Why turn the browser's audio processing off?

Because echo cancellation, noise suppression and automatic gain are all designed to make a call intelligible, not to preserve a recording for editing. Echo cancellation in particular misjudges reverb in a room and damages the voice. With them off you keep the raw room, which is exactly what a cleanup pass wants as its starting point.

Will the recording be stereo and 48 kHz?

The tool requests both, but a request is not a guarantee. If the input device cannot deliver stereo or that sample rate, the capture uses what is available, and the summary after the take states the channel count and rate that actually arrived, so there is no ambiguity about what you have.

Why is there an automatic stop at twenty four minutes?

It is a safety net for a recording left running rather than a target to work towards. A long single take in a browser tab is more fragile, so ending the file at a natural point is safer than relying on the tab staying awake indefinitely.

My take sounds echoey. Is something wrong?

No, that is the intended behaviour. Turning off echo cancellation and noise suppression means the room is captured as it sounds. The trade is deliberate: this is the state a cleanup pass wants, and tools that reduce steady noise and tame reverb can work on it afterwards, which they could not if processing had already altered the signal.

Can it record a phone call or a remote guest?

Only what reaches the microphone. A remote participant on a call is captured if the speaker output is loud enough for the microphone to pick up, which is usually a poor result. A proper remote interview needs the far side recorded on their own machine, or the system audio captured separately and the two takes aligned later.

Interview Recorder

You Might Also Need

Multi-Guest Leveler

Balance levels for multiple guests

Voice Recorder

Record your voice in browser

Podcast Recorder

Record podcast episodes with chapters

Noise Reducer

Reduce background noise in audio

Interview Recorder captures a conversation at the microphone's best quality, asking for 48 kHz stereo and turning the browser's own processing off. It is the take you intend to edit rather than to use as it comes.

What is Interview Recorder?

Press record and the capture requests the highest quality the input can offer: stereo where available, 48 kHz preferred, and with echo cancellation, noise suppression and automatic gain all disabled. A safety stop ends the take automatically at twenty four minutes.

What the capture asks for and why:

A channel count of two where the device supports it, so the recorded file can be manipulated as a stereo pair.
A preferred sample rate of 48 kHz, the standard for spoken production and the rate most interfaces run at.
Echo cancellation turned off, because it damages a voice when it mistakes reverb for an echo to remove.
Noise suppression turned off, because it is tuned for calls rather than for programme material and leaves artefacts a cleanup pass would rather start from a clean slate.
Auto gain turned off, so the level does not ride up and down with the conversation.
An automatic stop at twenty four minutes as a safety net rather than a target, together with the standard preview and download panel.

A word on the constraints: these are requests, not commands. A device that cannot deliver stereo or 48 kHz will deliver what it can, and the label after the take reports the channel count and sample rate that actually arrived. Disabling the browser's processing is the substantive choice, and it is the opposite of what a conferencing tool does. The point is to keep the raw room: a little reverb and a little background noise can be reduced later with proper tools, whereas processing applied during capture cannot be undone. That trade is worth stating plainly because a take made this way will sound less polished than a call recording when you first play it back.

What it cannot promise:

The requested stereo and sample rate are best effort, so a mono input produces a mono take.
With processing off, room echo and background noise are preserved, which means a poor room stays audible.
Twenty four minutes is a hard ceiling per take, so a longer session needs two recordings.
It records the microphone only, so a remote participant on a call is not captured unless the speaker output is picked up by the microphone.
There is no level metering, so an input set too high will clip without warning.
Nothing here separates the speakers afterwards; that is processing applied to the finished file.

Use it for a conversation you will edit: an interview, a two hander segment, or any spoken take where the room will be dealt with afterwards. Record with the microphone close enough to keep the direct sound ahead of the reflections, and plan for the twenty four minute ceiling by taking a fresh file at a natural break rather than mid sentence. Because the browser's processing is off, the room is part of the recording, so the most valuable preparation is to move the microphone closer to the speaker and to soften the hardest surface in the room; both are free, and both reduce how much repair the take needs later.

How to use Interview Recorder

Press record and allow microphone access; the capture requests stereo at 48 kHz with browser processing off.
Keep an eye on the elapsed timer, since the take stops automatically at twenty four minutes.
Press stop at a natural break if the conversation runs longer and start a fresh file.
Check the reported channels and sample rate in the label, then download the WAV for editing.

When to use Interview Recorder vs related tools

Interview Recorder is for conversations destined for post production, which is why it disables the processing that a call app would apply. Microphone Recorder lets you pin a specific input when the interview is recorded through an interface, and Scheduled Recorder handles the case where the session starts without you at the keyboard. Afterwards, Noise Reducer deals with the room and Multi-Guest Leveler evens out the speakers.

Privacy & Security

This tool runs entirely in your browser — no data ever leaves your device. There is no server round-trip, no upload, no logging, and no account required. Your input is processed locally using client-side JavaScript and is never stored, transmitted, or accessible to anyone else. When you close the tab, everything disappears.