Overview
Explore the theory behind the Short-Time Fourier Transform (STFT) in this 35-minute video tutorial. Gain a clear understanding of this crucial tool for AI audio and music engineering, essential for extracting spectrograms used in deep learning audio models. Learn about the Fourier Transform problem, signal segmentation, STFT intuition, windowing techniques, overlapping frames, and the transition from DFT to STFT. Discover STFT outputs, time/frequency trade-offs, and key parameters. Visualize sound through spectrograms and understand their importance in audio processing. Access accompanying slides for further study and join a community of AI audio enthusiasts to enhance your learning experience.
Syllabus
Intro
Join the community!
Fourier Transform Problem
CONSIDER SMALL SEGMENTS OF THE SIGNAL
STFT intuition
Windowing
Overlapping frames
From DFT to STFT
Outputs
Example STFT output
Time / frequency trade off
STFT parameters
Hann window
Visualising sound
Spectrogram
What's up next?
Taught by
Valerio Velardo - The Sound of AI