VCodecOneVpl C++ library

v2.1.1

Table of contents

• Overview
• Versions
• Library files
• VCodecOneVpl class description
  • VCodecOneVpl class declaration
  • getVersion method
  • transcode method
  • setParam method
  • getParam method
  • executeCommand method
• Build and connect to your project
• Installation on Linux
• Installation on Windows
• Simple example
• Test application

Overview

VCodecOneVpl C++ library provides hardware video encoding / decoding (H264, HEVC and JPEG) for Intel HD Graphics based on oneVPL API. VCodecOneVpl class inherits interface and data structures from open source VCodec library (provides interface for video codecs, source code included, Apache 2.0 license) and also includes Logger open source library (provides functions to print logs, source code included, Apache 2.0 license). The library provides simple interface to be implemented in different C++ projects and compatible with Linux and Windows. It is written with C++17 standard. The library is supplied as source code only in form of CMake project.

Encoding time for 11th Gen Intel(R) Core(TM) i5-1145G7E on Ubuntu 22.04 LTS:

codec / resolution	2560x1440	1920x1080	1280x720	640x512
H264	11.6 msec	8.6 msec	4.4 msec	2.6 msec
HEVC	23.4 msec	15.2 msec	9.3 msec	5.2 msec
JPEG	8.2 msec	4.8 msec	2.5 msec	1.2 msec

Decoding time for 11th Gen Intel(R) Core(TM) i5-1145G7E on Ubuntu 22.04 LTS:

codec / resolution	2560x1440	1920x1080	1280x720	640x512
H264	7.3 msec	4.8 msec	2.5 msec	1.3 msec
HEVC	7.4 msec	4.3 msec	2.2 msec	1.3 msec
JPEG	8.7 msec	5.2 msec	2.6 msec	1.3 msec

Versions

Table 1 - Library versions.

Version	Release date	What’s new
1.0.0	01.12.2022	First version.
2.0.0	29.09.2023	- Interface changes to VCodec. - Added decoding support. - Added JPEG support.
2.0.1	12.11.2023	- Frame class updated.
2.0.2	10.01.2023	- VCodec interface updated. - Examples updated. - Documentation updated.
2.0.3	25.04.2024	- Test application updated. - Documentation updated.
2.0.4	17.06.2024	- Documentation updated. - VCodec submodule updated.
2.1.0	10.07.2024	- CMake updated. - VCodec submodule updated. - Repository structure changed.
2.1.1	04.11.2024	- Update VCodec interface with variable bitrate params.

Library files

The library is supplied only by source code. The user is given a set of files in the form of a CMake project (repository). The repository structure is shown below:

CMakeLists.txt ---------------- Main CMake file of the library.
3rdparty ---------------------- Folder with third-party libraries.
    CMakeLists.txt ------------ CMake file which includes third-party libraries.
    Logger -------------------- Folder width Logger library source code.
    VCodec -------------------- older width VCodec library source code.
example ----------------------- Folder with simple example of VCodecOneVpl usage.
    CMakeLists.txt ------------ CMake file for example application.
    main.cpp ------------------ Source code file of example application.
test -------------------------- Folder with codec test application.
    CMakeLists.txt ------------ CMake file for transcode test application.
    main.cpp ------------------ Source code file of transcode test application.
src --------------------------- Folder with source code of the library.
    Impl ---------------------- Folder with implementation of video codec.
        VCodecOneVplImpl.cpp -- C++ implementation file.
        VCodecOneVplImpl.h ---- Header file which includes VCodecOneVplImpl class declaration.
    CMakeLists.txt ------------ CMake file of the library.
    VCodecOneVpl.cpp ---------- C++ implementation file.
    VCodecOneVpl.h ------------ Header file which includes VCodecOneVpl class declaration.
    VCodecOneVplVersion.h ----- Header file which includes version of the library.
    VCodecOneVplVersion.h.in -- CMake service file to generate version file.

