Audio Signal Compression Using Wavelet Transform: A MATLAB Implementation
Session
Computer Science and Communication Engineering
Description
In modern signal processing, the need to efficiently analyze and store non- stationary signals has become increasingly important, particularly for audio applications where both time and frequency localization are essential. Traditional Fourier-based methods provide global frequency information but fail to indicate when these frequencies occur, making them suboptimal for real-world, time-var- ying signals. This paper investigates the use of wavelet transforms, specifically the Daubechies wavelet for audio signal compression. A practical implementa- tion is developed in MATLAB, where the discrete wavelet transform (DWT) is applied to decompose audio signals into approximation and detail coefficients. Thresholding techniques are employed to remove insignificant coefficients, re- ducing data size while preserving perceptual quality. Results include coefficient visualization, compressed signal reconstruction, and Power Spectrum Density (PSD) analysis comparing original and compressed signals. The method achieves significant compression ratios without perceptible quality loss, demonstrating wavelet transform’s potential for audio data reduction.
Keywords:
Wavelet Transform, Audio Compression, MATLAB, Daubechies, Time-Frequency Analysis, Signal Processing
Proceedings Editor
Edmond Hajrizi
ISBN
978-9951-982-41-2
Location
UBT Lipjan, Kosovo
Start Date
25-10-2025 9:00 AM
End Date
26-10-2025 6:00 PM
DOI
10.33107/ubt-ic.2025.111
Recommended Citation
Pira, Rigon; Azizi, Argjend; Halimi, Gentrit; and Tafa, Zhilbert, "Audio Signal Compression Using Wavelet Transform: A MATLAB Implementation" (2025). UBT International Conference. 43.
https://knowledgecenter.ubt-uni.net/conference/2025UBTIC/CS/43
Audio Signal Compression Using Wavelet Transform: A MATLAB Implementation
UBT Lipjan, Kosovo
In modern signal processing, the need to efficiently analyze and store non- stationary signals has become increasingly important, particularly for audio applications where both time and frequency localization are essential. Traditional Fourier-based methods provide global frequency information but fail to indicate when these frequencies occur, making them suboptimal for real-world, time-var- ying signals. This paper investigates the use of wavelet transforms, specifically the Daubechies wavelet for audio signal compression. A practical implementa- tion is developed in MATLAB, where the discrete wavelet transform (DWT) is applied to decompose audio signals into approximation and detail coefficients. Thresholding techniques are employed to remove insignificant coefficients, re- ducing data size while preserving perceptual quality. Results include coefficient visualization, compressed signal reconstruction, and Power Spectrum Density (PSD) analysis comparing original and compressed signals. The method achieves significant compression ratios without perceptible quality loss, demonstrating wavelet transform’s potential for audio data reduction.