MP3
Commonly known as MPEG-1 Audio Layer III, this is a lossy audio format that uses auditory models, filter banks, and entropy coding to reduce data volume; it once served as the primary format for online music and portable media players.
Explanation
MP3 is the common name for MPEG-1 Audio Layer III (MPEG-1 音频层 III), which is part of the ISO/IEC 11172-3 audio standard. The MPEG-1 version was finalized in 1992 and supports sampling rates of 32, 44.1, and 48 kHz; The backward-compatible extension of MPEG-2 added support for lower sampling rates and bit rates; the informally termed MPEG-2.5 further extended this to even lower sampling rates, but it is not part of the official hierarchy of the original ISO MPEG-2 standard.
MP3 is a lossy perceptual codec. The encoder uses a multiphase filter bank and a modified discrete cosine transform to analyze frequency components, estimates which errors are less perceptible based on auditory masking models, then quantizes the spectrum and stores it using Huffman coding. The encoding process permanently discards some information, and the decoder cannot restore the original sample when outputting PCM. Different encoders may employ different psychoacoustic models, noise control, and bitrate allocation strategies; therefore, the same bitrate does not guarantee the same level of distortion.
The bitstream consists of relatively independent audio frames, each carrying header information such as version, layer, sampling rate, bitrate, and channel mode. Layer III can utilize bit reserves, lending unused data space from certain simple frames to subsequent, more complex frames; therefore, the physical frame boundaries do not always correspond one-to-one with the data consumed by a single segment of audio. Combined stereo can reduce stereo redundancy through center-side or intensity stereo, with the specific method varying depending on the encoder and bitrate.
MP3 supports fixed bitrate, average bitrate, and variable bitrate. Fixed bitrate ensures that all frames use the same nominal bitrate, facilitating early transmission and estimation; variable bitrate allocates more data to complex segments and is typically controlled by quality targets rather than a single value per second. File duration and seamless playback may also depend on Xing, Info, or LAME headers that record the number of frames, encoding delay, and padding.
`.mp3` files typically consist of a direct concatenation of MPEG audio frames and may include ID3 tags at the beginning or end. ID3 is not part of MPEG audio encoding but is a separate system for storing metadata such as title, artist, and cover art. MP3 files can also be embedded within certain container formats or transmission systems; the file extension merely describes common packaging conventions.
Converting an MP3 to FLAC, WAVE, or an MP3 with a higher bitrate will not restore information lost during encoding; re-encoding with lossy compression may even introduce additional distortion. While spectral cutoff, encoder latency, or media information can provide clues about the source, none of these factors alone can prove that a file originated from a specific master or bitrate.