Advanced Audio Coding
A family of lossy audio codecs developed under the MPEG standards framework, including configurations such as AAC-LC and HE-AAC, is used in broadcasting, streaming, mobile devices, and MPEG-4 files.
Explanation
Advanced Audio Coding (AAC, 高级音频编码) is a family of lossy audio codecs within the MPEG standards framework. AAC was initially standardized as MPEG-2 Part 7 in 1997, later incorporated into MPEG-4 Audio, and expanded in scope and applications through various audio object types. It is not a single “one-size-fits-all” encoder with fixed parameters, but rather a set of configurations designed for different decoding requirements.
AAC primarily uses the modified discrete cosine transform (MDCT) to process audio; the encoder selects either long or short windows based on transient, spectral, and auditory models, and performs entropy coding on quantized spectra, scaling factors, and channel-related information. Techniques such as temporal noise shaping, prediction, and joint stereo can reduce audible distortion or data volume across different object types. The standard specifies the bitstream and decoding process, while psychoacoustic models and bitrate control are determined by the encoder implementation.
AAC-LC (Low Complexity) is the most common base configuration in consumer media. HE-AAC adds spectral band replication to AAC-LC to reconstruct high frequencies at lower bitrates; HE-AAC v2 also supports parametric stereo. The subsequent xHE-AAC is designed for adaptive transmission ranging from low-bitrate speech to music. Player support for “AAC” does not imply support for all object types; multichannel audio, HE extensions, or specific transport syntax may still exceed a device’s capabilities.
Raw AAC frames can use formats such as ADTS, ADIF, or Low-Overhead Audio Transport Multiplexing (LOTM), and are often encapsulated as tracks within MP4, M4A, 3GP, or MPEG-2 Transport Streams. The ADTS header provides synchronization and parameters for consecutive frames, while MP4 places configuration, timing, and sample tables within the container structure. The same AAC audio can be repackaged between different containers without re-encoding; changing the container should not result in a change in audio quality.
M4A is not another name for AAC. A `.m4a` file can contain AAC, ALAC, or other audio formats; conversely, AAC is not exclusive to M4A. Media information should report the container, codec type, sample rate, number of channels, and bitrate.
AAC is a lossy codec; once decoded to PCM, WAVE, FLAC, or ALAC, only the decoded result is preserved. Bitrate, audio format, and encoder quality collectively influence distortion; one cannot simply assume that a particular AAC file is necessarily superior to an MP3 file at the same bitrate, nor can one infer from a high bitrate label that the source has not undergone prior lossy encoding.