Why can't Windows Sound Recorder record system audio?

Sound Recorder is hard-coded to use the default microphone input only. There is no UI to switch it to a loopback device, and the underlying UWP API the app uses does not expose system audio capture. This is partly an Apple-style privacy choice and partly that Microsoft expects users with system-audio needs to use OBS, Audacity, or one of the bundled OEM tools (Dell SmartByte, Lenovo Vantage). For a quick mic capture Sound Recorder is fine; for system audio you install Audacity and use WASAPI loopback as described in the guide. The output also caps at 128 kbps M4A AAC, which is a hard ceiling regardless of input device.

Is Audacity safe to install?

Yes if you download from audacityteam.org. The 2021 acquisition by Muse Group introduced concerns about telemetry and account linking; the actual implementation is opt-in optional telemetry that you can decline at first launch and disable later under Edit, Preferences, Application. The audio engine and editing code are unchanged from the pre-acquisition codebase and remain open source under GPLv2. Avoid third-party download portals (softonic, cnet, similar) — they sometimes wrap the installer with adware. The Microsoft Store version is also legitimate and skips the optional telemetry prompt entirely.

What's the difference between WASAPI loopback and Stereo Mix?

Stereo Mix is a hardware/driver-level virtual device exposed by some Realtek, Conexant, and Creative audio chipsets. It captures the post-mix output stream as a regular input, and any recording app sees it without special configuration. WASAPI loopback is a Windows API mode (Windows 7+) where any app can request the output stream of any endpoint. WASAPI works on every modern Windows machine because it does not depend on driver support; Stereo Mix only works where the hardware vendor enabled it. WASAPI is also per-endpoint (you choose Speakers loopback specifically) and per-session, which makes it more reliable but slightly more configuration-heavy.

How do I record a Discord call on Windows?

Open Audacity, set Audio Host to Windows WASAPI, Recording Device to your speakers loopback, Recording Channels to 2. Add a second mono track (Tracks, Add New, Mono Track) and set its input to your microphone. Press Record. Audacity captures the Discord call audio (other speakers) on track 1 and your voice on track 2 in a single session. After the call, normalize the levels per track and export as MP3 or WAV. Note: recording calls is regulated by wiretapping and consent laws in some jurisdictions (two-party consent states in the US, GDPR in the EU). Get explicit consent from participants before recording.

Why is my recording quiet on Windows?

Settings, System, Sound, choose your input device, click Properties. Look at the Input volume slider — anything below 50 typically produces a quiet recording. Push it to 75-100 and check the input level meter while you talk. Some USB mics also expose a hardware gain knob that overrides the OS slider. For headsets, check the headset's mic mute and gain toggles on the inline controller. If post-recording the file is still quiet, normalize it in Audacity (Effect, Volume and Compression, Normalize) or push it through [audio-compressor](/audio-compressor) which raises perceived loudness as part of the re-encode.

Can I record audio from a specific app only?

Standard WASAPI loopback captures the entire output endpoint, mixing every app together. To isolate a single app you need software-level routing. Voicemeeter Banana (free, vb-audio.com) and Audio Router (open-source on GitHub) both let you route specific apps to virtual audio cables and record only those cables. The setup is more involved than WASAPI loopback (you create a virtual cable, set the target app's output to it, point Audacity at the cable as input) but it is the only way to capture, say, only Discord audio while ignoring Spotify in the same session. For one-off recordings, just mute the apps you do not want.

How do I convert M4A from Sound Recorder to MP3?

Use [m4a-to-mp3](/m4a-to-mp3) — drop the Sound Recorder M4A file into the page, the conversion runs in your browser using WebAssembly ffmpeg, and the resulting MP3 downloads back to your machine. The file never uploads anywhere. Most M4A-to-MP3 conversions complete in a few seconds for a typical 5-minute recording. For background on the format itself see [what is M4A](/blog/what-is-m4a). If you have many files to convert, the page handles batch input — drop them all and download a zip of the converted MP3s. For larger archives, ffmpeg on the command line ('ffmpeg -i input.m4a -b:a 192k output.mp3') runs faster than any browser tool.

How to Record Audio on Windows: 2026 Guide

Windows has more recording options than macOS but worse default ergonomics. The pre-installed Sound Recorder app handles single-microphone captures and nothing else. System audio recording is officially possible through Stereo Mix on hardware that supports it, but Stereo Mix is missing on most modern laptops because vendors stopped enabling it in their audio drivers around 2018. The reliable path in 2026 is Audacity's WASAPI loopback mode for system audio and OBS Studio for screen-plus-audio captures.

This guide covers what each tool actually does, the Stereo Mix versus WASAPI trade-off, the OBS pipeline for streamers and tutorial creators, and the cleanup chain that turns the raw recording into a shareable file.

Sound Recorder: The Pre-Installed Microphone Capture

Windows 10 and 11 both ship with Sound Recorder (rebranded from Voice Recorder in the Windows 11 update). Open the Start menu, type Sound Recorder, hit Enter. Click the red record button, click stop. Recordings save to your Documents, Sound recordings folder by default, named "Recording.m4a" with a timestamp suffix.

The output format is M4A AAC at roughly 128 kbps mono — Microsoft fixed this at the codec level and there is no settings panel to change it. For most voice-note use cases the bitrate is fine. For music or anything where you care about the high frequencies, Sound Recorder is the wrong tool — its 16 kHz internal pre-emphasis is tuned for spoken word.

Sound Recorder records the default microphone only. It cannot select a specific input device from inside the app; if you have multiple mics you change the default under Settings, System, Sound, Input. It cannot record system audio. There is no built-in trim or edit feature beyond start-stop — for trimming you push the M4A to audio-trimmer or open it in Audacity.

