Psychoacoustic Masking Model
Definition
The perceptual model used by lossy codecs such as MP3 and AAC to decide which frequency components can be discarded or coarsely quantised without audible loss. The specific masking thresholds and sub-band structures differ by codec family, leaving identifiable spectral patterns in the encoded file.
- Applies to
- Lossy audio codecs such as MP3 and AAC
- Function
- Decides which frequency components to discard or coarsely quantise
- Forensic use
- Codec identification from residual spectral artefacts
Common questions
Why does the masking model help identify which codec encoded a file?+
Each codec family implements its own masking thresholds and sub-band splits, leaving characteristic quantisation artefacts and frequency cutoffs in the compressed signal even after it is converted to an uncompressed format.
Can this model help detect that a file was re-encoded after editing?+
It cannot directly show an edit, but comparing the spectral fingerprint expected from a claimed codec against what is actually observed can flag transcoding or re-encoding inconsistent with the file's stated origin.
Related terms
- Double Compression
- The condition in which a video has been encoded at least twice through a lossy codec. The term is also used loosely...
- MEMS Microphone
- A Micro-Electro-Mechanical Systems microphone: the capsule type used in virtually all modern smartphones and tablets. MEMS microphones have a characteristically flat midrange...
- Quantisation Noise
- The distortion introduced when a continuous audio signal is rounded to discrete digital values. Lossy codecs concentrate quantisation noise at predictable frequency...
- Self-Noise (Noise Floor)
- The electrical noise inherent in a microphone circuit and capsule, measured in dB(A) equivalent input noise. Different microphone types have characteristic noise...
- Spectral Fingerprint
- The characteristic frequency response profile of a recording device, including roll-off curves, resonance peaks, and self-noise level. Comparing a recording's spectral fingerprint...