Audio Filter Using FFT
Integrated multiple sensors, PLCs, and HMI systems to create an automated manufacturing cell with predictive maintenance capabilities.
Full write-up
This project focused on the design and implementation of a robust digital signal processing system aimed at enhancing audio clarity by removing unwanted environmental and equipment-induced noise. The primary goal was to develop a tool capable of isolating and suppressing disturbances—such as background chatter, high-frequency hiss, or low-frequency hums—while preserving the integrity and naturalness of the desired audio signal.
Technical Implementation
The system was built using Python, utilizing the scipy.signal library for processing and tkinter for a graphical user interface (GUI). I implemented 4th-order Butterworth filters (Low-pass, High-pass, Band-pass, and Band-stop) due to their maximally flat frequency response in the passband, which minimizes signal distortion. To ensure high-fidelity results, the project utilized zero-phase filtering (applying the filter forward and backward) to prevent phase shifts that could degrade audio quality. Fast Fourier Transform (FFT) algorithms were integrated to convert signals from the time domain to the frequency domain, allowing for the precise identification of noise frequencies.
Key Concepts Utilized
Digital Filter Design:
Configuring Cutoff frequencies, filter orders, and filter types (IIR Butterworth) to target specific noise profiles.
Frequency Domain Analysis:
Leveraging FFT and spectrograms to visualize and analyze the frequency spectrum of signals. * Signal Processing Techniques: Time-domain waveform analysis, signal normalization, and zero-phase filtering to maintain signal phase relationships.
Engineering Programming:
Application ofNumPy,SciPy, andMatplotlibfor mathematical modeling and data visualization.
Outcome
The project successfully demonstrated effective noise attenuation, verified through comparative spectrogram analysis which showed the elimination of high-energy noise bands. The final application provided a practical platform for cleaning audio signals without compromising the intelligibility of speech or the quality of music