Digital Signal Processing with Swift
summary
Concrete pipeline: capture PCM with AVAudioEngine, use Accelerate/vDSP to run a 2048-frame FFT, return normalized magnitudes and plot peaks in UIKit in real time. Uses a Hann window, converts interleaved samples into vDSP split complex, sets log2n and calls vDSP_create_fftsetup and vDSP_fft_zrip, notes Nyquist sampling limits and that a circular buffer would improve accuracy. Playback demos show visible peaks; a Shazam-style fingerprinting implementation is attempted then deferred.
AI-written summary · about
chapters (8)
- 0:00Introduction to DSP
- 2:01The Fourier transform
- 3:01iOS audio frameworks
- 4:09Theoretical foundations
- 6:19Capturing audio input
- 8:08Processing audio signals
- 12:02Visualizing audio data
- 14:07Best practices and resources
related talks
01h 12mSpeeding Up with Accelerate: Experience Optimizing C++ Code
27mICYMI: Enums Are...
15mSwiftUI for React Devs via an NFT drum machine
18mPrototyping Custom UI in SwiftUI
26mShaping Sounds in Swift
16mSwiftUI API Design Lessons:Bridging Procedural APIs with theDeclarativeWorld
48mReal Physical Computing in Swift
25mAI-Assisted Swift in Gamifying an iOS App