Noise shaping is a digital signal processing technique that modifies the spectral distribution of quantization error by feeding the quantization error through a feedback filter. Parameters: (1) the input signal is quantized at fixed bit depth; (2) the quantization error — the difference between the unquantized and quantized signal — is passed through a shaping filter (typically high-pass or custom); (3) the filtered error is fed back and added to subsequent samples before quantization, creating correlated (colored) noise that pushes energy away from sensitive frequency bands. Persistence mechanism: implemented as digital filter algorithms in DAC firmware, audio processing software, and embedded quantizers; maintained through engineering standards, textbooks, and industry practice in digital audio. [formal: noise_shaping | substrate: behavior | horizon: a life | explicit: yes | epoch: 0.01]
Accepted ontology entry
noise-shaping
Noise shaping is a digital signal processing technique that modifies the spectral distribution of quantization error by feeding the quantization error through a feedback filter. Parameters: (1) the input signal is quantized at fixed bit de…
Definition
Why it is in scope
Human-made digital signal processing technique: a method that shapes the spectral distribution of quantization noise by feeding back quantization error through a filter, pushing noise energy to less perceptually sensitive frequency bands. Built to persist as a standard technique in audio DAC design and digital audio processing.
Names and aliases
- noise-shapingen · CANONICAL
Relations from this entry
- cmsnet9gh03q31q131146onk3DEPENDS_ON →
Noise-shaping needs quantization to operate now — it shapes quantization error by feeding it back through a filter. Remove quantization and noise-shaping has no error to shape. The removal test (Law 8) passes.
- cmsnet9gh03q31q131146onk3SERVES →
noise-shaping serves quantization — it is a technique designed to improve quantized audio by redistributing quantization noise to less perceptually salient frequency bands. Its purpose by design is to make quantization better.
- cmsnet9gh03q31q131146onk3DERIVED_FROM →
Noise shaping evolved from the study of quantization error. Quantization is the fundamental concept that predates noise shaping; noise shaping is a specialized technique that shapes quantization error spectra, building directly on quantization theory.
- cmsrzc22l00r2h7yukrx2iasoDERIVED_FROM →
Dither (random noise added before quantization, 1949) preceded noise shaping. Noise shaping (1970s) extends dither by controlling the spectral distribution of quantization noise — it evolved from the dither concept, with dither existing first and feeding into noise shaping.
Relations to this entry
- sigma-delta-modulation← DEPENDS_ON
Sigma-delta modulation requires noise shaping to operate — the feedback loop integrates the quantization error and shapes its spectral distribution, pushing quantization noise out of the signal band. Remove noise shaping and the modulator becomes a simple 1-bit quantizer with no advantage over uniform quantization at the Nyquist rate. Removal test per Law 8: without noise shaping, sigma-delta modulation ceases to be sigma-delta modulation. Epoch 0.5 > 0.31, direction confirmed.
Record identity
- Created
- Aug 13, 2026, 9:50 PM UTC
- Content hash
- 243b0a82963a3fffffc3e156d72093729bebbeaab98304dab7454b44cec0b17b