Run your microphone through a chain of VST3 plugins and use the result as a virtual microphone in any app.
MicVST is a lightweight Windows tray app. It takes one microphone, runs it through the VST3 effects you choose (EQ, de-noise, compression, …) and sends the processed signal to a virtual audio cable - which any application (Discord, OBS, Zoom, Voicemod, games, …) can then use as a microphone input.
Think of it as a minimal alternative to VoiceMeeter / Wave Link when all you want is: one input, plugins on it, out as a virtual mic.
Mic → MicVST (your VST3 chain) → VB-Cable (CABLE Input → CABLE Output) → Discord...
Grab the latest portable build - no installer, no dependencies, just run the .exe:
(Statically linked, x64. No Visual C++ Redistributable required.)
The .exe is not code-signed, so Windows SmartScreen may warn “Windows protected your
PC” - click More info → Run anyway. Each release lists the file’s SHA-256 so you can
verify your download (Get-FileHash MicVST.exe -Algorithm SHA256 in PowerShell).
You also need a virtual audio cable. The free VB-Cable is recommended - MicVST detects it automatically. VoiceMeeter and Virtual Audio Cable work too.
- Install VB-Cable and reboot. (If no cable is installed, MicVST shows an orange hint with the download link.)
- Run
MicVST.exe. On first start it picks your default microphone as input and the detected virtual cable (e.g. CABLE Input) as output. Both are changeable in the audio setup and are remembered across restarts. - Build your plugin chain (bottom list, “+ Plugin”):
- add any VST3 effects you like (e.g. a noise-suppressor, EQ, compressor),
- optionally insert the built-in Mono → Stereo node at the point where you want stereo (before it the chain runs mono = less CPU),
- drag rows by the handle on the left to reorder; the trash icon on the right removes a plugin (with confirmation). Double-click a row to open the plugin’s editor.
- In your target app (Discord/OBS/Zoom/…), select CABLE Output as the microphone.
- Done. Device and plugin settings (including each plugin’s state) are saved automatically.
Right-click the tray icon → “Run at Windows startup” to launch MicVST silently into the tray on
boot (no window, engine running). For a stable setup, copy the .exe to a fixed location first
(e.g. C:\Tools\MicVST\) and enable autostart from there.
- One mic → ordered VST3 plugin chain → virtual cable output
- Automatic cable detection (VB-Cable / VoiceMeeter / Virtual Audio Cable) with a download hint when none is found
- Channel-aware routing - the chain can start mono and switch to stereo at a placeable Mono → Stereo node (so a mono noise-suppressor doesn’t cost double)
- Horizontal in/out level meters with a dB scale
- Drag-to-reorder plugin list, per-row bypass, remove with confirmation, per-plugin editor windows
- Scans the standard VST3 folder; add custom VST3 folders via “+ Plugin → Add VST3 folder…” (a plugin that crashes the scan is remembered and skipped next time)
- Persistent settings (
%APPDATA%\MicVST\config.xml), low latency, silent tray autostart - Optional, opt-in update check (off by default): one request to the GitHub releases API on startup; no telemetry, no auto-installer
- Single portable
.exe, no install, no runtime dependencies
Requires Windows 11 x64 and Visual Studio 2022 with the Desktop development with C++
workload (MSVC + Windows SDK + CMake). JUCE is fetched automatically via CMake FetchContent
(JUCE 8.0.13) - nothing else to install.
$cmake = "C:\Program Files\Microsoft Visual Studio\2022\Community\Common7\IDE\CommonExtensions\Microsoft\CMake\CMake\bin\cmake.exe"
# Configure (fetches JUCE)
& $cmake -S . -B build -G "Visual Studio 17 2022" -A x64
# Release build -> build\MicVST_artefacts\Release\MicVST.exe (self-contained, ~8 MB)
& $cmake --build build --config Release --target MicVST
# Optional: unit tests (exit code 0 = ok)
& $cmake --build build --config Release --target MicVSTTests
.\build\MicVSTTests_artefacts\Release\MicVSTTests.exeThe build uses the static MSVC runtime (/MT), so the resulting .exe runs on any Windows 11 x64
machine without the Visual C++ Redistributable.
- An AudioProcessorGraph wires the input device → your plugin chain → the output device,
driven by JUCE’s
AudioDeviceManager(WASAPI shared mode). - Per hop,
min(source.outs, target.ins)channels are connected, so variable channel width (the mono→stereo transition) is routed correctly and automatically. - The output device is just a normal render endpoint - pointing it at a virtual cable’s input endpoint lets the cable’s driver loop the audio to its output (capture) endpoint, which other apps read as a microphone. No special “output plugin” needed.
MicVST adds only the smallest buffer necessary to run your VST3 chain - typically one 10 ms block (480 samples at 48 kHz), the standard Windows shared-audio period. There's no extra buffering on top: audio comes in, goes through your plugins, and goes straight out.
A few things worth knowing:
- The buffer size is a single fixed value (e.g. 480) because MicVST uses WASAPI shared mode - the same low-overhead path Windows uses for everything else. Windows sets one engine period for shared audio, so there's nothing to tune there, and nothing is being wasted. MicVST simply shows it (with the live end-to-end latency) under the device list.
- Total latency you hear is mostly downstream of MicVST. The signal path is
Mic → MicVST → virtual cable → your app (Discord/OBS/…). The virtual cable and the receiving app each read through Windows shared audio too, and the cable has its own buffering. If you want to trim the cable's part, VB-Cable exposes a latency / internal-sample-rate setting in its own control panel (VBCABLE_ControlPanel.exe). - For talking into Discord/OBS/Zoom, latency is inaudible - your microphone signal travels one way to your listeners, so a few milliseconds never matter. (Low buffer sizes only matter when you monitor yourself live or play an instrument in real time, which isn't what MicVST is for.)
- ASIO / ASIO4All won't help here. ASIO is built for a single exclusive in-and-out device and doesn't fit a mic-in / virtual-cable-out setup - and the receiving app would still read through shared audio anyway. MicVST is already on the most direct path Windows offers for this job.
- rnnoise (noise suppression) is available as a VST3 here:
https://github.com/werman/noise-suppression-for-voice/releases - drop the
.vst3intoC:\Program Files\Common Files\VST3\and restart. It runs at 48 kHz; for low CPU use the mono variant and place it before the built-in Mono → Stereo node. - Third-party software (JUCE, your VST3 plugins, VB-Cable, …) is subject to its own licenses. This repository contains only the MicVST source.
MicVST is free software licensed under the GNU General Public License v3.0 - see LICENSE. It uses JUCE, whose free open-source tier is GPL, so MicVST is GPL too. Third-party software (JUCE, your VST3 plugins, VB-Cable, …) is subject to its own licenses; this repository contains only the MicVST source.
Copyright (C) 2026 Philip Zimmermann
This program is free software: you can redistribute it and/or modify it under the terms of the
GNU General Public License as published by the Free Software Foundation, either version 3 of the
License, or (at your option) any later version. It is distributed WITHOUT ANY WARRANTY; without
even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.