Missing first words in transcriptions
Last updated: July 30, 2026
Available on: Mac, Windows, iOS, Android
Missing the first few words of your dictation? This is almost always a wireless microphone wake-up delay — wireless mics like AirPods can be slow to start, so anything you say in that window isn't captured. Most people fix this in under 2 minutes.
Quick checks
Pause briefly before speaking. Wait a moment after pressing the hotkey before you start talking — especially on wireless mics.
Hold the hotkey long enough. Releasing too quickly cancels the dictation. Hold the key while you speak, then release.
Check whether you're on a wireless mic. Open Flow Settings → General → Microphone. If AirPods or a wireless headset is selected, that's almost certainly the cause — continue to the steps below.
Does your mic work in other apps? Record a short clip in Voice Memos (Mac/iOS) or Sound Recorder (Windows). If it's silent there, the problem is system-level, not Flow.
Note: Flow shows a notification when something is blocking the mic, each with a button that takes you straight to the fix. Watch for Microphone Permission Required (Mac), Microphone access is disabled (Windows), No microphone detected, Microphone unavailable (another app is holding the mic, or the driver is stuck), Microphone error, Microphone disconnected, or a silence notification such as Is your microphone muted?, We couldn't hear you, or Microphone is not working. If your mic disconnects mid-dictation, the Insert button on the notification inserts the text recorded up to that point.
How to fix missing first words
Step 1: Switch to a wired or built-in microphone
Built-in laptop and phone microphones, and wired USB microphones, have no wake-up delay. This is the most reliable fix.
Mac and Windows
Choose your microphone directly in Flow:
Open Flow Settings → General → Microphone.
Select your built-in microphone or a wired USB microphone. AirPods appear with a ⚠ and a short warning; other wireless devices appear without a warning indicator.
Dictate a test phrase. If all your words appear, you're done.
iOS and Android
Mobile has no in-app mic picker on Android, so switch the device at the system level:
Disconnect your wireless device in system Settings → Bluetooth, or plug in a wired accessory.
Dictate a test phrase. If all your words appear, you're done.
Note: On iOS, Flow defaults to the built-in microphone via the Use Built-In Mic toggle in Flow Settings → Audio. Turn it off if you want to use a wireless headset mic. On Android, Flow uses the system default microphone.
Important: If you use your Mac with the lid closed (clamshell mode), the built-in microphone won't work. On macOS 12 and later, Flow detects clamshell mode and disables the built-in mic in the picker with an explanatory tooltip. On macOS 11, clamshell mode isn't detected — if your lid is closed there, select an external USB or wireless mic manually.
Tip: In the Settings mic picker, use Edit Ranking mode to drag and drop your microphone priority order and forget disconnected devices. When you plug in a new mic, Flow also prompts you to switch ("[device name] detected — Switch / Don't Switch"), and tells you when a previously selected mic is unavailable.
Step 2: Pause briefly before speaking
If you'd rather keep using a wireless mic, give it time to wake up:
Press your dictation hotkey.
Wait a moment before speaking so the mic can finish starting up.
Dictate a test phrase. If all your words appear, this workaround is enough for you.
Tip: Start with a filler word like "so" or "okay." Smart Formatting removes these automatically.
Note: Flow discards a small amount of audio at the very start of every recording, no matter which microphone you use. That baseline sits on top of any wireless wake-up delay, so a brief pause helps on wired mics too.
Step 3: Select your microphone explicitly (Mac and Windows)
Auto-detect can add a connection delay. Naming your device directly removes it:
Open Flow Settings → General → Microphone.
Select your specific microphone from the list instead of Auto-detect. Auto-detect shows the active device name in parentheses (for example, "Auto-detect (MacBook Pro)").
Dictate a test phrase. If all your words appear, you're done.
Step 4: Check system audio settings (Windows)
Open Windows Settings → System → Sound → Input.
Confirm your preferred microphone is set as the default input device.
Click Test your microphone and speak — the level meter should move.
Return to Flow and dictate a test phrase. If all your words appear, you're done.
Step 5: Verify what was actually recorded
This tells you whether the problem is recording or transcription:
Open the Flow Hub and go to the Home page.
Find your recent dictation in the Activity list, click the three-dot menu (⋮) on the dictation card, and select Extract audio.
