Exploring DSC: A Comprehensive Guide to High Refresh Rate Technology and Its Applications
AnnChouHave you ever noticed while playing Counter-Strike: Global Offensive that sometimes, when moving your character and turning to shoot, the screen freezes for half a second or even longer, disrupting what should be smooth gameplay?
Or while playing the campaign mode in Call of Duty: Modern Warfare 2, have you encountered moments where the screen splits into two halves during intense combat scenes?
The problem is resolved!
But what's the reason behind this?
It actually comes down to the limitation of data transmission bandwidth. Imagine the data transfer between your computer and the monitor as a water pipe system, where the data is water, and the bandwidth is the width of the pipe.
When you need to quickly fill a container (the monitor) with water (transmit a large amount of image data), if the pipe is wide (sufficient bandwidth), the water (data) can flow smoothly and quickly into the container (the monitor), and the container will be filled on time (the screen updates normally).
However, if the pipe is narrow (insufficient bandwidth), and you're trying to fill it quickly with a large amount of water (high resolution, high refresh rate data demands), problems arise. The water flows slowly through the pipe (data transfer is slow), and the container can't be filled in time (screen update delay, stuttering). Additionally, since the container (monitor) has a small capacity (low resolution, low refresh rate), the filling process might cause backflow or splashing (errors or screen tearing), resulting in an unstable, incomplete, or incorrect state in the container (monitor), which leads to display issues like misaligned images during screen tearing, or color distortion when the color is off. All of these problems occur because the data transmission bandwidth is insufficient, preventing the stable and accurate transfer of data to the monitor.
How can we solve this problem?
This is where DSC (Display Stream Compression) technology comes in. Let's break it down in simple terms to understand what DSC is and how it can enhance our experience.
I. DSC Technology Overview
(1) What is DSC?
DSC, or Display Stream Compression, is a video compression algorithm developed by the Video Electronics Standards Association (VESA). With the continuous advancement of display technology, including increasing resolutions and frame rates, existing physical interfaces often face bandwidth limitations when transmitting high-resolution, high-frame-rate video data. The introduction of DSC technology addresses this challenge by enabling the transmission of higher display resolutions and frame rates over current physical interfaces (such as DP 1.4 and above, HDMI 2.1, and embedded DP (eDP) 1.4 and above).
DSC achieves impressive compression ratios, reducing images to 8 bits per pixel (bpp). For common 24 bpp images, the compression ratio can reach 3x; for 30 bpp images, the compression ratio is even higher, at 3.75x. We will explore the core factors behind DSC’s excellent compression performance as we analyze its principles.
(2) The Development of DSC
The development of DSC marks its rise in the field of display technology. In 2014, DSC made its debut, and by 2016, it was officially incorporated into the DisplayPort 1.4 standard, laying a solid foundation for its widespread use in computer displays. In 2017, HDMI 2.1 also adopted DSC, further expanding its application. Since then, DSC has continued to evolve, playing a key role in subsequent versions of DisplayPort, such as 2.0 and 2.1, and gradually becoming an essential component of modern display technology.
(3) How Does DSC Work?
We can use the analogy of a parcel delivery process to explain how DSC works. Imagine you need to send a lot of items (image data) from one place to another (from the source device to the monitor). If you don’t do anything to optimize the process and just throw everything into a big box (uncompressed data transfer), this box will likely be very large, heavy, and possibly even too big to fit, making the transportation difficult and time-consuming (taking up excessive bandwidth and slow transfer speeds).
DSC acts like a super-smart courier packer. It first carefully examines the items (analyzing the image data) and notices that some items are very similar or even identical (like adjacent pixels or similar areas in the image, such as color, brightness, etc.). Then, it uses a series of specialized packing methods to reduce the size and weight of the items (compressing the data) while still ensuring everything can be safely and efficiently delivered (transmitted with minimal loss).
