noiseshaping

Noise shaping is an audio processing technique used to improve the perceived quality of low-bitrate audio reproduction, particularly in digital systems with limited resolution. The method involves altering the frequency distribution of quantization noise—an inherent artifact of digital audio encoding—to make it less audible to human hearing. By emphasizing noise at frequencies where the human ear is less sensitive, noise shaping helps mask the distortion, resulting in a more natural sound despite the reduced bit depth.

The process typically employs a filter to attenuate noise in the mid-frequency range, where human perception is most acute, while allowing higher and lower frequencies to retain more noise. This is often achieved through a technique called pre-emphasis, where the input signal is boosted in the mid-range before quantization, and de-emphasis is applied during playback to restore the original frequency balance. Some advanced noise shaping methods, such as sigma-delta modulation, use feedback loops to dynamically adjust noise distribution in real time, further enhancing audio clarity.

Noise shaping is commonly applied in digital audio interfaces, compact disc players, and low-bitrate codecs like MP3 and AAC. While it does not eliminate quantization noise entirely, it effectively reduces its perceptibility, making it a valuable tool in digital audio production and playback systems. The technique is particularly useful in scenarios where high fidelity is impractical due to hardware limitations or bandwidth constraints.