Play the downloaded audio file. If the audio has all your words but the transcript doesn't, contact support. If the audio is also missing the first words, it's a mic wake-up delay — return to Step 1.
Note: Extract audio is unavailable for recordings older than 14 days, and Flow shows a message such as "Cannot download audio as it is no longer available (older than 14 days)". On enterprise Never Store (Zero Data Retention) plans, audio is never saved. If dictation history is turned off, the Hub shows "History disabled" — either because you set a Never Store policy yourself (change it in Settings → Data and Privacy) or because your organization has disabled retention of your dictation history.
Common issues
Bugs fixed in recent updates — update Flow to resolve
These issues were fixed in recent Wispr Flow updates. Update to the latest version to resolve them:
Text missing even though the audio recorded correctly: Flow's transcription quality check was too aggressive and discarded good transcripts, especially when audio was quiet or contained numbers. In severe cases, up to a full minute of dictated text could be lost.
Accidental short taps generating unexpected text: Very brief, unintentional taps could insert text that was never spoken. Short accidental taps now resolve cleanly with no output.
First dictation after opening the app failed or cut off words (iOS): The first dictation after launching Flow on iOS could fail silently or miss opening words while the microphone finished starting up.
Dictation failed on the first attempts after a cold start (iOS): Starting dictation from the keyboard, the action button, or quick-note shortcuts right after a cold start could fail once or twice before succeeding.
iOS keyboard showed an "Audio was silent" warning: When a dictation captured no speech, the iOS keyboard displayed a warning. It now resets quietly back to the microphone.
Microphone slow to activate after canceling and quickly retrying (Mac and Windows): Pressing the hotkey, canceling, and pressing again could make the next dictation behave erratically or fail to start.
Text missing when dictating several times in a row: A timing issue when starting a new dictation immediately after finishing one caused text to be dropped.
Empty results pasted during dictation: Text insertion occasionally produced an empty result.
FAQs
My dictation cut off — did my mic drop?
Desktop dictations stop at 20 minutes, with a warning notification at 19 minutes. Everything recorded up to the stop point is still transcribed and pasted — nothing is discarded. For longer captures, break the dictation into multiple sessions.
How do I check the live volume indicator?
Open Flow Settings → General → Microphone and click Change (labeled Change microphone during onboarding). A live volume bar appears next to the selected device — speak and watch it move to confirm Flow is receiving audio.
How does hands-free mode work?
Double-tap your dictation hotkey to start hands-free mode. Flow keeps recording until you press the hotkey again to stop and submit. Press Esc (or your custom dismiss shortcut) to discard any dictation, hands-free or not.
My dictation produced no audio on Mac — what happened?
A dictation can capture no audio if the mic was briefly unavailable, so that dictation produces no text and the next one works normally. If it keeps happening, select your microphone explicitly (Step 3) and check the notifications Flow shows you.
On iOS, I spoke but nothing was typed and no error appeared — what happened?
When the iOS keyboard detects no speech, it resets quietly back to the microphone. Speak a little louder, wait a moment after the keyboard appears before speaking, or switch to the built-in microphone in Flow Settings → Audio.
Why don't I see my virtual microphone in the list?
Virtual microphones, such as those created by audio routing software, are hidden from Flow's list unless they're your system default. Set the virtual mic as the system default to make it selectable.
Which microphones work best with Flow?
Built-in, wired USB, and wired 3.5mm mics have no wake-up delay and are the most reliable. AirPods and wireless headsets vary, so pause before speaking when you use them.
In the picker, your built-in mic is labeled Built-in mic (recommended). Only AirPods are flagged with a ⚠ — Beats-branded Bluetooth devices and other wireless headsets are not flagged, even though the same pause-before-speaking advice applies.
Still stuck?
If first words are still missing after these steps, reach out to our support team:
Mac: Press Cmd+/, or click Help in the menu bar or Hub sidebar → Talk to support.
Windows: Click Help in the Hub sidebar → Talk to support.
iOS: Open the menu → Talk to Support. Android: Open the drawer menu → Report an issue.
Include your platform, microphone type, and what you've already tried. If the extracted audio contains words the transcript is missing, say so — most mic issues are resolved in one reply.