Jetson Linux API Reference
38.2 Release
01_video_encode (video encode)

Overview

The video_encode sample application demonstrates how to encode H.264/H.265/AV1 video streams.

The application YUV reads input buffers from a file, performs video encoding, and saves the encoded bitstream to an elementary .264, .265, or av1 file.

The application runs on file source simulated input buffers, and so does not require a camera.

Supported video formats are:

  • H.264 (Orin, Thor)
  • H.265 (Orin, Thor)
  • AV1 (Orin only)

Supported YUV formats are:

  • YUV420
  • YUV444
  • NV24
  • P010_10LE
  • NV24_10LE
    Note
    For 10-bit encoding, YUV input must be 16-bit MSB aligned.
    For AV1 YUV420 8 bit is supported only.
    NV24 and NV24_10LE is not supported in Jetson Thor.

CUDA Integration

The application includes comprehensive CUDA support for GPU-accelerated encoding:

  • CUDA Memory Types: Supports CUDA Device and CUDA Pinned memory allocation
  • GPU Detection: Automatically detects GPU capabilities using isGPUEnabled()
  • Memory Management: Intelligent memory type selection based on GPU availability
  • Performance Optimization: GPU-accelerated processing paths for enhanced performance

Thor Platform Support

On Thor platforms, the application utilizes GPU-accelerated encoding for:

  • Enhanced GPU resource management and allocation
  • Optimized memory bandwidth utilization between CPU and GPU
  • Improved multi-stream encoding capabilities with GPU coordination
  • Better integration with CUDA-based system-wide resource scheduling
Note
Some features marked as "Not supported for T264" indicate Thor-specific limitations or alternative implementations through GPU-accelerated paths.

Building and Running

Prerequisites

  • You have followed steps 1-3 in Building and Running.
  • If you are building from your host Linux PC (x86), you have followed step 4 in Building and Running.
  • For CUDA functionality, ensure CUDA runtime is properly installed and configured.

To build

  • Enter: 
     $ cd /usr/src/jetson_multimedia_api/samples/01_video_encode
 $ make

To run

  • Enter: 
    $ video_encode <in-file> <in-width> <in-height> <encoder-type> <out-file> [OPTIONS]

CUDA-specific Options

The application supports several CUDA-related command-line options:

  • -alloc_type_oplane <num>: Allocation memory type for output plane buffer
    • 0: Default (system-managed allocation)
    • 1: CUDA Pinned (host memory accessible by GPU)
    • 2: CUDA Device (GPU memory)
    • 4: Surface Array

To view supported options

Enter: 

   $ ./video_encode --help

Example

   $ ./video_encode ../../data/Video/sample_outdoor_car_1080p_10fps.yuv 1920 1080 H264 sample_outdoor_car_1080p_10fps.h264

CUDA-enabled Example

   $ ./video_encode input.yuv 1920 1080 H264 output.h264 -alloc_type_oplane 2

This example uses CUDA Device memory for optimal GPU performance.

Flow

The following diagram shows the flow through this sample.

  • The Output Plane receives input in YUV frame format and delivers it to the Encoder for encoding.
  • The Capture Plane transfers encoded frames to the application in bitstream format.
  • The encoded bitstream is written to a file.
  • For the Output Plane the application supports MMAP, DMABUF, and USRPTR memory types. For the Capture Plane it supports MMAP memory type.
    Note
    USRPTR is not supported in Jetson Thor for output plane.

CUDA Processing Flow

When CUDA is enabled, the processing flow includes additional GPU-accelerated paths:

  1. Memory Allocation: CUDA-aware memory allocation using NvBufSurf APIs
  2. GPU Detection: Runtime detection of GPU capabilities via isGPUEnabled()
  3. Memory Type Configuration: Automatic configuration of CUDA memory types using setCudaMemType()
  4. Buffer Management: Intelligent buffer management for CUDA Device/Pinned memory
  5. Performance Optimization: GPU-accelerated encoding with optimized memory transfers

Thor Platform Flow

On Thor platforms:

  1. Resource Discovery: GPU-based resource enumeration and capability detection
  2. Memory Management: Enhanced memory allocation through GPU-aware APIs
  3. Stream Management: Multi-stream encoding coordination via GPU scheduler
  4. Performance Monitoring: Real-time performance metrics and adaptive GPU optimization

Key Structure and Classes

The sample uses the following key structures and classes.

Element Description
NvVideoEncoder Contains all video encoding-related elements and functions.
Enc_pollthread A pointer to the thread handler for the encoding capture loop.

The NvVideoEncoder class packages all video encoding-related elements and functions. Key members used in the sample are:

Member Description
output_plane Specifies the V4L2 output plane.
capture_plane Specifies the V4L2 capture plane.
createVideoEncoder Static function to create video encode object.
subscribeEvent Subscribe event.
setOutputPlaneFormat Sets the output plane format.
setCapturePlaneFormat Sets the capture plane format.
dqEvent Dqueue the event which reports by the V4L2 device.
isInError Checks if under error state.

Class NvVideoEncoder contains two key elements: output_plane and capture_plane. These objects are derived from class type NvV4l2ElementPlane. The sample uses the following key members:

Element Description
setupPlane Sets up the plane of V4L2 element.
deinitPlane Destroys the plane of the V4L2 element.
setStreamStatus Starts/stops the stream.
setDQThreadCallback Sets the callback function of the dqueue buffer thread.
startDQThread Starts the thread of the dqueue buffer.
stopDQThread Stops the thread of the dqueue buffer.
qBuffer Queues the V4L2 buffer.
dqBuffer Dequeues the V4L2 buffer.
getNumBuffers Gets the number of V4L2 buffers.
getNumQueuedBuffers Gets the number of buffers currently queued on the plane.

CUDA Support

CUDA-Specific APIs

The sample includes additional CUDA-specific functionality:

API Function Description
isGPUEnabled Checks if GPU configuration is enabled for CUDA processing.
setCudaMemType Sets CUDA memory type (Device/Pinned) when GPU is enabled.
setCUDASliceIntrarefresh Sets CUDA-specific slice intra-refresh parameters.
setCudaConstantQp Sets CUDA-specific constant QP values.

Memory Management APIs

API Description
NvBufSurf::NvAllocate Allocates buffer surfaces with CUDA memory type support.
NVBUF_MEM_CUDA_DEVICE Enum for CUDA device memory allocation.
NVBUF_MEM_CUDA_PINNED Enum for CUDA pinned (host) memory allocation.
V4L2_CUDA_MEM_TYPE_* V4L2 controls for CUDA memory type configuration.

GPU Integration on Thor

The GPU acceleration provides enhanced capabilities:

  • Unified Memory Management: Coordinated allocation across CPU/GPU memory domains
  • Performance Optimization: Real-time performance monitoring and adaptive GPU tuning
  • Multi-Stream Support: Efficient GPU resource sharing for concurrent encoding streams
  • System Integration: Deep integration with Thor platform GPU capabilities

Platform-Specific Considerations

Platform Feature Thor (T264) Other Platforms
CUDA Support GPU-accelerated Direct CUDA API
Memory Types GPU-managed Standard allocation
Multi-stream GPU-scheduled Application managed
Resource limits Dynamic GPU allocation Static configuration
Note
Features marked as "Not supported for T264" may have equivalent functionality through GPU-accelerated paths or platform-specific alternatives.