-
Preprocessing Stage
During the preprocessing stage, the data undergoes various conversions depending on the encoding type and pixel format to make it more suitable for subsequent compression. If the encoding used is RGB, it will first be converted to the reversible YCGCO format. This is because YCGCO-encoded data is easier to manipulate and compress in the later stages of the compression algorithm.
For pixel formats like "simple 4:2:2," interpolation is used to fill in missing chroma samples by estimating the chroma values of neighboring pixels, converting the format to 4:4:4. For example, by analyzing the chroma information of surrounding pixels, the algorithm can reasonably guess and complete the missing chroma values for the current pixel, making the overall pixel data more structured and conducive to efficient compression.
-
Encoding and Prediction Step
The second step is encoding prediction, which can be simplified as follows: the device first uses the information from surrounding pixels to predict the target pixel values. Then, the predicted pixel values are compared with the original pixel values to calculate an error value. This error value is transmitted to the receiver, which uses the same method to predict the target pixels and adds the error value to get closer to the original pixel values. In this way, only the error values need to be transmitted, reducing the overall data volume.
DSC employs several techniques for encoding prediction, including Median-Matching Adaptive Prediction (MMAP), Block Prediction (BP), and Midpoint Prediction (MP).
When choosing a prediction method, DSC bases its decision on the composition of the image pixels. It's similar to packing a box for shipping—when deciding on the best packing strategy, we consider the type, quantity, and distribution of items in the box (the pixel composition of the image). For example, if the box mainly contains small items of the same type (a region in the image with high pixel similarity), the MMAP method may be chosen as it is better suited for handling such items. If the box contains items arranged in a certain pattern (pixels with specific distribution), the BP method may be more appropriate. If the items fall between two common states (pixel values that lie between two reference points), the MP method might be selected. The goal is always to minimize the space occupied by the packed box (minimizing the encoding error).
-
Bitrate Control and Buffering Mechanism
The bitrate control algorithm in DSC plays a critical role by tracking two key factors: color uniformity and buffer fullness. For example, when large areas of the image have uniform or flat colors, this indicates that the bit depth can be lowered to further compress the data. On the other hand, when the buffer is nearing capacity, the quantization bit depth of the pixel groups must be adjusted to ensure that data is processed in time without overflowing the buffer.
With this bitrate control algorithm, DSC dynamically adjusts the quantization bit depth of pixel groups within the set bitrate range to minimize artifacts that can arise during compression. This process is similar to the packing process when shipping: we observe the distribution of items in the box (color uniformity) and the remaining space in the box (buffer fullness). If the items in the box are arranged neatly and there is plenty of remaining space (areas with smooth color transitions), we can pack the items more tightly (lower quantization bit depth), as they are less likely to be deformed (simple color information that doesn’t need high precision). However, if the items are irregularly shaped and there is limited space left (areas with rich color variation), we need to be more careful with the packing strategy (adjusting the quantization bit depth appropriately) to ensure that everything fits (within bandwidth limits) and minimizes damage from compression (reducing artifacts).
By carefully balancing data compression and visual fidelity, this approach ensures that the image maintains its visual quality while being effectively compressed, meeting the "visually lossless" requirement. This means users are unlikely to notice any perceptible compression artifacts while viewing the image.
-
Indexed Color History (ICH) and Entropy Encoding Applications
The Indexed Color History (ICH) is a clever design in DSC that enhances compression efficiency. The idea behind ICH is to store frequently occurring pixel information from the image as an index. For example, in certain graphical scenes, such as UI elements in games or backgrounds with large areas of uniform color, these repeating pixels can be stored in a 32-entry Indexed Color History (ICH) buffer. During transmission, there’s no need to repeatedly send the detailed information of these pixels; instead, only the index data is transmitted, significantly reducing the amount of data transmitted and improving overall compression efficiency.
