1. Choose your recording app: Windows, Mac, or web

You do not need paid software to record your voice: Windows includes a recorder, while Mac offers Voice Memos and QuickTime Player. Use a browser recorder if you cannot install an app, and check that it lets you download the finished audio.

Windows recording apps save files automatically. On Mac, Voice Memos keeps an app library with optional iCloud sync, while QuickTime lets you save a standalone file to a folder you choose.

Recording apps, file formats, and save locations
App Best fit Audio format Where the audio goes
Sound RecorderWindows 11 Everyday voice recording with format choices AAC (.m4a), MP3, or WAV; options vary by app version Documents > Sound Recordings by default
Voice RecorderWindows 10 Simple voice notes with automatic saving M4A Documents > Sound Recordings by default
Voice MemosmacOS Quick notes and optional iCloud sync M4A when exported to Finder App library; drag a recording into Finder to export it
QuickTime PlayermacOS Direct input, quality, filename, and folder control High: AAC in M4A
Maximum: uncompressed AIFF-C
The folder selected in File > Save
Browser recorderWeb Borrowed computers or no app installation Service-dependent: MP3, WAV, WebM, or OGG Downloads, unless you choose another destination

Windows 11: record with Sound Recorder

Sound Recorder comes pre-installed on many Windows 11 computers. If it is missing, install Microsoft's Sound Recorder app from the Microsoft Store.

  1. Open Sound Recorder. Click Start, type Sound Recorder, and open the app.
  2. Select your microphone. Use the microphone dropdown near the bottom-left of the recording window. Choose your laptop microphone, USB headset, or external microphone by name. If your version has no input selector, choose the device under Settings > System > Sound > Input.
  3. Choose the recording format before starting. Open the app's Settings and find Recording format. Current versions offer options including AAC, MP3, and WAV. AAC recordings use the .m4a extension. Choose M4A/AAC for compact voice notes, MP3 for broad compatibility, or WAV for an uncompressed editing source. Format changes apply to new recordings, not existing files.
  4. Start recording. Click the red Record button or press Ctrl + R while the app is active. Speak normally. For your first attempt, record only 20 seconds.
  5. Mark important moments. If your version includes a flag or Add marker control, click it during a take to bookmark a topic change or a section that needs attention. These are app bookmarks, not spoken annotations. Do not rely on them appearing in an exported file or transcript.
  6. Stop and listen. Click Stop, then play the recording back. Check for silence, crackling, excessive room noise, and words cut off at the beginning.
  7. Rename and locate the file. Use the recording's menu to rename it and choose Show in folder or the equivalent file-location command.
Default Windows folder Documents\Sound Recordings

If OneDrive manages your Documents folder, the recordings folder sits inside that redirected location. Show in folder is the quickest way to find the actual file.

Windows 10: record with Voice Recorder

Windows 10's Voice Recorder follows the same basic process, but it has fewer recording controls.

  1. Open Settings > System > Sound. Under Input, select the microphone you want to use.
  2. Open Start, search for Voice Recorder, and launch it.
  3. Click the microphone-shaped Record button or press Ctrl + R.
  4. Click the flag button during recording to add a marker. Use Pause if you need a break without ending the session.
  5. Click Stop. The app saves the recording automatically.

Voice Recorder saves audio as M4A. Unlike newer Windows 11 Sound Recorder versions, it does not provide a recording-format menu for choosing MP3 or WAV.

Find your recordings in Documents > Sound Recordings, or right-click a recording and choose Open file location. If you need another format, convert the saved file with an audio editor. Changing the filename extension does not convert the audio.

Mac: use Voice Memos for quick notes and iCloud sync

Voice Memos is the simplest choice for reminders, study summaries, and ideas you want available on your other Apple devices.

  1. Open Voice Memos from Applications or Spotlight.
  2. Click the red Record button and begin speaking.
  3. Pause and resume as needed, then click Done.
  4. Rename the recording so you can recognize it later.

Voice Memos uses your selected system input. To change microphones, go to System Settings > Sound > Input before recording.

For iCloud sync, sign in with the same Apple Account on your devices and enable Voice Memos in iCloud settings. Syncing is optional; recording does not require it.