VCodecOneVpl class description

VCodecOneVpl class declaration

VCodecOneVpl class declared in VCodecOneVpl.h file. Class declaration:

namespace cr
{
namespace video
{
/**
 * @brief VCodecOneVpl class.
 */
class VCodecOneVplImpl : public VCodec
{
public:

    /// Class constructor.
    VCodecOneVplImpl();

    /// Class destructor.
    ~VCodecOneVplImpl();

    /// Set parameter value.
    bool setParam(VCodecParam id, float value) override;

    /// Get parameter value.
    float getParam(VCodecParam id) override;

    /// Encode/Decode video frame.
    bool transcode(Frame& src, Frame& dst) override;

    /// Execute command.
    bool executeCommand(VCodecCommand id) override;
};
}
}

getVersion method

The getVersion() method returns string of current version of VCodecOneVpl class. Method declaration:

static std::string getVersion();

Method can be used without VCodecOneVpl class instance:

cout << "VCodecOneVpl class version: " << VCodecOneVpl::getVersion() << endl;

Console output:

VCodecOneVpl class version: 2.1.1

transcode method

The transcode(…) method encodes and decodes video frame (Frame class). Video codec encodes/decodes video frames frame-by-frame. Method declaration:

bool transcode(Frame& src, Frame& dst);

Parameter	Value
src	Source Frame class object. To encode video src frame must have NV12 raw pixel format. To decode video data src frame must have compressed pixel format (field fourcc of Frame class): HEVC, JPEG, H264.
dst	Result Frame class object. To encode video data dst frame must have compressed pixel format (field fourcc of Frame class): HEVC, JPEG, H264. In contrary, src frame must be NV12.

Returns: TRUE if frame was encoded/decoded or FALSE if not.

setParam method

The setParam(…) method sets new video codec parameters value. Method declaration:

bool setParam(VCodecParam id, float value) override;

Parameter	Description
id	Video codec parameter ID according to VCodecParam enum.
value	Video codec parameter value.

Returns: TRUE is the parameter was set or FALSE if not.

VCodec.h file of VCodec library defines IDs for parameters (VCodecParam enum) and IDs for commands (VCodecCommand enum). VCodecParam declaration:

namespace cr
{
namespace video
{
enum class VCodecParam
{
    /// [read/write] Log level: 0-Disable, 1-Console, 2-File, 3-Console and file.
    LOG_LEVEL = 1,
    /// [read/write] Bitrate, kbps. For H264 and H265 codecs.
    BITRATE_KBPS,
    /// [read/write] Minimum bitrate, kbps. For variable bitrate mode.
    MIN_BITRATE_KBPS,
    /// [read/write] Maximum bitrate, kbps. For variable bitrate mode.
    MAX_BITRATE_KBPS,
    /// [read/write] Bitrate mode: 0 - constant bitrate, 1 - variable bitrate.
    BITRATE_MODE,
    /// [read/write] Quality 0-100%. For JPEG codecs.
    QUALITY,
    /// [read/write] FPS. For H264 and H265 codecs.
    FPS,
    /// [read/write] GOP size. For H264 and H265 codecs.
    GOP,
    /// [read/write] H264 profile: 0 - Baseline, 1 - Main, 2 - High.
    H264_PROFILE,
    /// [read/write] Codec type. Depends on implementation.
    TYPE,
    /// Custom 1. Depends on implementation.
    CUSTOM_1,
    /// Custom 2. Depends on implementation.
    CUSTOM_2,
    /// Custom 3. Depends on implementation.
    CUSTOM_3
};
}
}

Table 2 - Video codec params description. Some params not supported by particular VCodecLibav library.

Parameter	Access	Description
LOG_LEVEL	read / write	Logging mode. Values: 0 - Disable. 1 - Only file. 2 - Only terminal (console). 3 - File and terminal.
BITRATE_KBPS	read / write	Bitrate, kbps. Default 5000 kbps. For H264 and H265(HEVC) encoding. According to this value, FPS and GOP size video codec calculate parameter for H264 or H265(HEVC) encoding. - Sets special settings for noisy video if bitrate is 0 for hardware encoding. - For software JPEG encoding bitrate incising gives higher quality.
MIN_BITRATE_KBPS	read / write	Minimum bitrate, kbps. Only for H264 and H265(HEVC) codecs. Default 2000 kbps. For variable bitrate mode. For H264 and H265(HEVC) encoding.
MAX_BITRATE_KBPS	read / write	Maximum bitrate, kbps. Only for H264 and H265(HEVC) codecs. Default 8000 kbps. For variable bitrate mode. For H264 and H265(HEVC) encoding.
BITRATE_MODE	read / write	Only for H264 and H265(HEVC) codecs. Bitrate mode: 0 - constant bitrate (default), 1 - variable bitrate. In variable bitrate mode for software encoder
QUALITY	read / write	Quality 0(low quality)-100%(maximum quality). Only for hardware JPEG encoding. Not supported by JPEG SW encoding.
FPS	read / write	FPS. For H264 and H265(HEVC) encoding only. According to this value, FPS and GOP size video codec calculate parameter for H264 and H265(HEVC) encoding.
GOP	read / write	GOP size (Period of key frames) for H264 and H265(HEVC) encoding. Value: 1 - each output frame is key frame, 20 - each 20th frame is key frame etc.
H264_PROFILE	read / write	H264 profile for H264 encoding: 0 - Baseline. 1 - Main. 2 - High. For hardware encoding only.
TYPE	read / write	Type of encoder/decoder: 0 - hardware (default). 1 - software.
CUSTOM_1	read / write	Number of thread for software encoder/decoder. Thread’s count for software encoder\decoder H264 / HEVC(H265) (not for JPEG): 1 - 32, default value 1. Note: in case one thread software encoder produces only one slice per frame. In case multiple threads frame can have multiple slices which can effect on RTSP servers.
CUSTOM_2	read / write	Not supported by VCodecLibav library.
CUSTOM_3	read / write	Not supported by VCodecLibav library.