This is similar to the packing process in shipping, where we have small groups of items that often appear together (frequently occurring pixel information), such as a set of stationery (pencils, erasers, rulers, etc.). We can place this set into a small box and label it (the index). When packing the larger box, we just need to find the label on the small box to identify the items inside, rather than unpacking and labeling each item individually (avoiding the need to transmit large amounts of pixel data). This approach saves a lot of space and labeling work (reducing data volume).
-
Image Slicing and Tile Composition
To better compress data and facilitate subsequent transmission, DSC divides the entire image into slices of different heights and widths. There are various common slicing combinations, such as dividing based on image width, which may result in 100% or 25% of the image width, and slicing based on height, with different configurations like 8 rows, 32 rows, or 108 rows.
By slicing the image into smaller sections, each slice can be encoded simultaneously. This slicing mechanism allows the encoding resources to be more flexibly allocated based on the characteristics of the image content, optimizing compression efficiency.
Dividing the image into slices is similar to packing a large number of items (image data) for shipping. If everything is thrown together without sorting (no slicing), it can lead to a disorganized, hard-to-manage, and poorly protected package (which is inefficient for compression and transmission). However, cutting the image into slices is like categorizing and packaging the items into smaller parcels (slices). Each small parcel is easier to handle. For example, when packing a large box with many small items, if you separate them into several smaller packages, the packing process becomes easier, and the risk of items being crushed or damaged during transport is reduced (lowering the chance of data loss or errors). In image transmission, slicing makes each part of the data easier to manage and optimize, improving overall compression and transmission efficiency.
At this point, we can see that by significantly compressing data at each stage, DSC achieves excellent compression efficiency.
II. DSC Applications
(1) Questions about DSC
-
Does DSC introduce any latency?
The processes of compression, transmission, and decompression all take some time. Although these algorithms are optimized, they inevitably introduce a small amount of latency. The overall delay added by DSC is only about 0.5 microseconds, which is extremely low compared to other video processing technologies.
-
What is "visually lossless"?
Although DSC is not mathematically a strict lossless compression method, it achieves a level of visual quality comparable to lossless compression. "Visually lossless" means that, for an average observer, the image or video compressed with DSC appears nearly identical to the original uncompressed version in terms of visual perception, with almost no noticeable differences. This is due to the limitations of the human eye in perceiving minute image details and changes. Under normal viewing conditions, the human eye struggles to distinguish very small color differences, brightness changes, or subtle alterations in the image structure. For example, extremely fine color gradients between adjacent pixels may be imperceptible to the human eye. DSC technology takes advantage of this characteristic by selectively processing information that is less sensitive to human vision, thereby reducing the data size without affecting the overall visual quality.
-
What kind of display performance can I achieve with DSC?
HDMI 2.1: In its native non-DSC mode, HDMI 2.1 supports 4K resolution at 144Hz refresh rate and 8K resolution at 30Hz refresh rate. With DSC technology, it can support up to 10K resolution at 120Hz refresh rate. However, it's important to note that the DSC feature in HDMI 2.1 can only be used in FRL (Fixed Rate Link) transmission mode, and it is not a default feature—it is determined by the device manufacturer.
DisplayPort 1.4: When using a single DP 1.4 cable connected to a monitor that supports DSC compression, you can achieve the following display performances:
-
4K at 120Hz (lossless)
-
4K at 144Hz (DSC/4:2:2)
-
4K at 240Hz (DSC/4:2:0)
-
8K at 30Hz (lossless)
-
8K at 60Hz (DSC/4:2:2)
-
8K at 120Hz (DSC/4:2:0)
With multi-stream support, DisplayPort 1.4 can drive 4 independent displays at 4K 120Hz or, using DSC, 2 independent displays at 4K 240Hz or 8K 60Hz.
DisplayPort 2.0/2.1: Without DSC, DisplayPort 2.0/2.1 supports up to 4K at 240Hz, 8K at 60Hz, and 2K at 500Hz (UHBR20), with 10-bit color depth. With DSC, it can support up to 16K resolution at 60Hz.
Resolution support under different interface versions.