Need a regular audio file? Drag the completed recording from Voice Memos into a Finder folder or onto the desktop. This exports an M4A file you can attach, upload, or archive.

Mac: use QuickTime Player for a standalone audio file

QuickTime Player is a better fit when you want direct control over the input, recording quality, filename, and destination folder.

  1. Open QuickTime Player.
  2. Choose File > New Audio Recording.
  3. Click the arrow beside the Record button.
  4. Select your microphone and recording quality:
    • High: compressed AAC audio, saved as M4A. A practical choice for spoken recordings.
    • Maximum: uncompressed AIFF-C audio. This creates a much larger file and is unnecessary for most voice notes.
  5. Click Record, speak, then click Stop.
  6. Choose File > Save, enter a descriptive filename, and select a folder.

Watch the input meter before committing to a long take. Also, do not confuse QuickTime's monitoring-volume slider with microphone gain. Adjust the microphone's input level in System Settings > Sound > Input or with the microphone's hardware controls.

Example: recording a 15-minute project briefing

A remote worker selects a USB headset in QuickTime, chooses High quality, and saves the result as project-briefing.m4a. Saying "Progress," "Blockers," and "Next steps" aloud creates useful section cues for both listeners and a later transcript.

Web: record your voice in a browser

A browser recorder is useful on a borrowed computer or when you cannot install an app. Choose one with a clear Download or Export option and a privacy policy that explains whether recordings leave your device.

  1. Open the recorder in your browser.
  2. Click Allow when the browser requests microphone access.
  3. Select the microphone in the website's controls, if available. Otherwise, check the browser's microphone settings and your operating system's selected input.
  4. Record a short test and play it back.
  5. Record the full take, click Stop, and download the audio before closing or refreshing the tab.

Browser downloads go to your Downloads folder unless you choose another location. Available formats depend on the recorder; some export MP3 or WAV, while others produce WebM or OGG.

"In your browser" does not automatically mean "stored only on your computer." For confidential recordings, check the service's processing and retention practices before pressing Record.

Audio formats: compact files or an editing source?

Choose M4A/AAC for compact everyday recordings, MP3 for broad compatibility, or uncompressed WAV for editing. The difference in storage becomes noticeable on longer takes.

Example file sizes for a one-hour recording
Format and example settings Approximate size Practical use
AAC in M4A128 kbps 58 MB Voice notes, briefings, and uploads
MP3128 kbps 58 MB Sharing across many apps and devices
Uncompressed WAV44.1 kHz, 16-bit mono PCM 318 MB Editing and keeping an uncompressed source
Same recording length. Different file sizes.
M4A / AAC128 kbps 58 MB
MP3128 kbps 58 MB
WAV / PCM44.1 kHz · 16-bit mono 318 MB

Each bar represents one hour of audio on the same 0-318 MB scale. These are estimates in decimal MB; metadata and variable-bitrate encoding affect the result. Stereo PCM doubles the listed WAV size.

What M4A, AAC, and PCM mean
M4A and AAC
M4A is a file container; AAC is the compressed audio format commonly stored inside it. The container name and the audio encoding are not the same thing.
PCM WAV
PCM stores uncompressed digital audio samples. It is useful as an editing source, but takes more space than the compressed examples above.

Not every built-in recorder exposes bitrate, sample-rate, or channel settings. Renaming an extension does not convert audio, and converting a compressed recording to WAV cannot restore detail already lost.

2. Set up your microphone and fix silent recordings

A clear recording starts with the correct microphone, a healthy input level, and permission for the app to access it. Start with input volume around 70% to 80%, then adjust using the level meter and a short playback test.

INPUT-VOLUME STARTING POINT 70-80%

A starting control setting, not a required meter reading. Adjust for your microphone and voice.

SPEECH PEAKS, IF YOUR APP SHOWS dBFS -12 to -6dBFS

Leave room below 0 dBFS, where digital clipping begins.

Check the input meter before recording

A moving meter proves that the computer is receiving a signal. Playback tells you whether that signal sounds good. You need both.

