Design and Hardware Implementation of Audio Filtering

Abstract

In order to isolate, amplify, and analyse particular frequency components within audio signals, audio filtering is a crucial step in digital signal processing. In this study, a real-time audio filtering system that uses several Finite Impulse Response (FIR) filters on a System-on-a-Chip (SoC) platform is designed and hardware realised. Four working modes are included in the proposed system: band-pass filtering, low-pass filtering, high-pass filtering, and direct signal transmission. Using the HDL Coder methodology, the filter architectures are created in MATLAB Simulink and automatically converted into hardware descriptions that can be manufactured. Real-time switching between the different filtering modes is made possible by a changeable intellectual property (IP) core, which enables the dynamic selection of the appropriate filter response without the need for hardware redesign or reconfiguration. System performance is evaluated using stereo audio signals sampled at 44.1 kHz. In order to guarantee compatibility with common audio processing needs, the filter coefficients are simultaneously constructed based on a reference sampling frequency of 48 kHz. The outcomes of the experiment show that unwanted frequency components can be effectively attenuated while maintaining the desired audio content\'s quality. For real-time embedded audio processing applications on SoC platforms, the implementation validates the applicability of FIR-based filtering approaches.

Introduction

This study presents the design and implementation of a real-time audio filtering system using Finite Impulse Response (FIR) filters on the Xilinx Zynq-7000 System-on-Chip (SoC) platform. Digital audio signals are often degraded by unwanted noise and frequency components, making filtering essential for applications such as speech enhancement, noise reduction, music processing, and communication systems. FIR filters are preferred because of their linear phase response, stability, and predictable performance. Unlike software-based processing, FPGA-based systems provide high-speed, low-latency, and deterministic execution suitable for real-time applications.

The proposed system supports three filtering modes: Low-Pass Filter (LPF), High-Pass Filter (HPF), and Band-Pass Filter (BPF). Incoming audio is processed by selecting the appropriate FIR filter coefficients through an AXI4-Lite control interface, while continuous audio streaming is handled using the AXI4-Stream interface. The entire system is designed in MATLAB/Simulink using HDL-based development and implemented on the Xilinx Zynq-7000 SoC, which integrates FPGA programmable logic with a dual-core ARM processor for efficient hardware–software co-design.

The system architecture consists of several functional blocks, including audio input, data conversion, filter control, FIR filtering, buffering, spectrum analysis, and audio output. Stereo audio sampled at 44.1 kHz is converted from floating-point to fixed-point format for efficient FPGA processing. Three 60th-order FIR filters, designed using the Blackman window method, are implemented for low-pass, high-pass, and band-pass filtering. After filtering, the signal is converted back to floating-point format for playback and spectral analysis. The HDL Workflow Advisor automates hardware code generation, IP core creation, synthesis, implementation, and deployment onto the FPGA.

Simulation and hardware testing confirmed that each FIR filter achieved the expected frequency response. The low-pass filter preserved low-frequency components while suppressing high frequencies, the high-pass filter retained high-frequency components and attenuated low frequencies, and the band-pass filter passed only the desired frequency range while rejecting frequencies outside the passband. Power spectrum and spectrogram analyses verified that the filters met their design specifications.

Hardware implementation on the Xilinx Zynq-7000 SoC demonstrated continuous real-time audio processing with negligible latency. The AXI-based communication successfully enabled dynamic filter selection and reliable interaction between the ARM processor and FPGA logic. Overall, the proposed system provides a flexible, high-performance, and low-latency platform for real-time digital audio filtering and offers a foundation for future applications such as adaptive filtering, speech enhancement, audio equalization, and active noise cancellation.

Conclusion

The design and hardware implementation of FIR-based audio filters utilizing the SoC platform was the main emphasis of the work reported in this project. Using the MATLAB FDA Tool, low-pass, high-pass, and band-pass filters were created and incorporated into the hardware design. The developed filters met the necessary frequency-domain parameters, as confirmed by both simulation and experimental tests. Effective filter selection was made possible via the AXI-Lite interface, and the active working mode was visually confirmed by LED indications. In real-time operation, the system showed consistent performance and precise audio filtering. In general, the experiment confirmed that SoC-based FIR filters work well for digital audio signal processing applications.

References

[1] J. Yu, H. Zhong and P. Yan, \"Implementation of FIR filter based on Xilinx IP core,\" 2020 2nd International Conference on Information Technology and Computer Application (ITCA), Guangzhou, China, 2020, pp. 79-85, doi: 10.1109/ITCA52113.2020.00024. [2] A. Anand and R. B, \"Design and Implementation of Real-Time Audio Processing on Zynq-7000,\" 2023 International Conference on Computing, Communication, and Intelligent Systems (ICCCIS), Greater Noida, India, 2023, pp. 286-291, doi: 10.1109/ICCCIS60361.2023.10425313. [3] Proakis, J. G. (2007). Digital signal processing: principles, algorithms, and applications, 4/E. Pearson Education India. [4] Crockett, L. H., Elliot, R. A., Enderwitz, M. A., & Stewart, R. W. (2014). The Zynq book: embedded processing with the ARM Cortex-A9 on the Xilinx Zynq-7000 all programmable SoC. Strathclyde Academic Media. [5] Crockett, L. H., Elliot, R. A., & Enderwitz, M. A. (2015). The zynq book tutorials for zybo and zedboard. Strathclyde Academic Media. [6] The MathWorks, \"Implement an Audio Filter on a Zynq Board,\" in HDL Coder™ Documentation, The MathWorks, Inc., Natick, MA, USA, 2026. [7] The MathWorks, “HDL Coder Documentation,” The MathWorks, Inc., Natick, MA, USA.

Copyright

Copyright © 2026 Kandukuri Harika, Dr. M. Madhavi Latha . This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.