Guide to graphic display settings for game players (4) (the gamer's graphics & Display Settings Guide)

Guide to graphic display settings for game players (4)

Translated by koroush Ghazi

Level 4 graphic processing (part 3)

Step 7 GPU processing data

Now the GPU can access data in VRAM, And it undertakes a series of complex graphic computing tasks in a place usually called graphics pipeline. This "Pipeline" has many stages (stages) which are used to constantly convert pure 3D data into visible things on a 2-dimensional display screen:

1. transformation: Starting from vertex data, a vertex is a point in 3D space. Generally, some vertices can be combined to form a wireframe of a 3D image ), these 3D images are called polygon ). In the first phase of the pipeline, the pure 3D vertex data will be converted to an image vertex format to create all 3D objects, and place these 3D pairs correctly in 2D space-that is, your monitor. Everything will be correctly scaled and positioned based on the relationships between them. In this case, you can use the vertex shader to apply some special effects to the vertex, such as a warping object.

2. lighting-after conversion, all objects are illuminated based on the position of the light source and the reflected surface of the object. The illumination-related data is still in vertex format, so it is basically just an application of a large number of points and lighting colors of different colors.

3. viewpoint-then you will observe these objects from a specific angle and distance, and you will find that these objects will be further adjusted. That is to say, those objects closer to the observer will be larger, while those farther away will be smaller, and each object will be properly rotated and moved.

4. Clipping-objects that cannot be viewed by the observer due to the limitation of the current field of view (FOV) are deleted. For example, if your field of view is 90 degrees (most games use this value), the objects in the remaining 270 degrees around you will be removed and you will not be able to see them. This is done to reduce the useless computing workload of the GPU. If this is not done, the objects you cannot see will also be computed by the GPU.

5. Triangle setup-at this time, all vertices will form a triangle or polygon. This is the basic 3D image on the computer screen that we are familiar.

6. Rasterization-pixels are the smallest graphic information unit on your display. Because everything in a 3D scenario will eventually be displayed in a two-dimensional plane as pixels, the objects in the scenario will be allocated pixels at a specific position to be displayed on the screen.

7. culling-this process is a supplement to clipping. pixels that are still invisible will also be removed: for example, the pixels from which the object moves back to your side, objects blocked by other objects will be removed, because they are invisible to the current observer, so it is not worth wasting additional time on them.

8. Texture & shading-at this stage, each pixel of the image will be applied with a variety of colors or effects. A texture is a 2D image that is added to the surface of a 3D object to make it look more realistic. The pixel shader can combine a series of colors, depths, and textures to create special effects on relevant pixels, for example, fold, smooth, or transparent effects on the surface of an object.

Performance prompt: Currently, all video cards can operate multiple pipelines at the same time, and this "multithreading" allows the video card to share the work of the above stages in parallel, so that the execution can be faster. Obviously, the more pipelines a video card has, the higher the execution efficiency. This is why the current video card can smoothly render real 3D images. In the past, rendering complex 3D scenes (such as movie effects) was slow, sometimes taking hours or even days, but now rendering may occur many times in a second. Games are computation-sensitive tasks on home computers, so do not underestimate how much pressure your computer is under when you are playing a complex 3D game. This is not just a video card. The more modern the game is, the more complicated it is, the more unstable it will bring to your computer, causing system crashes and deadlocks.

 

Step 8 transfer data to the frame Cache

 

 

 

The data obtained from the above steps is usually called a separate frame and will be transferred to a temporary storage area in VRAM. This is called frame buffer ). As the name suggests, real frames are cached here, and these frames will be transmitted to the computer display for display.

Two images are cached at the same time. This process is called double buffering ). A complete frame is saved in the master (Front) cache, and another complete frame is being formed in the secondary (Backup) cache. Frames cached by the master node are displayed on the monitor at any time. When the content in the master cache is displayed in the monitor, the video card is preparing another frame in the cache, as long as your monitor is ready to refresh itself, frames in the cache will be displayed as "exchanged", and the master cache (which has already become a secondary cache) clears itself and assembles another frame. This page switching technology can help solve potential bottlenecks and be applied in 3DProgramQuick frame display improves performance.

The higher the Resolution, the more complex the image is, and the larger the data size, the more space occupied by each frame cache. If you add additional effects to this frame, such as smoothing the Antialiasing with a sharp tooth shape, or clearing the heterosexual filter from the texture, the image (FRAME) will be added).

Performance prompt: Generally, only two frames are saved in the frame cache. However, triple buffering can provide a third or third (Teriary) frame in the frame cache. When vertical synchronization (vsync) is enabled, the performance will be improved, while waiting for the display to signal it, it is ready to accept the new frame, it gives the video card more space to operate the complete frame. However, this will also use more VRAM to increase the "card" and "pause" phenomenon (stuttering and hitching) of the screen ). In addition, this increases control latency and graphic flickering-For more details, see the triple buffering chapter.

 

Step 9 transfer the image to the screen

 

 

 

Images are now ready to be displayed, and they are accessed from the frame buffer via random memory to the analog-to-analog converter (ramdac, random access memory digital to analog convertor) to the video cable, finally, it is sent to your monitor. Here ramdac is required to convert the numerical information (0 and 1) in your VRAM to the analog format (video signal voltage) required by most monitors ). Ramdac can often convert frames into visible images multiple times within one second.

Ramdac is important because it affects the maximum possible update rate and image quality on the screen at popular resolutions (which is related to the display capability of the Monitor. Fortunately, the ramdac of most graphics cards is good enough, so they are not the focus of our attention-but the refresh rate of the display itself is usually the limiting factor rather than ramdac.

Importantly, traditional CRT monitors rely on ramdac, but if you use a pure digital connection such as DVI (Digital Video Interface) between your video card and the monitor, your monitor no longer depends on ramdac. When DVI is used as the display input, the result is that ramdac is ignored, because the original digital signal is transmitted to your LCD screen intact, however, if a standard VGA link is used on the LCD, the digital signal needs to be converted to analog (through ramdac) and then converted back before the LCD display, in this way, it is no longer efficient.

 

Performance prompt:This step emphasizes that the LCD display will provide the best image quality when using DVI connections, because there is no possibility of data downgrading during repeated conversions. However, there are still many reasons to connect the LCD display in analog mode-see the response time chapter. If you have a CRT Display and an old-fashioned video card, but your image quality does not reach its best, it may be that the ramdac capability on the old-fashioned graphics card is not good enough. In the old-style graphics card, it is particularly likely to produce blurred 2D images, such as Windows desktop blur and high-resolution 3D quality reduction. These are definitely not good for your eyes.

 

This chapter covers how 3D images work on your system. We hope that you can better understand how your system works when playing 3D games, so you can arm yourself in solving problems and improving performance. Remember that my game and System Configuration Guide contains a lot of performance prompts, such as tweakguides tweaking companion. Therefore, please read the suggestions from the above chapters.

The next chapter focuses on the graphics settings that gamers often need. An accurate understanding of how they work will once again strengthen the problem solving and performance optimization capabilities.

 

Previous

Next article