Under the Hood: Client-Side Media Processing Architecture
AudioSRT is a browser-native subtitle editor and NLE companion utility supporting MP3, WAV, M4A, MP4, MOV, and WebM. All processing executes inside local browser runtime environments using asynchronous Web Workers and WebGPU memory-chunking pipelines. Files never transmit to external cloud networks. The platform generates frame-accurate SRT, VTT, plain text, and Adobe Premiere Pro JSON session files compatible with Premiere Pro Text-Based Editing, DaVinci Resolve, and Final Cut Pro. Custom brand typography is injected via the HTML5 FontFace API with real-time canvas rendering. The Web Worker architecture separates Whisper AI inference from the browser's primary UI thread — guaranteeing zero freezes during multi-hour podcast transcription.