HoloSing

Karaoke in AR :)

Comment

Inspiration

Karaoke is fun—but it becomes magic when it blends with immersive technology. We wanted to make karaoke more than just reading lyrics on a screen. We imagined singing inside an AR environment where lyrics float in front of you, allowing people to sing, entertain, and practice songs whenever, wherever. Our goal was simple: bring karaoke into the world of spatial computing and build the foundation for an XR-powered social music experience.

What it does

HoloSing is a karaoke experience built with Unity for the XREAL Beam to be used with the XREAL glasses that lets users:

  • Search for songs directly from YouTube

  • Extract clean audio automatically

  • Display synchronized lyrics pulled from captions (SRT, VTT, JSON)

It’s a full end-to-end karaoke pipeline—from YouTube search → backend processing → Unity playback → real-time lyric sync—all running in one seamless flow.

How we built it

Unity

  • Handles the karaoke experience—search UI, playback controls, subtitle parsing, and lyric display

  • Uses TextMesh Pro for crisp typography

  • Integrates AR Foundation + XREAL XR support for future AR overlays

Backend (Node.js / Express)

  • Uses ytdl-core to extract audio

  • Fetches subtitles via youtube-captions-scraper

  • Provides a REST API that Unity calls to retrieve audio + lyrics

Challenges we ran into

  • Syncing lyrics across formats — SRT, VTT, and YouTube captions each behave differently

  • Latency between backend and Unity — Extracting audio + captions takes time; we optimized pipelines by caching the files for demonstration.

  • API limitations — We first tried to find karaoke specific APIs to be used, but they all required permission to be given beforehand.

  • AR integration — Coordinating Unity's AR stack with the XREAL SDK took a lot of iteration

What we learned

  • Best practices for clean backend → Unity pipelines
  • AR Foundation + XREAL plugin setup
  • Managing Unity’s Android build ecosystem

What’s next for Vibers

Karaoke is inherently a social event, and our next steps are to build out multiplayer integration and audio inputs. We want to allow users to connect with each other remotely, sing together with each other, and even have a karaoke scoring system for a little bit of competitive fun. Imagine singing with floating holographic lyrics anywhere—your room becomes the stage.

Built With

Share this project:

Updates