Google Meet rolls out multi-device adaptive audio merging

Original Article ↗ Hacker News Discussion ↗

Feature overview & availability

  • New Meet feature merges audio from multiple nearby devices to act like one shared mic/speaker array, aiming to fix echo/howling in hybrid rooms with many laptops.
  • Several commenters who tested it say it “just worked” and solved a long‑standing pain point.
  • Others ask whether similar tech has existed for years in products like around.co, and whether it will come to Zoom/Teams.
  • The feature is gated behind high‑tier “Gemini” Workspace plans or an AI add‑on; many see this as paywalling a non‑AI feature to pad Gemini revenue.

Echo, feedback, and duplex issues

  • Many users complain more generally about hearing their own voice (echo) due to remote participants using speakers + open mics.
  • Practical tips: briefly muting to “reset” echo cancellation; pausing both sides for a few seconds so algorithms relearn silence.
  • Debate over whether major platforms truly support natural full‑duplex. Some say all do, but noise/echo cancellation and “last‑N speakers” limits effectively make calls semi–half‑duplex.

Meet vs Zoom/Teams and others

  • Strong split: some consider Meet the best balance of simplicity, speed, browser‑only operation, calendar integration, and “just works” behavior.
  • Others find Zoom superior for:
    • Higher bitrate and clearer screen sharing (especially code),
    • Better noise/echo handling and audio options,
    • Layout control (true fullscreen with faces overlay), multiple displays, remote control, music/teaching.
  • Teams is often criticized as heavy, buggy, and UX‑complex but praised where deeply integrated with chat, channels, and document permissions.
  • Several lament Meet’s limited layouts, inability to easily get 1:1 full‑screen sharing, and weak whiteboarding/annotation vs Zoom/Slack.

Hybrid and remote meeting practices

  • Some organizations discourage hybrid meetings altogether (everyone joins individually) to avoid audio/participation asymmetries; others say this is unrealistic and prefer better tech like this feature.
  • Common problems: bad conference‑room hardware, limited meeting rooms, people joining from desks in open offices, and inconsistent policies across big enterprises.

Technical challenges & skepticism

  • Multiple commenters are impressed by the audio engineering: cross‑device sync, multi‑mic echo cancellation, and array‑like processing in heterogeneous laptops.
  • Debate over implementation:
    • Some think it requires heavy cloud processing or even neural nets; others argue classical DSP and level management may suffice.
    • Sync via inaudible/ultrasonic signals is suggested; others think normal audio correlation is enough.
  • A few are skeptical it will work robustly outside controlled demos and note that audio issues often stem from users not muting or not using headphones.