T& in hotel English; The meaning of l

T& in hotel English; L means that T&L technology is one of the specifications that can be seen on graphics accelerator cards recently. The original meaning of this term is: Transforming and Lighting, light and shadow conversion. In the graphics accelerator card, the biggest function of T&L is to deal with three-dimensional effects such as the overall angular rotation of graphics and the shadow of light source.

At present, most graphics accelerator cards are dealing with 3D stereoscopic images and other images. In order to make the images more realistic, the T&L function on the graphics card can accurately and clearly display all the points on the 3D object on the screen through angle calculation.

if we think of ourselves as graphics chips, then when the central processing unit (CPU) asks us to draw a complete 3D pattern, the CPU should give us a basic idea of what to draw, especially when those things don't have the most fundamental impression in our minds, and the CPU can only give us a design drawing similar to that used in construction sites.

There are many connection points on this 3D design, and each connection point can be regarded as a turning point of a line. Through these connection points, the graphics chip can clearly identify the structure of the whole object. However, the positions of these connecting points on this 3D design are fixed. In other words, no matter we look at the object from any angle, the positions of the connecting points have not changed, only the angles have changed. However, if the object pattern only goes to the Transforming stage, there are only connecting points on the surface of the object, and there is no color block or any material mapping performance, so it is very similar to a three-dimensional 3D perspective.

Just like engineers who draw construction site design drawings and painters who draw plane three-dimensional patterns, painters only draw the three-dimensional patterns they see in front of them, although they are also three-dimensional patterns, but the angles are fixed; Engineers, on the other hand, must carefully describe all the important details, so that the construction workers working according to the design drawings can know the construction direction no matter which direction they cut in. This blueprint sent by CPU is the same as the engineer's construction blueprint, which allows users to see the 3D changes of the whole object no matter from which angle, and this is the essence of Transforming.

If we really want to study what Transforming does in the graphics chip, it may be to control the above-mentioned link points. In order to convert these connecting points that constitute 3D objects into the angles we see from the screen, the Transforming mechanism in the graphics chip must "pretend to move" these connecting points. Therefore, even after pretending to move the connecting points, the original 3D object structure remains unchanged. As long as the angle on the screen changes, the Transforming mechanism will "pretend to move" again, which is quite frequent in the operation of 3D animation, so the Transforming function has attracted considerable attention for graphics chips in recent years.

a graphics chip with Transforming mechanism can effectively reduce the burden of the system except that it does not need to process data through the platform of the system (processor, main memory and chipset, etc.). Moreover, it also acts as an operation mechanism specially designed for processing the above data, which is more powerful than a system that needs to process multiple data at the same time. Therefore, the graphics chip can draw many more connection points than the processor alone, and the image presented by the structure is more detailed.

For the human brain, it is a bit vague to understand a pattern represented by connecting points, just like the game of "Lianliankan" when I was a child. However, through the action of connecting points, many polygons will be formed on the surface of 3D solid objects, and most of them are triangles. After that, the graphics chip will stick a material map on each polygon.

on the other hand, the meaning of Lighting is to respond to the projection angle of the light source. In addition to fixing the cut-in angle of the light source, we can calculate the change of the light source of the shadow and the reflective surface with the user's observation angle, and through these two functions, a fairly realistic 3D object can be presented to our eyes.

generally speaking, the performance of t&l is usually based on "how many polygons can be processed in a unit time and the light source operation". GeForce 2, like nVidia, can process 25 million polygons per second, and some products can even achieve the efficiency of 3 million polygons.

Vertex Skinning function: Most people should not see the word Vertex Skinning, because this specification is not listed in most graphics display card specifications, and usually only the polygon processing number in T&L is used as the main specification introduction. However, with the increasing of 3D animation software and materials, the importance of Vertex Skinning is also rising.

netizens who love to play video games may wish to recall that there must be many 3D action games such as monsters, heroes, sportsmen and role-playing in the video games at hand, and the 3D stereoscopic images of these game characters need to be fully displayed on the screen, which requires a more powerful computing mechanism through graphics chips.

One of the animation film-making techniques is to record the movement angle and joint changes of real people through clothes with human bone patterns. The same situation often happens in the operation of graphics chips, especially by using the concept of human bone movement to simulate the presentation of 3D objects in animation. In other words, if a complete 3D object finds out the main framework of its motion (human bones, for example), then the graphics chip can operate the outline and graphics wrapped outside the framework by the presentation of the framework, and calculate the image change relationship between actions by the motion mode.

However, if we really want to use the "human bone" mode to reverse the image of 3D objects, there will be many parts of data added to the image by the graphics chip itself, and this added technology is called Vertex Skinning. Vertex Skinning was first proposed by nVidia, but later other graphics chip manufacturers followed suit and adopted the same concept. Let's take this noun literally, that is, "skinning action of connecting points". In other words, the connection point and the backbone formed by the connection point often have a distance that is difficult to connect with the backbone of another part, and this distance will make the presented object unable to achieve a smooth and smooth real image. At this time, the Vertex Skinning mechanism in the graphics chip will automatically add the connection mechanism to make the whole picture more smooth.

generally speaking, a system or graphics chip needs a central processing unit to execute application software to achieve the above effect. however, it can also be achieved by using the computing mechanism (hardware computing) in the graphics chip. for example, ATI has indicated that its 4-Matrix Skinning is achieved by hardware computing.

