Lossless Audio
An audio data format or compression method capable of fully reconstructing the encoded digital audio samples from the input after decoding; the term "lossless" describes the reversibility of the data and does not directly indicate the source of the recording, the sampling specifications, or the subjective quality.
Explanation
无损音频 refers to audio that, after encoding, storage, or transmission, can be fully restored to its original digital sample data upon decoding. Uncompressed PCM can be stored losslessly, while formats such as FLAC, ALAC, WavPack, and MLP utilize audio correlations to reduce data volume while ensuring that the decoded result matches the input exactly.
“Lossless” describes a specific data relationship: the decoded output contains no information loss relative to the encoded input. It does not indicate whether the input originated from lossy encoding at an earlier stage, or whether it has undergone resampling, dynamic processing, or analog duplication, nor does it guarantee the quality of the master tape, recording, or playback equipment. If an MP3 file is decoded and saved as FLAC, the resulting FLAC file is lossless with respect to this specific input, but it cannot restore the data as it existed before MP3 encoding.
Lossless compression is typically achieved by predicting adjacent samples, utilizing channel correlations, and efficiently encoding the residuals. The more predictable the music, the higher the compression ratio tends to be; noise and complex high-frequency content are more difficult to compress. The encoding level can alter the search complexity and file size, but as long as the format and implementation are correct, the decoded sample rate should remain the same.
Whether a file is lossless cannot be determined solely by its file extension. The WAVE container can hold either PCM or lossy-encoded data; M4A can encapsulate AAC or ALAC; and some formats even offer both lossy and lossless modes. Effective verification methods include reading codec identifiers, decoding checksums, comparing sample hashes, or using the format’s built-in checksums. Spectrograms can only reveal partial clues about the source; low-pass filtering may also result from recording equipment, production filtering, or sample rate conversion, and cannot be used alone as proof of lossless quality.
Sampling rate and bit depth are independent of lossless status. Both 16-bit/44.1 kHz FLAC and 24-bit/192 kHz FLAC can be lossless; the difference lies in the parameters of the stored PCM. High-specification files may simply be upsampled versions of lower-specification content. Lossless transcoding can change the container, metadata, and compression ratio while preserving the sample data, whereas bit depth reduction, sample rate conversion, and loudness processing generate new data.