Tutorials, deep dives & use cases
Everything we know about stem separation, audio production workflows, and making AI work for real creators.
How AI Stem Separation Actually Works
A plain-English breakdown of source separation — from classic NMF methods to Meta's HTDemucs Hybrid Transformer architecture that powers AudSep today.
The DJ's Guide to Isolating Acapellas in 60 Seconds
Stop hunting for official vocal stems. Learn how to rip a clean acapella from any track in under a minute using AudSep — and blend it live with perfect pitch.
Remove Background Noise from Podcast Recordings
HVAC hum, street noise, and keyboard clicks ruin podcast audio. Here's a step-by-step workflow for getting broadcast-clean dialogue using AudSep's denoiser mode.
Sampling in 2026: How Producers Use AI Stems
Modern beat makers don't flip vinyl anymore — they flip stems. We explore how producers use AudSep to legally sample elements from reference tracks, then rebuild them into something entirely new.
How Music Teachers Use Stem Separation in the Classroom
From transcribing solos to teaching chord voicings, isolated stems are a secret weapon for music education. Here's how teachers have incorporated AudSep into lessons.
Build a Perfect Karaoke Track in 3 Steps
Upload. Remove vocals. Export instrumental. It really is that simple. We walk through the exact settings to get a clean, full-band karaoke backing track with no bleed-through.
HTDemucs vs MDX-Net: Which Model Should You Use?
Both models are excellent, but they excel in different scenarios. We ran 200 test tracks through both and break down exactly when you should pick each one.
How Film Composers Use Stem Separation for Reference Tracks
When a director says "I want it to sound like Zimmer", composers need to study that reference deeply. Here's how to tear apart a film score stems-by-stem to understand what makes it tick.