H.265 development background
Although H.264 is an epoch-making digital video compression standard, with the rapid development of the digital video industry chain, the limitations of H.264 have gradually emerged, and due to the complete solidification of the core compression algorithm of the H.264 standard, it cannot pass Adjust or expand to better meet current high-definition digital video applications.
The trend of video application development in the following aspects is becoming increasingly obvious:
High definition: The application format of digital video has been fully upgraded from 720P to 1080P. In some video application fields, 4K×2K and 8K×4K digital video formats have even appeared;
High frame rate: The digital video frame rate is upgraded from 30fps to 60fps, 120fps or even 240fps application scenarios;
High compression rate: Transmission bandwidth and storage space have always been the most critical resources in video applications. Therefore, obtaining the best video experience in limited space and pipelines has always been the unremitting pursuit of users.
Since digital video applications are facing the above trends in development, if H.264 encoding continues to be used, the following limitations will occur:
The explosive growth in the number of macroblocks will cause too many codewords to be occupied by macroblock-level parameter information such as prediction modes, motion vectors, reference frame indexes, and quantization levels for encoding macroblocks. The code words are significantly reduced. That is: the information of the image content represented by a single macroblock is greatly reduced, resulting in a greatly increased low-frequency similarity after 4×4 or 8×8 block transformation, and a large amount of redundancy will occur;
With the substantial increase in resolution, the amplitude of the motion vector representing the same motion will be greatly increased. A motion vector prediction value is used in H.264, and Columbus index coding is used to encode the motion vector difference. The characteristic of this coding method is numerical value The smaller it is, the fewer bits it uses. Therefore, as the amplitude of motion vectors increases significantly, the compression rate of the method used to predict and encode motion vectors in H.264 will gradually decrease;
The degree of parallelism is relatively low: Some key algorithms of H.264, such as the use of two context-based entropy coding methods, CAVLC and CABAC, and deblock filtering, require serial encoding and the degree of parallelism is relatively low. For CPUs with very parallel programs such as GPU, DSP, FPGA, ASIC, etc., the serial processing of H.264 has increasingly become a bottleneck restricting computing performance.
H.265 discussion
H.265 is also known as High Efficiency Video Coding or HEVC. It is the abbreviation of HighEfficiencyVideoCoding. It is the latest standard in a series of video compression standards. Like its predecessor H.264, H.265 was originally developed for the broadcast industry by the ITU-T Video Coding Expert Group and the ISO/IEC Moving Pictures Expert Group.
H.265 is a video compression codec and the successor to H.264. It's built on a similar concept to its predecessor, but is becoming commonplace thanks to the rapid adoption of 4K content everywhere. Without compromising video quality, H.265 allows video compression at half the bitrate - H.264 - making it theoretically twice as efficient. When we compressed H.265 to the same bitrate as H.264, we found that H.265 provided significantly improved video quality. Thanks to all these advantages and upgrades, H.265 makes it easier to stream and download 4K videos - something that was not possible in H.264 because it takes up a lot of space and has a high bitrate.
How the h.265 codec works
The H.265 codec is based on the same general ideas and structure as the previous H.264 standard. As before, it has a source video consisting of a sequence of frames encoded (or compressed) by an encoder - this results in a compressed bitstream. This compressed bitstream is stored or shared, and a decoder on the other end decompresses the bitstream to create the original sequence of frames. In terms of H.265 codec working, the steps performed by the encoder include:
- The first step is about predicting each unit and performing the necessary calculations (i.e. subtracting it from the original picture unit). In this step, the encoder receives a frame of video units (macroblocks) displaying 16x16 pixels. This macroblock forms a prediction in two ways - from the current frame (intra-prediction) or from a previously transmitted frame (inter-prediction).
- Quantize and transform residuals. Residuals are the differences between the prediction and the original picture units. This is accomplished by outputting a set of coefficients using an approximate form of the discrete cosine transform, each coefficient being related to a weight value of the underlying mode. These base patterns can later be combined together to create the initial residuals.
- The converted output, mode information, prediction information, and headers are entropy encoded (i.e., a lossless data compression scheme is performed ).
H.265 decoder, on the other hand, performs the following steps:
- Entropy decoding - that is, reversing the steps performed during entropy encoding - and extracting the original elements from the encoded sequence.
- Invert transformation and rescale.
- Predict each unit and add it to the output of the inverse transform
- Reconstruct the final decoded video image.
Applications of h.265
The H.265 codec uses a different macroblock encoding method than H.264, called a Coding Tree Unit (CTU). CTU processes information with higher coding efficiency and supports 64x64 macroblocks. This makes the H.265 format very useful in many applications. Some such applications include:
- H.265 supports a wide color gamut - such as NTSC, Rec.601, PAL, Universal Film, SMPTE170M, sRGN, SYCC and more.
- Provides seamless video streaming and data sharing for next-generation HDTV displays and content capture systems.
Why use h.265 compression
The simple answer to this question is to be more efficient. The less bitstream a codec has available to compress and decompress an image without sacrificing its quality, the more efficient it is. From this perspective, H.265 is more efficient than the H.264 codec. Learn more about H.264 vs. H.265. Beyond that, as things have evolved, so have the resolutions people are viewing at, and the resolutions of the screens people are using to watch videos. With so many constraints and modifications, it makes sense to move to compression techniques designed for modern video requirements.
Advantages of h.265
To summarize, here are the main advantages and improvements provided by the H.265 compression standard:
- Provides half the compression ratio of H.264.
- Supports 64x64 pixel macro blocks, while H.264 supports 16x16 pixel macro blocks.
- Video compression depends on predicting motion between frames, and H.265 has a better motion prediction mechanism. This results in improved compression standards.
- Inter prediction is more detailed in H.265 than in H.264.
- The resolution supported by H.265 is much higher than the 8K Ultra HD supported by H.264.
- H.265 has a much lower bitrate compared to the H.264 codec, making the entire process more efficient.
The full text mainly analyzes H.265 encoding in audio and video development. For in-depth study of advanced audio and video first-level H265 technology, you can refer to the "Audio and Video Beginner to Mastery Manual" and click to view detailed categories.
Summarize
H.265 coding framework flow chart.
