the hybrid channel derives from the concept of masking and is a simplified application for controlling the overlapping relationship between two images.

5. Displacement maps and bump maps

In 3D software, the channel is not limited to processing plane maps. He is also used to represent more complex materials and even to build models.

Imagine, we have to model a coin: the complex patterns and patterns on its surface will certainly cause a lot of trouble for our work. With the usual modeling methods, it can hardly be completed. Some people may ask: Can we use a flat image to represent the undulations of three-dimensional surfaces (like an elevation map), and let the computer automate tedious modeling work? The answer is: Yes, and this key plan is the channel.

A channel is attached to the surface of the object by means of a so-called "displacement map". At this point, the computer will work as follows: Apply the node on the surface of the map to the luminance information of the pixels on the map channel along the surface at the point. Different degrees of traction and stretching in the direction of the line are either recessed or convex: it is like matching a map depicting elevations with color, and pinching the mountains and valleys on the mud. Now, as long as we draw a two-dimensional image with a plane drawing tool, and then convert it into a displacement map and give it an object, a very textured coin will leap to the surface.

Although displacement maps can greatly reduce the amount of modeling work, the number of polygons is too numerous due to the lack of optimization of the resulting model. This can cause a dramatic increase in rendering time. For this reason, people have come up with a good way to compromise. Without increasing the complexity of the model, the bump effect on the surface of the object is approximated to the real model generated in the displacement map. This is the bump mapping algorithm. The bump map also uses the channel as the source of information. Through special surface mapping and light and shadow processing, the object's highlights and shadows are displayed, making the light and shadow effect reach a convincing degree in most cases.

Now, we use the bump map to create a coin model and compare it with the previous displacement model. We will find that when the “front view” of the plane being processed (line of sight perpendicular to the plane), the latter has the same outstanding performance as the former, but the rendering time is much better than the former; and in the “sideview” (the line of sight is perpendicular to the plane In the comparison of the normals), the displacement map still shows the true undulations of the surface of the object, while the plane treated with the bump map is flat. This also exposes the drawbacks of bump mapping. The above two algorithms have their own advantages and disadvantages. However, when deciding which algorithm to use to combine the channel and the model to achieve the desired effect, the perspective becomes a decisive factor.

6 . Vector channel

In order to reduce the amount of data, people will once again analyze the digital image that is plotted point by point, and use complex calculation methods to convert the point, line, surface, and color information on it into a simple mathematical formula; this type of formula is called "Vector graphics"; and the formulaized channels are called "vector channels." Although vector graphics can compress the amount of image information hundreds or thousands of times, its calculation method is too complicated and the conversion effect is often unsatisfactory. As a result, he can only use it when presenting concise outlines and clear-colored geometric figures; it rarely comes in handy when dealing with real effects such as photographs. The “paths” in Photoshop, several preset maps in 3DS, masks in illustrator, flash, and other vector drawing software all belong to this type of channel.

The ancients said: "Do not build a platform in the sand." If you want to have extraordinary skills, you must work hard to learn the basics. The application of the channel is the only way for the artists in the art industry to go from getting started to becoming proficient. It is also a colorful movement for this course. I hope that everyone can learn from this bit by bit, and in the near future, let his works exude dazzling light of art. (Text/Chen Yuxi)