Keyframe Interpolation Function

We first mentioned the imaging technology of raw meat forming on human bones. Now let's get to know how the graphics chip is realized when it can't be imaged through the backbone. And this technology, we often call it Keyframe Interpolation (key frame addition technology).

In fact, there are many places where 3D animation technology needs to use Keyframe Interpolation. When running, we only need to find out the starting action and ending action, and then make the rest of the linking actions by software or hardware under a certain change process. And this is the main spirit of Keyframe Interpolation.

For example, the facial expressions in character animation, no matter how hard you try to piece together the backbone of the face, you just can't describe expressions such as smile or frustration through the backbone structure. At this time, Keyframe Interpolation comes in handy. In fact, this is not difficult at all. You only need to let the system find the static patterns of two actions, such as the starting point and the ending point (these two pictures are the so-called Keyframe), and then add the main picture between these two pictures as necessary, so that the system can complete the change patterns of the rest details according to the direction you add. At this time, we can see the vivid character animation on the screen and deal with polygons

For graphics chips, polygon processing is one of the most important tasks. Including all the connecting points must be interpreted as polygons, and then the data is transmitted back to the graphics chip and then to the screen, or after the operation backbone, the polygon data after operation is returned to the operation flow of the graphics chip, so as to find the most perfect 3D animation. The work of T&L is almost always performed on these polygons one by one. Therefore, if we can understand the polygon processing, we can clearly understand the simple steps of image forming.

Generally speaking, most polygons on the screen or 3D objects are composed of triangles. Therefore, some graphics chips will explain the processing quantity of triangles instead of polygons. However, the more triangles you can handle, the better.

As we mentioned before, the blueprint drawn by the engineer must draw all the connecting points clearly, and such clear structural data is mainly to make the 3D object clearly presented and have basic data at any angle. However, because the screen will only show the image of the local angle of the object, many triangles or polygons in this complete 3D object will be covered by other triangles or polygons in front.

Let's take a 3D house as an example. The data of a 3D house needs to include all the furniture in the room and the 3D object information of the objects. However, if we can only look into the house from the side door, other 3D objects in the house will be hidden by the wall next to the door. Although the graphics chip can still calculate the complete transformation of all the data of 3D objects, the only thing that is really used (or seen by the user's eyes) is the screen.

Therefore, people are beginning to wonder: Since the image formation needs to be completed by many polygons, including the connecting points of polygons, these polygons and connecting points can't be displayed on the screen at some angles, will the transmission of such complete data in the computer drag down the processing efficiency of graphics chips? In order to maximize the performance of graphics chips, of course, many manufacturers have begun to pay attention to this topic and made many solutions. For example, ATI's Hyper-Z is one of the solutions to this problem. And we will introduce this relationship a little in the next few units.

the relationship between polygons and Fill Rate

another important topic is the relationship between triangle processing and so-called Fill Rate. To understand this topic, we must first know that in graphics chips, the mechanism for dealing with triangles is also called Pixel rendering ("pixel rendering" function), and this mechanism can only deal with one triangle or polygon at a time.

this is a very important concept, because Pixel rendering contains several pipelines to handle the same three shapes. If there are fewer pixels in a triangle or polygon than the number of pipelines, then several pipelines in this Pixel Rendering will be idle.

for a 3D stereoscopic image, the more connected points it forms, the more triangles or polygons it forms. And these triangles will become more and more triangles with pixels smaller than the total number of pipeline in Pixel Rendering because the distance between connecting points is getting closer and closer. This situation is a serious loss to the performance of graphics chips. As a result, the pattern with many small triangles will make the graphics chip with high Fill Rate specification unable to show its real performance. This is why graphics chips with high Fill Rate must be matched with correct display settings in order to truly show the performance of graphics chips.

hardware acceleration t&; L

T& L is transform &; The abbreviation of Lighting means coordinate conversion module and light source module. Have t&; The drawing chip with L function is only used in 3D computer workstations and high-end desktop computers. But from the development direction of graphic processing technology now, T&; L's hardware acceleration function will surely become an essential feature of future drawing chips.

T& L engine is mainly used for complex coordinate processing and light source image calculation, so that users can feel the real light and shadow of objects. In the past, there was no T&; On the platform of L engine, most of the coordinate processing work and light and shadow special effects need to be performed by the CPU, so it takes up too much computing time of the CPU, which causes the overall picture not to be displayed very smoothly. If t&; L engine will greatly reduce the CPU's load when processing 3D, and enable the CPU to have more resources to handle more wonderful 3D special effects and provide better visual effects. Therefore, at present, the 3D graphics computing power of the drawing chip has surpassed that of the CPU. The simple architecture of 3D drawing pipeline is shown in the figure.

The dynamic images you see in the game are actually composed of a series of static images that are displayed quickly (the basic requirement is 3 images per second). The more static images that the computer can display, the smoother the scene will be. The computer must go through a series of steps to draw a static picture, including Transform, Lighting, Triangle Setup, Rendering and so on. 3D application software (such as 3D games) will provide every action during the game, including the movement of the camera, the relative actions of other objects, the change of the fine level, the calculation of the physical engine and other changing factors. Then these data are sent to 3D drawing pipeline through API (such as DirectX or OpenGL), processed and displayed on the screen, while T&; L refers to coordinate transformation and projection.

coordinate transformation module