РусскийRUEnglishEN

Windows permissions

In brief

The main difference between the Windows client and the macOS one: the only permission that requires user action is microphone access. Whisperer captures system audio (the other person's voice) automatically — no separate permission is needed for that. Screenshots and global hotkeys on Windows also work without permissions.

In other words: on Windows, granting microphone access once is enough, and both sides of the call will be heard. This is simpler than on macOS, where the other person's voice requires a separate "Screen Recording" permission.

When to use this

  • On first launch of the client, when the permission wizard opens.
  • When your voice ([Me]) is missing from the transcript — the microphone wasn't granted.
  • When you can't hear the other person ([Other]) — the problem is usually in the choice of output device, not in permissions.
  • After a Windows update or reinstalling the app.

Permissions and audio capture on Windows

What Permission needed? How it's captured
Microphone (your voice, [Me]) Yes — Settings → Privacy & Security → Microphone Microphone capture, audio
System audio (the other person's voice, [Other]) No System audio capture from the default output device
Screenshots (vision) No PNG in %USERPROFILE%\Downloads\Whisperer_Screenshots\ and the clipboard
Global hotkeys No System-level key interception
No-logs / "stealth" (window excluded from screen capture) No, but requires Windows 10 2004+ (build 19041) Window excluded from screen capture

Step-by-step

  1. Launch the permission wizard. On first start, Whisperer opens the wizard (PermissionWizard) if microphone access isn't granted. Follow its prompts — that's the easiest path.
  2. Grant microphone access manually (if needed). Open Settings → Privacy & Security → Microphone. Turn on the master toggle "Microphone access", the toggle "Let apps access your microphone", and, if present, the toggle for desktop apps ("Let desktop apps access your microphone") — Whisperer falls under it.
  3. Restart Whisperer. After granting access, quit and reopen the app so the change applies.
  4. Check the other person's audio. System audio is taken only from the default output device. If the other person isn't heard, see the steps below.
  5. If the other person isn't heard, check the default output device. Right-click the sound icon in the tray → "Sound settings" (or Settings → System → Sound). Make sure the default output device is the one the call actually plays through (the speakers/headphones you hear). Loopback captures only that device.
  6. Check the transcript. Start a trial session and make sure both roles appear — [Me] (microphone) and [Other] (system audio).
  7. (Optional) Enable no-logs/stealth. Hiding the window from screen capture works on Windows 10 version 2004 and newer. On older builds the toggle is there but has no effect.

Screenshots

📸 [Screenshot: the Windows client's permission wizard (PermissionWizard) — microphone access status]

📸 [Screenshot: Settings → Privacy & Security → Microphone with the toggles on]

📸 [Screenshot: Sound settings — choosing the default output device]

Common mistakes

  • Microphone not granted — no voice from you. The transcript is empty on the [Me] side. Turn on access in Settings → Privacy & Security → Microphone and restart the app.
  • Can't hear the other person — wrong default output device. System audio is taken only from the default device. If the call plays through headphones but speakers are set as default, the other person won't be captured. Set the right device as default. (Picking a specific device inside the client isn't implemented yet.)
  • Expecting it to work "like on macOS". On Windows, system audio does not need the "Screen Recording" permission — the audio is captured automatically. Don't go looking for such a toggle.
  • Stealth/no-logs doesn't work. Hiding the window from screen capture requires Windows 10 2004+ (build 19041). On older versions the toggle has no effect — update Windows.
  • The desktop-apps toggle is off. On some Windows builds, access for desktop apps is enabled by a separate toggle — check it if the master permission isn't enough.

Best practices

  • Go through the permission wizard on first launch — it covers the single required item (microphone).
  • Remember the labeling logic: [Me] = microphone (permission needed), [Other] = system audio (no permission needed, but the correct default output device matters).
  • Before an important call, do a short trial session and make sure both roles appear in the transcript.
  • Don't switch the default output device in the middle of a session — it will break capture of the other person's voice.

Related articles