Where to check your microphone input
Operating system Settings path What to do
Windows 11 Settings > System > Sound > Input Select the microphone, open its properties, and watch the input level or use the microphone test.
Windows 10 Settings > System > Sound > Input Select the device and speak while watching Test your microphone.
macOS System Settings > Sound > Input Select the microphone and watch the Input level meter.

A 70% to 80% input-volume setting is a starting point, not a required meter reading. Microphone sensitivity, distance, and speaking volume determine the right setting. Some USB microphones use a physical gain dial instead of an operating-system slider.

Speak at the loudest level you expect during the recording. Lower the gain if the meter repeatedly reaches its limit or the playback sounds crunchy. Raise it if speech is faint, but first move closer to the microphone.

In an app with a dBFS meter, aim for speech peaks around -12 to -6 dBFS. This leaves headroom below digital clipping at 0 dBFS. Clipped audio loses detail that turning the volume down afterward cannot restore.

What input gain, dBFS, and clipping mean
Input gain
The amount of amplification applied to the microphone signal before or during capture. Raising it also makes captured room noise more noticeable.
dBFS
Decibels relative to full scale, a digital level measurement. The top of the scale is 0 dBFS; ordinary recording levels sit below it.
Clipping
Distortion caused when the signal exceeds the available recording range. Lowering playback volume afterward does not recover the missing detail.

A percentage input slider and a dBFS meter measure different things. Do not treat a 75% slider position as a target signal level.

Silent recording? Check permissions first.

A connected microphone is not enough. The operating system must also allow the recording app to access it. Blocked access can prevent recording entirely or leave the app without a usable microphone signal.

Windows microphone permissions

  1. On Windows 11, open Settings > Privacy & security > Microphone.
  2. Turn on Microphone access.
  3. Turn on Let apps access your microphone.
  4. Enable Sound Recorder if it appears in the app list.
  5. For browsers and other desktop software, turn on Let desktop apps access your microphone.
  6. Reopen the recorder and make another short test.

On Windows 10: use Settings > Privacy > Microphone. Check device-level access, Allow apps to access your microphone, the Voice Recorder permission, and desktop-app access where relevant.

macOS microphone permissions

  1. Open System Settings > Privacy & Security > Microphone.
  2. Enable access for the affected recorder or browser if it appears.
  3. Quit and reopen the app if prompted, then test again.

Some built-in Apple apps are not listed there. A missing entry alone does not prove that access is blocked.

For a third-party app that has never requested access, attempt a recording inside that app to trigger its permission request.

Browser recordings have a second permission gate

Your operating system can permit Chrome, Edge, Firefox, or Safari to use the microphone while an individual website remains blocked.

Open the website's permissions through the browser's address-bar controls. Set Microphone to Allow, confirm the selected device, and reload the page. Both permission layers must permit access.

The microphone troubleshooting chain

Work from the physical connection to the final playback. Skip the website permission check for a desktop recording app.

Microphone connected

Check the cable and hardware mute switch.

Correct input selected

The input meter should move as you speak.

OS permission allowed

Permit access for the recorder or browser.

If blocked: allow access

Website permission allowed

Browser recording only.

If blocked: allow the site

Test recording plays clearly

Listen to the entire 20-second test.

The animated meter illustrates a working input; this diagram does not access your microphone.

Built-in microphone, USB headset, or condenser mic?

You can record your voice without buying anything. Still, microphone placement often matters more than the recording app.

Built-in laptop microphone

Good for: quick reminders and informal study notes.

AdvantageAlready available. No extra equipment to connect.

Trade-offCaptures room sound, fan noise, typing, and vibrations through the laptop.

USB headset microphone

Good for: briefings, calls, interviews, and spoken notes.

AdvantageStays close to your mouth as you move.

Trade-offBreath noise and rubbing cables can enter the recording.

USB condenser microphone

Good for: narration and longer recordings in a quiet room.

AdvantageOffers flexible placement and detailed capture.

Trade-offAlso captures keyboard noise and room reflections without careful setup.

Laptop microphones sit inside a machine that produces noise. Cooling fans create airborne sound; typing and desk movement send vibrations through the chassis. Raising the input gain amplifies those distractions along with your voice.