getParam method

The getParam(…) method returns video codec parameter value. Method declaration:

float getParam(VCodecParam id);

Parameter	Description
id	Video codec parameter ID according to VCodecParam enum (see Table 2).

Returns: parameter value or -1 if the parameters not supported.

executeCommand method

The executeCommand(…) method executes video codec command. Version 2.0.2 doesn’t support commands. Method will return FALSE. Method declaration:

bool executeCommand(VCodecCommand id);

Parameter	Description
id	Video codec command ID according to VCodecCommand enum.

Returns: method returns FALSE in any case.

VCodec.h file of VCodec library defines IDs for parameters (VCodecParam enum, table 2) and IDs for commands (VCodecCommand enum). VCodecCommand declaration:

enum class VCodecCommand
{
    /// Reset.
    RESET = 1,
    /// Generate key frame. For H264 and H265 codecs.
    MAKE_KEY_FRAME
};

Table 3 - Video codec commands description. Some commands maybe unsupported by particular video codec class.

Command	Description
RESET	Not supported by VCodecOneVpl library.
MAKE_KEY_FRAME	Not supported by VCodecOneVpl library.

Build and connect to your project

Before build you have to install OneVpl on your system. Follow Installation on Linux or Installation on Windows instructions. Typical commands to build VCodecOneVpl library:

cd VCodecOneVpl
git submodule update --init --recursive
mkdir build
cd build
cmake ..
make

If you want connect VCodecOneVpl library to your CMake project as source code you can make follow. For example, if your repository has structure:

CMakeLists.txt
src
    CMakeList.txt
    yourLib.h
    yourLib.cpp

Create folder 3rdparty in your repository. Copy repository folder VCodecOneVpl to 3rdparty folder. New structure of your repository:

CMakeLists.txt
src
    CMakeList.txt
    yourLib.h
    yourLib.cpp
3rdparty
    VCodecOneVpl

Create CMakeLists.txt file in 3rdparty folder. CMakeLists.txt should contain:

cmake_minimum_required(VERSION 3.13)

################################################################################
## 3RD-PARTY
## dependencies for the project
################################################################################
project(3rdparty LANGUAGES CXX)

################################################################################
## SETTINGS
## basic 3rd-party settings before use
################################################################################
# To inherit the top-level architecture when the project is used as a submodule.
SET(PARENT ${PARENT}_YOUR_PROJECT_3RDPARTY)
# Disable self-overwriting of parameters inside included subdirectories.
SET(${PARENT}_SUBMODULE_CACHE_OVERWRITE OFF CACHE BOOL "" FORCE)

################################################################################
## CONFIGURATION
## 3rd-party submodules configuration
################################################################################
SET(${PARENT}_SUBMODULE_VCODEC_ONE_VPL                  ON  CACHE BOOL "" FORCE)
if (${PARENT}_SUBMODULE_VCODEC_ONE_VPL)
    SET(${PARENT}_VCODEC_ONE_VPL                        ON  CACHE BOOL "" FORCE)
    SET(${PARENT}_VCODEC_ONE_VPL_TEST                   OFF CACHE BOOL "" FORCE)
    SET(${PARENT}_VCODEC_ONE_VPL_EXAMPLE                OFF CACHE BOOL "" FORCE)
endif()

################################################################################
## INCLUDING SUBDIRECTORIES
## Adding subdirectories according to the 3rd-party configuration
################################################################################
if (${PARENT}_SUBMODULE_VCODEC_ONE_VPL)
    add_subdirectory(VCodecOneVpl)
endif()

