Tutorials are geared towards beginners by providing simple code examples for basic media features and excellent code examples to start working with the Intel Media SDK. All code examples come with Microsoft Visual Studio* solution files (Windows) and Makefile (Linux) to get you started.
Sets up Intel Media SDK session and perform queries to determine selected implementation and which API version is used
Decodes AVC stream into YUV file using system memory surfaces, showcasing simple synchronous decode pipeline flow
Adds use of video memory surfaces for improved decode performance
Encodes YUV frames from file into AVC stream using surfaces in system memory, showcasing simple synchronous encode pipeline flow.
Adds use of video memory surfaces for improved encode performance.
Adds asynchronous operation to previous example, resulting in further improved performance.
Transcodes (decode+encode) AVC stream to another AVC stream using system memory surfaces.
Same as previous sample but uses the Intel Media SDK opaque memory feature. The opaque memory type hides surface allocation specifics and allows the Intel Media SDK to select the best type for the execution in hardware or software.
Adds asynchronous operation to the transcode pipeline implementation, resulting in further improved performance.
Same as "simple_5_transcode" sample but uses video memory surfaces instead. While opaque surfaces use video memory internally, application-level video memory allocation is required to integrate components not in Intel Media SDK.
Same as "simple_5_transcode_opaque - async" sample but pipeline includes video frame processing (VPP) resize.
Video Processing and more
Showcases VPP using system memory surfaces. Highlights frame resize and denoise filter processing.
Adds use of video memory surfaces for improved VPP performance.
Similar to the “simple_2_decode” sample but adds VPP post-processing capabilities to showcase resize and ProcAmp
Similar to the “simple_3_encode_vmem” sample with additional code to illustrate how to configure an encode pipeline for low latency and how to measure latency.
Similar to the “simple_5_transcode_opaque” sample with additional code to illustrate how to configure a transcode pipeline for low latency and how to measure latency.
Similar to the “simple_3_encode_vmem” sample but adds VPP pre-processing capabilities to show frame color conversion from RGB32(4) to NV12.
OpenCL and the OpenCL logo are trademarks of Apple Inc. used by permission by Kronos.