frametype

Frametype, or frame type, is a descriptor used in digital video to identify how a particular frame was encoded and how it can be reconstructed. In most lossy video codecs, frames are categorized as intra-coded frames (I-frames), predicted frames (P-frames), and bi-directionally predicted frames (B-frames). I-frames contain a complete image and do not require other frames to be decoded. P-frames store only differences from a preceding reference frame, allowing compression by referencing past data. B-frames are encoded using references from both past and future frames, achieving higher compression at the cost of additional decoding delay and buffering.

Frame type affects decoding and access: I-frames can be decoded independently, while P- and B-frames need one or more reference frames, influencing how a sequence can be randomly accessed and how decode order differs from display order. Frame type also informs the structure of a group of pictures (GOP) or equivalent, which defines how frequently I-frames appear and how many P- and B-frames separate them.

In practice, frametype is indicated in the video bitstream via a header or syntax element and is exposed by decoding libraries as an indicator such as I, P, or B (sometimes labeled as key frame for I-frames). Different standards may use slightly different terminology, but the underlying concept is the same: the frame type determines its encoding method, dependencies, and role in the sequence.

Frametype is a fundamental concept in video encoding, streaming, and editing, enabling efficient compression while preserving the ability to seek, reconstruct, or edit sequences.