DSD Audio Noise Canceling Without Quantization Noise
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing noise canceling systems using direct stream digital (DSD) audio signals with high sampling frequencies and low quantization bits suffer from degradation of audio characteristics due to added quantization noise, which degrades the sound quality.
Innovation Solution
A signal processing apparatus and method that includes delta sigma modulators, filters, and bit expanders to process audio signals without passing them through delta sigma modulators that cause quantization noise, ensuring the audio characteristic is preserved while suppressing external noise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a noise canceling system uses DSD audio signals with high sampling frequency and low quantization bits, then the audio signal can be processed efficiently, but quantization noise is added that degrades audio characteristics
Solution Approach 1:
The patent divides the noise canceling system into separate signal paths: one for the DSD audio signal that bypasses the delta-sigma modulator, and another for the noise canceling signal. This segmentation allows the audio signal to avoid quantization noise while the noise canceling signal undergoes normal processing.
Solution Approach 2:
The patent introduces an intermediary approach by processing the noise canceling signal separately and adding it to the original DSD audio signal after the audio signal has bypassed the quantization-prone delta-sigma modulator stage, thus preventing quantization noise from being added to the audio signal.
2Object-affected harmful factors
If the audio signal passes through a delta sigma modulator for noise canceling processing, then external noise can be suppressed, but quantization noise is introduced that degrades sound quality
Solution Approach 1:
The patent extracts the audio signal from the delta-sigma modulator processing path, allowing it to bypass the quantization stage entirely. The noise canceling function is achieved through separate processing of the noise signal, which is then combined with the preserved audio signal.
Solution Approach 2:
Instead of passing the audio signal through the delta-sigma modulator and then removing noise, the patent inverts the approach by preserving the original audio signal and separately processing the noise canceling signal to be added afterward, thus avoiding quantization noise from the beginning.
3Manufacturing precision
If quantization bit number is increased to reduce quantization noise, then audio characteristic is improved, but device complexity and processing overhead increase
Solution Approach 1:
The patent changes the processing approach rather than simply increasing quantization bits. By routing the audio signal to bypass the delta-sigma modulator and processing noise canceling separately, the system maintains audio quality without requiring increased bit depth or more complex processing architecture.
Data Source
AI summary
A signal processing apparatus is provided which can suppress external noise without degrading an audio characteristic.Provided is a signal processing apparatus including: an A/D converter configured to output a digital signal having a predetermined sampling frequency and quantization bit number a, the A/D converter including a first delta sigma modulator that performs a first delta sigma modulation process on an input analog signal; a filter unit configured to pass an output of the A/D converter through a digital filter provided with a predetermined filter characteristic and output a digital signal having the sampling frequency and a quantization bit number b; a second delta sigma modulator configured to perform a second delta sigma modulation process on an output of the filter unit and output a digital signal having the sampling frequency and the quantization bit number a; and an addition unit configured to add an output of the second delta sigma modulator and an input digital signal having the sampling frequency and the quantization bit number.