For a headset, position the boom a few centimeters from the corner of your mouth, out of the direct path of your breath. For a desktop condenser microphone, start about 15 to 20 centimeters away, use a pop filter, and avoid tapping the desk.

A condenser microphone is not automatically the better choice. In a noisy home office, a close headset microphone can produce clearer speech than an expensive microphone sitting across the desk.

Other problems a permission switch will not fix

The meter stays flat
Check the headset's physical mute switch, microphone cable, USB connection, and selected input. Reconnecting a device can change the active microphone.
Speech is faint
Move closer, confirm that you are speaking into the microphone's active side, and then adjust gain.
Audio crackles or distorts
Lower gain, increase distance slightly, and test another USB port or cable if the problem persists.
You hear echo or squealing
Use headphones or turn down live monitoring. Speaker output can feed back into the microphone.

For a long recording, connect power and check sleep settings before starting. A 20-second test does not reveal what happens when a laptop goes to sleep halfway through an hour-long session.

3. Turn your recording into searchable text

Upload your saved M4A, MP3, or WAV file to SpeechText.AI to turn spoken audio into searchable text. Speaker identification helps separate conversations, domain-specific models support specialist vocabulary, and Word or PDF exports make the finished transcript easier to edit or share.

Solve the post-recording dilemma

Recording is easy. Finding one useful sentence in a 60-minute audio file is not.

Without a transcript, you must scrub through the timeline, recognize voices, and replay passages to recover a deadline, definition, or decision. A written version lets you search for "budget," copy a study explanation, or review a briefing without listening from the beginning.

SpeechText.AI fits naturally after the recording step. Keep using Sound Recorder, Voice Memos, QuickTime, or your browser recorder, then upload the saved audio for transcription.

From a voice recording to a usable document

Keep the original audio. Review the text. Export the version you want to share.

Record

Use your preferred computer or browser app.

Save audio

Locate or export the actual file.

M4A MP3 WAV

Upload to SpeechText.AI

Choose the language and a suitable model.

Speaker identification Domain-specific models Multi-channel processing

Review transcript

Check names, numbers, and specialist terms against the audio.

Export Word or PDF

Choose an editable document or a file ready to distribute.

Transcribe your recording in five steps

  1. Locate the actual audio file

    On Windows, use Show in folder. From Voice Memos, export the recording. From QuickTime or a browser, open the folder where you saved it.

  2. Upload the M4A, MP3, or WAV file to SpeechText.AI

    There is no need to convert a supported M4A recording to WAV first. Conversion cannot restore detail lost during the original recording.

  3. Choose the language and closest available domain model

    Match the model to the content, not the speaker's job title. A doctor discussing holiday plans does not need a medical vocabulary model.

  4. Generate and review the transcript

    Check names, numbers, acronyms, and technical phrases against the audio. Background noise, accents, and overlapping voices affect recognition accuracy.

  5. Export the result

    Choose Word for an editable document or PDF for reading and distribution once processing and review are complete.

For interviews or meetings, obtain the required recording permission from participants. Before uploading confidential workplace, medical, or legal audio, check your organization's rules and the service's data-handling terms.

What speaker identification and domain models actually do

Speaker identification
(diarization)
Separates speaker turns instead of presenting an interview as one uninterrupted block. Labels such as "Speaker 1" and "Speaker 2" do not automatically reveal people's real names. Review them, especially where participants interrupt each other.
Domain-specific models
Account for specialist vocabulary. Medical terminology, legal phrases, and technical acronyms present different recognition challenges from everyday conversation. SpeechText.AI's domain-specific options help address those differences.
Multi-channel processing
Works with the structure of recordings that place participants on separate audio channels. Keep those channels intact rather than mixing them into mono before upload.

Critical details still need human checking. A wrong dosage or contract amount matters more than a misplaced comma.

Why a stereo recording does not always separate speakers

A stereo file containing the same mixed conversation on both channels does not provide the same speaker separation as a recording with different participants on separate channels. SpeechText.AI's multi-channel audio processing can work with useful channel separation, but the recording must contain it in the first place.

For professional handoffs, this combination is more useful than an unreviewed block of text: recognize the speech, distinguish speakers, check specialist terms, and export a document people can work with.