This note describes quantization noise and how it can effect biquads in particular
Overview#
The difference between the ideal mathematical value and the actual value as processed in fixed-point arithmetic is known as quantization noise. There are many factors that can influence the level of the quantization noise, some of which that affect biquad filters are discussed below. This note focuses on Biquad filters as they have a feedback component in them that can amplify the noise.
Key Features#
Description of a biquad
Desciption of fixed point number formats and relation to quantization noise
Description of the quantization employed by the xcore.ai Vector Processing Unit (VPU)
Known Issues#
None
Required Tools#
None
Required Libraries (Dependencies)#
None
Support#
This package is supported by XMOS Ltd. Issues can be raised against the software at: http://www.xmos.com/support