Of course, it’s important to note that the actual display performance may also be affected by factors such as the specific capabilities of the graphics card, the monitor, and the quality of the transmission cables.
(2) DSC Support for Different Devices
-
Computer Graphics Cards and DSC
To check if your graphics card supports DSC, you can either refer to the official technical specifications or contact the manufacturer's customer service for confirmation. Typically, when the graphics card detects that the connected monitor supports DSC, the feature will be automatically enabled. For example, some NVIDIA graphics cards also provide related options in the driver control panel, where you can check if DSC is enabled. If DSC mode is active, there will be a corresponding indication in the 3D settings management section.
When purchasing a new graphics card, if you aim to use DSC for high-refresh-rate gaming display output, you should pay attention to the display interface standards supported by the card and the compatibility with corresponding DSC versions. For instance, if you want to achieve 4K at 144Hz or even higher resolutions and refresh rates, make sure the graphics card is compatible with interfaces that support DSC, such as DisplayPort 1.4 or higher versions, or HDMI 2.1. Additionally, the card's performance should be powerful enough to meet the rendering demands for gaming visuals at the desired resolution and refresh rate, to avoid issues like stuttering or performance bottlenecks due to insufficient GPU power.
-
Gaming Consoles and DSC
Support for DSC varies among mainstream gaming consoles. For example, the PS5 has an HDMI 2.1 port that supports a limited 32Gbps bandwidth rather than the full 48Gbps, and the manufacturer has not explicitly stated support for DSC. This somewhat limits its ability to enhance high-refresh-rate, high-resolution video transmission using DSC technology.
The Xbox Series X performs well in terms of resolution upgrades. Although it doesn't explicitly mention DSC support, it is capable of outputting up to 4K at 120Hz with 4:4:4 16-bit color. When connected to a display that supports DSC, it can potentially leverage the monitor's DSC feature to further optimize the visual output.
As for the Nintendo Switch, its hardware performance in terms of display is not as high-end, and its use cases for DSC are relatively limited.
-
Monitors and DSC
In the monitor sector, many brands and models of high-refresh-rate displays use DSC technology to simultaneously achieve high resolution and high refresh rates. The activation of DSC on the monitor typically requires the monitor itself to support this technology, as well as the connected signal source (such as a computer graphics card or gaming console) to also support DSC and be connected through the appropriate interface. After the devices complete identification and handshake processes, the DSC feature will be automatically enabled, providing players with practical visual advantages such as smoother gameplay and more accurate color reproduction.
(3) Current Issues with DSC
When using NVIDIA's 40-series graphics cards with DSC-supporting monitors, one common compatibility issue is the occurrence of black screen problems. Taking the Samsung Odyssey Neo G9 monitor as an example, when users connect it to an NVIDIA RTX 4090 graphics card, they may experience a black screen during tasks like switching game windows (e.g., using the Alt+Tab shortcut). The black screen duration can range from a few seconds to over ten seconds, significantly affecting the smoothness of user operations and the overall experience. This issue does not occur every time a window is switched, but its frequency can be quite high, causing considerable inconvenience to users.
In response to these issues, NVIDIA has released multiple driver updates in attempts to resolve the problem. Many players have resorted to playing games in borderless windowed mode or manually lowering the resolution to disable DSC in order to reduce the occurrence of the black screen issue.
With the upcoming release of NVIDIA's 50-series graphics cards, there is hope for improvements. From a hardware specification standpoint, the new cards are expected to support more advanced interface technologies, such as potentially offering native support for DisplayPort 2.1. DP 2.1 offers an impressive bandwidth of up to 80Gbps, representing a significant leap compared to DP 1.4. This will greatly enhance data transfer capabilities. However, we cannot overlook the impact of software. NVIDIA has continuously worked on optimizing its graphics card drivers, and with the release of the 50-series, these drivers will likely undergo deep optimizations for DSC technology and the new card features. Ultimately, we will have to wait and see what the results will be.
