Performance Implications of Frequency Domain Crossfading for Time-Varying FIR Filtering
ADC - Audio Developer Conference via YouTube
AI, Data Science & Cloud Certificates from Google, IBM & Meta
Live Online Classes in Design, Coding & AI — Small Classes, Free Retakes
Overview
Google, IBM & Meta Certificates – 40% Off
One plan covers every Professional Certificate on Coursera.
Unlock All Certificates
Explore a technical conference talk examining the performance trade-offs between time-domain and frequency-domain crossfading techniques for real-time FIR filtering applications. Learn how exchanging impulse responses during runtime typically requires double the operations when using time-domain crossfading, as it involves filtering input signals with both impulse responses separately before blending. Discover the potential benefits and limitations of implementing crossfades in the frequency domain, which could eliminate one IFFT operation but introduces computational costs through direct convolutions. Understand the specific conditions where carefully selected blending functions might make frequency domain convolution more efficient than traditional time-domain multiplication approaches. Presented by iustin sandu, a software engineer at d&b audiotechnik specializing in algorithm design and optimization, this 18-minute presentation from the Audio Developer Conference (ADC) 2024 delivers practical insights for audio developers working with real-time DSP systems.
Syllabus
Performance Implications of Frequency Domain Crossfading for Time-Varying FIR Filtering - ADC 2024
Taught by
ADC - Audio Developer Conference