Setup guide
Use any app as your microphone
A virtual microphone is a mic input that carries whatever audio you decide to put on it: a music app, your real mic, or both blended together. Here is how to build one on Windows, step by step.
The goal: a microphone you assemble
Every calling app, recorder, and streaming tool has a microphone dropdown, and normally that dropdown only lists physical devices. A virtual microphone is a software device that shows up in that same list, except you decide what audio it carries. Route a music app into it and the call hears your music. Route your real mic into it and the call hears you. Route both and the call hears the blend: your voice on top of the music, live, as one input.
That one mechanic covers a surprising number of jobs: playing a track to friends on a call, firing a soundboard-style clip mid-conversation, or feeding a recorder a finished voice-plus-audio mix instead of a bare mic signal. The rest of this guide builds it with patchd, a Windows virtual audio mixer where every bus presents itself as a microphone, but it starts with what Windows alone can and cannot do.
The generic Windows route, and where it stops
Windows has no built-in switch for "play this app into my mic." The traditional workaround is a bare virtual audio cable: you set the app's output device to the cable's input, and the cable's output appears as a microphone you can select. That genuinely works for a single source, and it is worth knowing.
It stops working the moment you want a blend. A bare cable carries exactly one feed, so mixing your real mic on top of the app means chaining Windows "listen to this device" tricks that add delay and give you no meters and no per-source control. You end up guessing at levels inside a call. A mixer solves this because blending sources onto one output is the entire job of a mixer.
Step 1: pull the app onto its own strip
In patchd, every source is a strip with its own fader, live meter, and mute. To bring one specific app in, pull it onto a cable strip: set that app's output device to the virtual cable, and the strip now carries just that app and nothing else. Your music player, a browser tab, a sampler, anything that plays audio on Windows can sit here. If pulling a single app onto its own channel is new to you, how to route audio per app on Windows walks through it in detail.
Use a dedicated cable strip rather than the always-present Desktop strip for this. Desktop carries all system audio, including sounds you never meant to broadcast, and, as you will see below, it is the fast lane to a feedback loop.
Step 2: route the strip to Bus 1
A bus is a virtual output you assemble a feed for, and in patchd every bus doubles as a microphone other apps can select. Under the strip's BUS header is a row of routing pills. Click the Bus 1 pill and it lights cyan: whatever plays on that strip now flows onto Bus 1. That is the entire routing model, a source reaches a destination when its pill is lit, and only then.
Step 3: add your real mic to the same bus
This step is optional, and it is the one that makes the trick genuinely useful. Light the Bus 1 pill on your mic strip too. Bus 1 now carries the blend: your voice and the app's audio, mixed live. A bus's contents are simply whichever strips have its pill lit, so the three useful states are one click apart: app only (music to the call, you silent), mic only (back to a normal microphone), or both (you talking over the track).
Step 4: select patchd Bus 1 as the microphone
Now open the target app, the call, the recorder, the broadcast tool, and open its microphone dropdown. Alongside your physical devices you will see patchd Bus 1. Select it. From that app's point of view it is just a microphone; it has no idea the signal is an assembled mix. What the other side hears is exactly what you routed onto the bus, nothing more and nothing less. This works the same in Zoom, Teams, Meet, Discord, OBS, or a recorder: if the app has a mic picker, the bus can be its mic. If you are feeding a recording app, the posture is the same, patchd does the routing and your recording app captures the bus.
Step 5: set levels so your voice wins
Each strip has one fader, and that fader is the strip's level everywhere it goes. Talk while the track plays and watch the two meters: if your voice is fighting the music, pull the music strip's fader down until speech sits clearly on top. Resist fixing it the other way, by cranking a quiet mic into distortion; if the whole blend arrives too quiet on the far side, how to fix a mic that is too quiet covers the clean ways to add gain before you touch anything else.
The classic mistake: the feedback loop
Here is the failure everyone hits once. You are on a call, and you route the Desktop strip onto Bus 1 so the call can hear your audio. But the call's own output, the other people's voices, is also playing through Desktop. So their voices flow onto the bus, back into the call, and out their speakers again: everyone hears themselves echoed, and if it keeps circulating it climbs into a shriek.
The fix is structural, not a volume tweak: put the app you want to share on its own cable strip (Step 1) and leave Desktop's Bus 1 pill unlit. The call's output then never touches the bus, and the loop cannot form. If you are hearing echo and are not sure this is the cause, how to fix mic echo walks through diagnosing it.
Where patchd is today
patchd is built so this whole page is a two-minute job: pull the app onto a strip, light two cyan pills, and pick the bus as your mic. It is in development now. Join the waitlist and we will tell you the moment it is ready to install.