Written by Christian Ahmer | 11/20/2023

MPEG

The MPEG (Moving Picture Experts Group) standard refers to a suite of digital video and audio compression specifications and formats developed by the group bearing the same name. MPEG standards include MPEG-1, MPEG-2, MPEG-4, and more recently, MPEG-H and MPEG-DASH. Each standard addresses different needs and has different applications, from video CDs to digital television, streaming media, and beyond.

MPEG-1 was one of the first compression standards for audio and video. It was designed to compress VHS-quality raw digital video and CD audio down to about 1.5 Mbit/s without excessively compromising quality. This level of compression made it possible to store a full-length movie on a CD-ROM and was the standard used for Video CD (VCD).

MPEG-2 expanded on MPEG-1 and provided the framework for digital television and DVDs. It offers higher quality video at the cost of higher bitrates, supporting interlaced video and broadcasting standards.

MPEG-4 is a broad standard that covers a wide range of multimedia applications, including web streaming and broadcast television. It introduced new video and audio codecs like Advanced Video Coding (AVC, also known as H.264) and High Efficiency Video Coding (HEVC, also known as H.265). These codecs are highly efficient at compressing video to low bitrates without significant loss in quality and are the backbone of modern video streaming services.

The process of MPEG video compression typically involves the following steps:

  1. Frame Types: MPEG video streams consist of three types of frames: I-frames (intra-coded), which are self-contained reference frames; P-frames (predictive-coded), which use data from previous I-frames or P-frames to encode only the changes in the scene; and B-frames (bi-predictive-coded), which use both previous and following frames to encode only the differences.

  2. Temporal Compression: By using P-frames and B-frames, MPEG takes advantage of temporal redundancy, meaning that many frames in a video are similar to their predecessors or successors. By only storing the differences between frames, MPEG can significantly reduce the amount of data needed.

  3. Spatial Compression: Spatial redundancy within a single frame is also reduced. MPEG uses techniques similar to JPEG for compressing the individual frames, like Discrete Cosine Transform (DCT) and quantization.

  4. Motion Estimation: MPEG compressors use motion estimation to determine how parts of an image have moved from one frame to the next. The compressor then only needs to record the motion (as vectors) and the changes in the image, rather than the entire frame.

  5. Entropy Coding: After quantization, the coefficients are processed using entropy coding, like Huffman coding or arithmetic coding, to further reduce the size.

The effectiveness of MPEG compression depends on the content of the video, the motion complexity, and the desired output quality. Fast-moving and complex videos require higher bit rates to maintain quality, while slower, simpler videos can be compressed more without noticeable quality loss.

The MPEG standards have played a crucial role in the proliferation of digital video, from television broadcasting and DVDs to streaming services on the internet. Each evolution of the standard has brought about improvements in compression efficiency, allowing higher quality video to be stored and transmitted even as the demand for resolution and frame rates has increased.