File 3rdparty/CMakeLists.txt adds folder VCodecOneVpl to your project and will exclude test application and example from compiling (by default the test application and example are excluded from compiling if the VCodecOneVpl included as sub-repository). Your repository new structure will be:

CMakeLists.txt
src
    CMakeList.txt
    yourLib.h
    yourLib.cpp
3rdparty
    CMakeLists.txt
    VCodecOneVpl

Next you need include folder 3rdparty in main CMakeLists.txt file of your repository. Add string at the end of your main CMakeLists.txt:

add_subdirectory(3rdparty)

Next you have to include VCodecOneVpl library in your src/CMakeLists.txt file:

target_link_libraries(${PROJECT_NAME} VCodecOneVpl)

Done!

Installation on Linux

There are several steps to launching VCodecOneVpl on Linux ( tested on Ubuntu 22.04 LTS):

1. Install ubuntu with last updates:

   sudo apt-get update
   sudo apt-get upgrade
   sudo reboot
   

2. Install LibVA:

   sudo apt-get install git cmake pkg-config meson libdrm-dev automake libtool
   cd Downloads
   git clone https://github.com/intel/libva.git
   cd libva
   ./autogen.sh --prefix=/usr --libdir=/usr/lib/x86_64-linux-gnu
   make
   sudo make install
   

3. Install gmmlib:

   cd Downloads
   git clone https://github.com/intel/gmmlib.git
   cd gmmlib
   mkdir build
   cd build
   cmake -DCMAKE_BUILD_TYPE=Release ..
   make -j"$(nproc)"
   sudo make install
   

4. Install intel media driver for VAAPI:

   cd Downloads
   git clone https://github.com/intel/media-driver.git
   mkdir build_media && cd build_media
   cmake ../media-driver
   make -j"$(nproc)"
   sudo make install
   

5. Install oneVPL-intel-gpu:

   cd Downloads
   git clone https://github.com/oneapi-src/oneVPL-intel-gpu onevpl-gpu
   cd onevpl-gpu
   mkdir build && cd build
   cmake ..
   make -j"$(nproc)"
   sudo make install
   

6. Install additional packets to oneVPL-intel-gpu:

   sudo apt update
   sudo apt install -y gpg-agent wget
   wget -qO - https://repositories.intel.com/gpu/intel-graphics.key | sudo gpg --dearmor --output /usr/share/keyrings/intel-graphics.gpg
   echo "deb [arch=amd64 signed-by=/usr/share/keyrings/intel-graphics.gpg] https://repositories.intel.com/gpu/ubuntu jammy/production/2328 unified" | sudo tee /etc/apt/sources.list.d/intel-gpu-jammy.list
   sudo apt update
   sudo apt install -y linux-headers-$(uname -r) flex bison intel-fw-gpu intel-i915-dkms xpu-smi
   sudo reboot 
   sudo apt install -y intel-opencl-icd intel-level-zero-gpu level-zero intel-media-va-driver-non-free libmfx1 libmfxgen1 libvpl2 libegl-mesa0 libegl1-mesa libegl1-mesa-dev libgbm1 libgl1-mesa-dev libgl1-mesa-dri libglapi-mesa libgles2-mesa-dev libglx-mesa0 libigdgmm12 libxatracker2 mesa-va-drivers mesa-vdpau-drivers mesa-vulkan-drivers va-driver-all vainfo hwinfo clinfo
   sudo apt install -y libigc-dev intel-igc-cm libigdfcl-dev libigfxcmrt-dev level-zero-dev
   

7. Install oneVPL:

   cd Downloads
   git clone https://github.com/oneapi-src/oneVPL.git
   cd oneVPL
   sudo script/bootstrap
   script/build
   sudo script/install
   

Installation on Windows

1. Install oneVPL:

   git clone https://github.com/oneapi-src/oneVPL.git
   cd oneVPL
   script/bootstrap.bat
   script/build.bat
   script/install.bat
   

Simple example

Example application generates image color pattern with moving rectangle, encodes with HEVC codec and writes compressed data to binary file “out.hevc”. Example shows how to create codec objects and how to encode video frames:

#include <iostream>
#include "VCodecOneVpl.h"

int main(void)
{
    // Set codec parameters.
    cr::video::VCodec* videoCodec = new cr::video::VCodecOneVpl();
    videoCodec->setParam(cr::video::VCodecParam::BITRATE_KBPS, 7500);
    videoCodec->setParam(cr::video::VCodecParam::GOP, 30);
    videoCodec->setParam(cr::video::VCodecParam::FPS, 30);

