Free Online Voice Recorder
Record voice from any microphone with live levels, pause and resume, then trim and download WebM, Ogg, M4A or WAV. Audio never leaves your device.
Loading tool…
How to record your voice online and save it as WAV
This recorder is for capturing audio you want to keep, such as a voice memo, a voice-over draft or a spoken message. If you only want to check that a microphone works, the Microphone Test is quicker; to capture your screen with narration, use the Screen Recorder.
- Choose a microphone, or leave System default selected. Device names appear after you allow access once.
- Decide on browser processing. Leave echo cancellation, noise suppression and auto gain off for natural voice-overs, or turn them on for calls in a noisy room.
- Press Start recording and allow microphone access. Speak and keep the level meter mostly in the green; red means the signal is clipping.
- Use Pause and Resume to skip breaks. The timer only counts recorded time. Press Stop when you finish.
- Play the take back. Download the browser’s original compressed file, or set start and end times and choose Download WAV for an uncompressed copy.
When an online voice recorder is the right tool
A browser recorder is handy when you need a clean audio file quickly and do not want to install software or sign in to anything. Typical uses include a voice memo to yourself, a spoken reply to send to a colleague, a draft voice-over for a slideshow, a language-learning exercise where you compare your pronunciation with a native speaker, or a short audition clip.
It is also useful on a borrowed or shared computer. Because the audio stays in the tab and nothing is uploaded, closing the page leaves nothing behind except the files you chose to download. On a phone, the same page works with the built-in microphone or a headset.
This tool is different from the site’s Microphone Test, which shows a live waveform and spectrum without keeping any audio, and from the Screen Recorder, which captures video of your screen. Use the recorder when the end goal is an audio file.
Worked example: a two-and-a-half-minute voice-over
Suppose you record a 2 minute 30 second (150 second) voice-over in Chrome. The browser records Opus audio in a WebM container, and the recorder asks for about 128 kbps, so the original download is roughly 150 × 128,000 ÷ 8 = 2,400,000 bytes, about 2.4 MB. The browser may choose a slightly different bitrate, so the size shown after recording is the true one.
Now you trim 1.2 seconds of silence from the start and stop at 148.5 seconds, leaving 147.3 seconds. Your browser decodes audio at its own rate, typically 48,000 samples per second. A mono 16-bit WAV of that selection is 44 header bytes plus 147.3 × 48,000 × 2 bytes of samples, which is 14,140,844 bytes, about 14.1 MB. Choosing stereo doubles the sample data to 28,281,644 bytes, and choosing 24-bit mono gives 21,211,244 bytes. The size estimate under the trim fields shows the same calculation before you download.
How the recorder works
When you press Start, the page asks the browser for a microphone stream with the device and processing options you selected. A Web Audio analyser reads the signal about 60 times per second to draw the waveform and the level meter. The meter shows the RMS (average) level in dBFS, where 0 dBFS is the loudest value digital audio can hold, plus a peak marker that briefly holds the highest recent value. A peak at or near 0 dBFS turns the meter red because the sound is clipping.
The browser’s MediaRecorder encodes the audio while you speak. Pausing stops both the encoder and the timer, so the timer equals the length of the finished file. When you stop, the recorder decodes the file back to raw samples. Trimming cuts those samples at the start and end times you set, mono export averages the channels, and the WAV encoder writes a standard 44-byte RIFF header followed by little-endian PCM samples. All of this is ordinary JavaScript running in the tab.
Common mistakes to avoid
Recording with the wrong input is the most frequent problem, especially with a laptop that has both a built-in microphone and a headset. Check the device list and watch the meter move before a long take. If the meter barely moves, raise the input level in your operating system sound settings rather than speaking louder.
Leaving noise suppression on for music or singing can make notes sound watery, because it is tuned for speech. Turn it off for anything other than calls. Finally, remember that starting a new take replaces the current one. Download the file you want to keep first, and keep the original download as a backup before trimming.
Useful tips and limitations
- Speak about a hand’s width from the microphone and slightly off to the side to reduce popping sounds on p and b.
- Choose mono WAV for spoken word. It is half the size of stereo and most microphones capture a single channel anyway.
- Record ten seconds of room silence at the start. It gives you a clean point to trim from and lets you hear background hum before the real take.
- If an editing app will not open the WebM file, download the WAV version instead; virtually every audio editor reads 16-bit PCM WAV.
Common questions
Which file format does the recorder save?
The original download uses whatever your browser records natively: usually WebM or Ogg with the Opus codec in Chrome, Edge and Firefox, and MP4 with AAC in Safari. The WAV option converts the recording to uncompressed 16-bit or 24-bit PCM in your browser.
Is there a recording time limit?
Each take can run for up to one hour. WAV exports are limited to about 400 MB so the browser tab does not run out of memory; trim the recording or pick mono if you reach that size.
Why does the trim only apply to the WAV download?
The original file is kept exactly as your browser encoded it, because cutting a compressed WebM or MP4 file would mean re-encoding it. Trimming happens on the decoded audio, which is then saved as WAV.
Can I record on an iPhone or Android phone?
Yes, in a current mobile browser. Safari on iPhone and iPad records MP4/AAC, and Chrome on Android records WebM/Opus. Keep the screen on during long recordings, because some phones suspend background tabs.
Why does the player show an unknown length for my WebM recording?
Some browsers write WebM files without a duration in the header. The recording is complete; the recorder shows the real length from its own timer and from the decoded audio.
Does the recorder reduce background noise?
Only if you turn on the browser’s noise suppression before recording. It cannot be applied afterwards in this tool.
Recording, playback and encoding run in this browser tab. Audio is not uploaded to ToolWorld Online and disappears when you close the page unless you download it.