Skip to content

Audio Weaver Multicore

Demos

Audio Weaver Multicore on NXP RT685

Distribute audio processing across multiple compute cores, and assign each processing block in your audio signal flow to a specific core for execution.

Demonstrating Karaoke solution by distributing the ML based Vocal removal model running on HiFi4 and mic processing running on the M33.

CHALLENGES
  • Implementing the distributed audio processing on multi core.
  • Preserving music and removing a vocal from song
SOLUTIONS
  • Audio Weaver multicore implementation
  • Aizip ML based Vocal removal model
KEY FEATURES
  • Multicore (M33 & HiFi4) Audio processing (Implementing, debugging and tuning)
  • AiZip MSS (ML based vocal removal model)
  • Audio Weaver mic processing (Reverb)
USE CASES
  • Karaoke System: Realtime removal of vocals from any song. Mix the music with the live mic input. Add real time audio effects (reverb) on the mic input
AiZipi for Karaoke Demo (1)
Hardware specs:

NXP RT 685 SoC with M33 (27MHz) and HiFi4(596Mhz)

 

Processing specs:

M33 (Audio IO + Mic processing): 22%

HiFi4 (MSS model) : 44%

MSS model: 480KB

Get in touch with our team to see how you can elevate your audio designs