    // Create NV12 frame and fill color planes by random values.
    const int width = 1280;
    const int height = 720;
    cr::video::Frame frameNv12(width, height, cr::video::Fourcc::NV12);
    for (uint32_t i = 0; i < frameNv12.size; ++i)
        frameNv12.data[i] = (uint8_t)i;

    // Create output HEVC frame.
    cr::video::Frame frameHEVC(width, height, cr::video::Fourcc::HEVC);

    // Create output file.
    FILE *outputFile = fopen("out.hevc", "w+b");

    // Params for moving object.
    int objectWidth = 128;
    int objectHeight = 128;
    int directionX = 1;
    int directionY = 1;
    int objectX = width / 4;
    int objectY = height / 2;

    // Encode and record 200 frames.
    for (uint32_t n = 0; n < 200; ++n)
    {
        // Draw moving object.
        memset(frameNv12.data, 128, width * height);
        for (int y = objectY; y < objectY + objectHeight; ++y)
            for (int x = objectX; x < objectX + objectHeight; ++x)
                frameNv12.data[y * width + x] = 255;
        objectX += directionX;
        objectY += directionY;
        if (objectX >= width - objectWidth - 5 || objectX <= objectWidth + 5)
            directionX = -directionX;
        if (objectY >= height - objectHeight - 5 || objectY <= objectHeight + 5)
            directionY = -directionY;

        // Encode.
        if (!videoCodec->transcode(frameNv12, frameHEVC))
        {
            std::cout << "Can't encode frame" << std::endl;
            continue;
        }

        // Write to file.
        fwrite(frameHEVC.data, frameHEVC.size, 1, outputFile);  
    }

    // Close file.
    fclose(outputFile);

    return 1;
}

Test application

Test application (VCodecOneVpl/test/main.cpp) for VCodecOneVpl C++ library shows how library works on Intel platform. Test application generates artificial video, compresses it according to user’s parameters (codec type, bitrate or JPEG quality, GOP size and H264 profile), writes results to binary file “out.h264”, “out.hevc” or “out.mjpeg” and decompresses. The application shows encoding and decoding time for each video frame. To run application perform commands on Linux or just run VCodecOneVplTest.exe on Windows:

cd <application folder>
sudo chmod +x VCodecOneVplTest
./VCodecOneVplTest

After start you will see output:

====================================
VCodecOneVpl v2.1.1 test
====================================

Enter Encoder type (0 - JPEG, 1 - H264, 2 - HEVC) :

Chose codec type (0 - JPEG, 1 - H264). If H264 codec chosen you will see message:

====================================
VCodecOneVpl v2.1.1 test
====================================

Enter Encoder type (0 - JPEG, 1 - H264, 2 - HEVC) : 1
Default params:
Bitrate, kbps 6000
FPS: 30
GOP size: 30
Video width 1920
Video height 1080
H264 Profile: BASELINE
Use default params (0 - no, 1 - yes) :

When params chosen test application will create out.h264 file and will start writing encoded frames until stop. User can set custom parameters (video resolution, bitrate, GOP size and H264 profile). If JPEG codec chosen you will see message:

Enter Encoder type (0 - JPEG, 1 - H264, 2 - HEVC) : 1
Default params:
Bitrate, kbps 6000
FPS: 30
GOP size: 30
Video width 1920
Video height 1080
H264 Profile: 0
Use default params (0 - no, 1 - yes) : 0
Set video width : 1280
Set video height : 720
Set bitrate, kbps : 2000
Set GOP size : 30
Set FPS : 30
Set profile (0 - BASELINE, 1 - MAIN, 2 - HIGH): 0
Set number frames: 100

When params chosen test application will start writing encoded frames until stop (if user set number of frames). During encoding the application shows encoded data size and encoding time and decoding time:

Data size 1382400 / 6579   encoding time  578601 us / decoding time 411758 us
Data size 1382400 / 157   encoding time  8536 us / decoding time 6739 us
Data size 1382400 / 160   encoding time  6291 us / decoding time 2202 us
Data size 1382400 / 167   encoding time  3667 us / decoding time 3988 us
Data size 1382400 / 166   encoding time  4124 us / decoding time 3744 us
Data size 1382400 / 161   encoding time  2970 us / decoding time 4891 us
Data size 1382400 / 186   encoding time  3133 us / decoding time 1781 us