In this year, the company has continually planned H.265. But it seems to support the chip of the H.265 code, but the industry is constantly being promoted, we have to keep up with the era. Take a break, go to understand the relevant things, so as not to ask for an answer.
First, standard
HEVC is an abbreviation of "High EffiCiency Video Coding", also called H.265. Many years ago, I finally lost 1.0 version by 2013. So far, 3 versions have been released. The latest version 3.0 in April 2015. Different versions have a little different, but the overall frame is different. For example, the field in front of the version of the VPS is VPS_RESERVED_THREE_2BITS, occupying 2 bits. In subsequent versions, this field is VPS_BASE_LAYER_ITERNAL_FLAG and VPS_BASE_LAYER_AVAILABLE_FLAG, respectively, respectively. There are still many similar.
Many of the online articles use the 1.0 version of the grammar, it is recommended to download 3 versions of the standard documentation when learning.
HEVC / H.265 Learn Camp Address: https: //hevc.hhi.fraunhofer.de/. There are standard download addresses, related papers, and HEVC test model HM code on this website.
Standard download address: http://www.itu.int/rec/t-rec-h.265/EN
About H.265 analytical tools, there are many networks, but they are all charges, and the price is not. The trial version either only uses a few days, or only 20 frames of the video. But for the initial understanding of H.265 seems enough.
Codecvisa Tools Download: http://www.codecian.com/
H265VIDEOESVIEWER Tools Download: http://www.jongbel.com/manual-analysis/hevch-265-video-es-viewer/
There are many articles on the Internet, and there are many articles that have been explained and analyzed here.
Second, test model HM
HM code download: https://hevc.hi.fraunhofer.de/svn/svn_hevcsoftware/tags/. It seems that the site does not provide a compressed package download, can only download it with SVN. The latest version 16.6 is currently available. If you download the entire SVN, the space will be very large, so you only need to download the version you want to study. Also, you must explain that the HEVC standard version corresponding to different versions is different. If the standard uses the latest, the HM code is recommended to use the latest.
HM can be compiled, truth, track code, and debugging in Linux or Windows, or Windows is convenient. I installed VS2010, double-click \ Build \ HM_VC10.SLN to compile. The generated file is located in: \ bin \ VC10 \ Win32 \ Debug. There are several EXE files:
Decoder: TappDecoder.exe Encoder: TAPPENCODER.EXE
Code Usage: 1, YUV Original Sequence: Suzie_QCIF.YUV 2, copy the encoder_intra_main.cfg file from the \ cfg directory. 3, new YUV configuration file Suzie_qcif.cfg, content: # ======== File I / O ======================== InputFile: Suzie_qcif.yuv InputBitDepth: 8 # input bitdepth FrameRate: 30 # frame Rate per second frameSkip: 0 # Number of frames to be skipped in input SourceWidth: 176 # Input frame width SourceHeight: 144 # Input frame height FramesToBeEncoded: 50 # 50 encoded
Enter: TAPPENCODER.EXE-C Encoder_intra_main.cfg -c suzie_qcif.cfg
The HEVC code stream file Str.bin will then generate.
Let's introduce the encoding parameters. There are many Encoder_xx.cfg files in the CFG directory, which is an important parameter configuration file, such as Profile, GOP, QP, and so on. BitStreamFile: Output code stream files, the example eNCoder_intra_main.cfg file is the default value str.bin. InputFile: Enter a file, the above example is a separate file configuration. SourceWidth, SourceHeight: Enter the width of the source file, high InputBitDepth: Enter a depth, common YUV is generally 8 bits. INTERNALBITDEPTH: The location of the encoder uses. Also equivalent to the output bit depth OutputBitDepth. Framerate: FramestobeEncoded: Coded frame number, note, when profile is Main-Still-Pictur, must be 1 Profile: Specify the code stream. Take the value: main main-refxt, high-throughput-revx, main-scc, and so on. Level: Specifies the LEVEL of the code stream. Value: NONE, 1, 2, 2.2.1, 3, 3.1, 4, 4.1, 5, 5.1, 5.2, 6, 6.1, 6.2, 8.5. Tier: Specifies the tier of the code stream, only main or high. Intraperiod: IFAC GOPSIZE: GOP size Note: Intraperiod and GoPSIZE These two parameters are a bit complicated, some are selected 1, that is, I frame, if not, set some parameters, see Encoder_lowdelay_main_rext.cfg. QP: Specify quantization parameters. The default is 30, the HEVC ranges from 0 to 51. The larger the value, the better the compression, the smaller the volume, but the quality is worse.
Li Li 2015.9.6
Our other product:
