MutePoint
Home / Admin guides / Noise suppression

Turning On Teams Noise Suppression — Both Clients, One Page

By the MutePoint admin desk · Published

New client: Settings and more (…) → Settings → Devices → Noise suppression — a toggle, on by default, persisting across meetings. Mid-meeting: More actions (…) → Audio settings. Classic client still carries the old four-way control (Auto / High / Low / Off) in the same Devices area. If your control looks nothing like the screenshots you found, that's the redesign, not you.

New Teams client (classic noted where different), Windows & Mac — checked against official vendor documentation.
One setting, two dashboards

The same control in each client

New client — a toggle Noise suppressionSettings → Devices · on by default
Classic client — four levelsAuto shown: the sensible default.

1. Set it in the new client

  1. Click Settings and more (…) by your avatar → Settings.
  2. Open Devices in the left rail.
  3. Switch Noise suppression on under the microphone block.
  4. Close — the choice persists across calls and meetings.
Why this works: Teams filters non-speech sound before it leaves your machine; the new client collapsed the old levels into a managed toggle, trading control for fewer wrong configurations — an admin's trade if ever there was one.

2. Change it mid-meeting without dropping

  1. Toolbar → More actions (…).
  2. Audio settings.
  3. Flip Noise suppression; it applies live and sticks for next time.
Why this works: the in-meeting panel writes the same underlying setting — a mid-call fix is a permanent one, which cuts both ways when you later want music heard.
An admin's honest note

Suppression on, noise still through?

Policy locks and hardware gates mean Teams' own suppression isn't always yours to fix. Krisp works underneath the client — no admin consent needed for your own mic — and a 7-day free trial with no card gives you a full working week to evaluate.

Evaluate Krisp free for 7 days Krisp for Teams users — the full picture →

3. On classic Teams, pick the level deliberately

  • Auto — the sensible default; adapts to what it detects.
  • High — strips all non-speech; needs supported hardware and bypasses while music-sharing.
  • Low — persistent low hums only (fans, AC).
  • Off — for when the room's sound *is* the content.
Why this works: organizations still on classic get finer control at the price of more ways to be wrong — document your recommended level for the team and save yourself the tickets.

4. Know the voice isolation option where policy allows

  1. Settings → Devices → look for Voice isolation.
  2. Complete the one-time voice enrolment.
  3. Toggle on — Teams now prioritises your enrolled voice.
Why this works: standard suppression preserves every human voice by design; isolation profiles yours. Availability is tenant-policy-dependent — if it's absent, that's your admin, not your build.

Still stuck?

  • Toggle absent? Web client doesn't have it — use the desktop app.
  • Toggle reverts? Tenant policy pinning — a conversation with IT, not a settings hunt.
  • Music being shredded? Working as designed; disable suppression for the performance.

Quick answers

Does the setting cover calls as well as meetings?

Yes — one toggle governs both, and changes made mid-meeting persist afterwards.

Where did Low and High go in new Teams?

Consolidated into the managed toggle during the client redesign. Only classic Teams still exposes the four levels.