Variable Sampling Rate Conversion and Polyphase Filter Implementation

This article presents the implementation of variable sampling rate conversion and polyphase filters. The program implements a complete variable sampling rate conversion pipeline that includes filter design, interpolation, and decimation processes. The filter design utilizes the window function method, where appropriate window functions are designed based on specific requirements to determine the optimal filter length. The core algorithm proceeds through interpolation, filtering, and decimation stages to generate the final resampled waveform file. A key aspect of this implementation involves the comparison between two convolution methods: direct convolution and polyphase decomposition convolution. The polyphase approach partitions the filter into parallel subfilters that operate at lower rates, significantly reducing computational complexity while maintaining signal integrity. Both methods are evaluated for their final output quality and computational efficiency in the resampling process, with the polyphase implementation demonstrating superior performance for real-time applications.