Stereo Mix: Sometimes There, Often Not

Stereo Mix is a virtual recording device that captures whatever is mixed to your speakers. It dates back to Vista-era audio drivers and was the default way to capture system audio on Windows for a decade. It still works on hardware that exposes it, with one tap of the configuration UI:

1. Right-click the speaker icon in the system tray. 2. Sound settings, More sound settings (this opens the legacy Sound Control Panel). 3. Recording tab. 4. Right-click in the empty area, check Show Disabled Devices and Show Disconnected Devices. 5. If Stereo Mix appears, right-click it, Enable.

Where it does not appear: most modern laptops with Realtek, Conexant, or Intel SST audio drivers. Manufacturers strip Stereo Mix out of the driver INF to comply with various OEM-licensing terms or to push users toward bundled recording apps. There is no software fix — if your driver does not expose Stereo Mix, you cannot enable it.

When Stereo Mix works, it is the lowest-friction option: any recording app sees it as a regular input. When it does not, skip to the WASAPI loopback approach below.

Audacity WASAPI Loopback: The Reliable System-Audio Path

WASAPI (Windows Audio Session API) is the modern audio API that replaced the old MME and DirectSound paths. WASAPI's loopback mode lets a recording app capture the output stream of any audio endpoint (a speaker, an HDMI output, etc.) without going through Stereo Mix. Audacity supports it natively.

Install Audacity from audacityteam.org (the only official source — third-party download portals sometimes bundle adware). Audacity is free and cross-platform; the post-3.0 versions ask about optional telemetry on first launch. Decline if you want a clean install.

Configure WASAPI loopback:

1. Audio Host: Windows WASAPI. 2. Recording Device: choose the entry that ends in (loopback) — typically Speakers (Realtek Audio) (loopback) or Headphones (loopback) depending on your default output. 3. Recording Channels: 2 (Stereo). 4. Project Sample Rate: 44100 Hz for music, 48000 Hz if you are recording for video.

Press the red record button, play whatever you want to capture, press stop. File, Export Audio, choose MP3 or WAV. The file contains exactly what the Speakers endpoint was playing, sample-accurate, no Stereo Mix driver involved.

The one limitation: the source app must keep playing while you record. Pausing Spotify or closing the browser tab also pauses the loopback — there is nothing to capture. WASAPI loopback also captures system sound effects (notifications, the Windows startup sound), so mute notifications before a recording session.

For mic-plus-system-audio in one session, set Audacity's Recording Device to the loopback entry on track 1, then add a second mono track and arm it with the microphone as input — Audacity supports per-track input selection in Tracks, Add New, Mono Track, then click the track's input dropdown.

OBS Studio: Screen Plus Audio for Tutorials and Streams

OBS Studio (free, obsproject.com) is the standard screen-recording tool when you need both video and audio. The audio configuration matters because OBS exposes desktop audio and microphone as two separate tracks by default.

Settings, Audio:

Desktop Audio: choose your default speaker output. OBS uses WASAPI loopback under the hood, so this works on systems where Stereo Mix does not.
Mic/Auxiliary Audio: choose the microphone.

Settings, Output:

Recording Format: MKV is OBS's default and is more crash-safe than MP4 (a crashed OBS produces an unfinished MKV that VLC can still play; an unfinished MP4 is often unrecoverable). Remux to MP4 after recording via OBS, File, Remux Recordings, or convert to audio-only with mp4-to-mp3.
Audio Track: pick Track 1 for a single mixed audio output, or enable multiple tracks if you want desktop audio and microphone in separate channels for post-production.

For audio-only recording with OBS, set the video bitrate low (1000 kbps is fine — it gets stripped out anyway), record, then run the result through mp4-to-mp3 or your preferred extractor to get a clean MP3. The extract audio from video guide covers the same conversion in detail.

Quality Settings: What to Pick

For voice (podcast, interview, voice memo): 44.1 kHz, 16-bit, mono, MP3 at 96-128 kbps or M4A AAC at the same. Anything higher is wasted bits.

For music (band rehearsal, live performance, mixing source): 44.1 kHz, 16-bit minimum, 24-bit if your interface supports it. Export to FLAC or WAV for the master, MP3 320 kbps for sharing. Background on bitrate trade-offs lives in audio bitrate explained.

For video sync (recording audio for a YouTube video, anything that pairs with footage): 48 kHz, 16-bit. Cameras and editing software default to 48 kHz; recording at 44.1 forces a sample-rate conversion that can drift over long takes.

Cleanup Chain After Recording

Same shape as the Mac chain: trim, compress, convert.

audio-trimmer crops dead air at both ends. audio-cutter removes mid-file sections — coughs, mistakes, irrelevant talk-over. Both run client-side in your browser; the file never leaves your machine. For a desktop alternative see how to cut audio in Audacity.

audio-compressor re-encodes to a sensible bitrate for sharing. A 30-minute WASAPI recording at 320 kbps stereo is about 70 MB; at 128 kbps mono it is roughly 28 MB.

For format conversion, Sound Recorder produces M4A which most podcast hosts and chat apps still prefer as MP3 — run it through m4a-to-mp3. Audacity exports of WAV go through wav-to-mp3. Both keep the conversion local in your browser.

Summary

Sound Recorder for one-off voice notes. Stereo Mix if your hardware happens to expose it. Audacity WASAPI loopback for reliable system audio capture. OBS for screen plus audio. After recording, the trim-compress-convert pipeline finishes the job. None of these tools require an account or upload your audio to anyone.