AVSampleFormat

AVSampleFormat is an enumeration used by FFmpeg’s libavutil to describe the format of audio samples. It encodes both the bit depth of each sample (for example 8-, 16-, 32-bit, or floating point) and the memory layout of the data (packed versus planar). This information is essential for correctly interpreting, allocating, and converting audio buffers across codecs, resamplers, and containers.

The enum covers a set of packed formats, where samples for all channels are interleaved in memory, and corresponding planar variants, where each channel’s samples are stored in a separate memory plane. Common packed formats include 8-bit unsigned, 16-bit signed, 32-bit signed, and 32-bit floating point. Their planar counterparts add a P suffix to indicate the separate-channel layout (for example, S16P, FLTP).

In FFmpeg, AVSampleFormat also includes sentinel values such as AV_SAMPLE_FMT_NONE to denote an invalid or unspecified format, and AV_SAMPLE_FMT_NB to indicate the number of distinct sample formats defined. The exact list of formats is defined in libavutil and includes a mix of integer and floating-point representations across packed and planar layouts.

Uses and utilities: APIs such as those for codecs, resamplers, and frame handling rely on AVSampleFormat to determine bytes per sample, how many samples constitute a frame, and how to access or convert audio data. Helper functions and macros in libavutil, such as those that convert a format to a human-readable name or report the per-sample byte size, operate on this enum to enable format-agnostic audio processing.