SpeechDuck

Automatic sidechain ducking for Stream Deck. SpeechDuck lowers your music, game, or desktop audio the instant you speak — in your mic or in Discord — and brings it back the moment you stop. One button to arm it, live level meters right on the key.

Your mic always wins.

Talk over anything. Automatically.

The mute button you never have to press.

Full description

Every streamer needs this, whether they know it yet or not. You're mid-song, someone drops into your Discord call, and now you're scrambling for a mixer, an OBS audio filter, or a second monitor just to turn the music down before you can be heard — and turn it back up two minutes later, if you remember. Some people fight through this with OBS filters. You don't have to. You already own a Stream Deck. Let it do the job.

SpeechDuck is real-time sidechain ducking, built to live entirely on your Stream Deck. The moment your mic picks up your voice — or a conferencing app starts talking back — SpeechDuck automatically pulls your music or game audio down to whatever level you set, and brings it back up the instant you go quiet. It's the same trick broadcast radio and podcast studios have used for decades, just one key press away.

This isn't a nice-to-have for the occasional stream. If you listen to music while you play, or you keep Discord open in the background, autoducking is something you use every single session, not just when things go wrong. Get it right once and your viewers stop noticing your audio entirely — which is exactly the point. No more chat messages telling you the game is drowning you out. No more overcompressed voice because you rode the game volume down manually and forgot to bring it back up.

Everything lives on one key. Tap it to arm or disarm the whole effect — no menus, no alt-tabbing out of your game. While it's running, the same key doubles as a live equalizer, showing you the signal from your triggers in real time, so you always know it's actually listening.

What you control

Triggers: pick any microphone, any output device, or a specific running application as what causes the ducking — mix and match as many as you like. An app you add before it's even open will be picked up automatically the moment it launches.

Targets: choose exactly what gets turned down — music apps, games, or an entire output device — each with its own "volume while ducked" level.

Attack / Sustain / Release: shape exactly how fast it ducks, how long it stays down through pauses in speech, and how gently it comes back.

Per-trigger threshold and multiplier: drag directly on the live level meter to set the exact point where ducking kicks in — no guessing at numbers.

Live meters everywhere: every trigger and target shows its real signal level, right in the settings and on the key itself.

Your own language: a language switcher built right into the settings panel — English, German, French, Spanish, Japanese, Korean, Chinese (Simplified and Traditional), and Russian. Russian in particular isn't an option Stream Deck itself offers at all; SpeechDuck adds it anyway.

Why it's fast

SpeechDuck polls your audio hardware roughly 60 times a second, so the reaction to your voice is effectively instant — no audible lag between the first syllable and the music dropping out of the way.

€9.99. One-time purchase. Set it up once, forget it's there, and let your viewers enjoy audio that just works.

Fixed: volume left at the ducked level after a shutdown or crash. If the PC was shut down, or the plugin was closed abruptly while it was ducking, your devices could stay quiet after the next start, and you had to fix them by hand. That can no longer happen.

Safe shutdown. When Windows begins a shutdown or sign-out, SpeechDuck immediately puts every ducked device and app back to its own level. If the shutdown is cancelled, ducking carries on as usual.

Reliable crash recovery. If the plugin is closed abruptly, the next start restores everything it had lowered. This now works with any number of targets; before, it failed whenever more than one was ducked. A device that isn't connected yet is restored as soon as you plug it in. An app is restored the next time it starts.

Never turns anything down. Recovery only raises volume back to where you left it. If you've already set a level by hand, SpeechDuck leaves it alone.

Correct starting level. Starting the plugin while someone is talking no longer assumes 100% volume, which used to push targets louder than you had them.

Self-healing. If the plugin stops unexpectedly, it restarts in the background and restores your volume straight away.