The space between two sounds is where stuttering happens — and where it's trained. Sound Bridge measures every transition you make, frame by frame.
This tour walks you through every part of the game. Three minutes. Eight steps.
Every transition resolves through three phases. You hold the FROM sound, slide into the TO sound without breaking your voice, then land on it. Press play to watch a clean S → T bridge.
The progress bar stretches to show which phase you're in. If your voice drops, the bridge breaks. Try the break button — the engine catches the gap and tells you exactly where it happened.
Behind the bar, three signals run continuously: the amplitude envelope of your voice, the pitch contour where voicing exists, and any voicing drops the engine catches mid-bridge.
This same signal layer powers every FluentPlay tool. Microphone-only, no wearables, no special hardware. Sound Bridge calibrates against your environment in the first 1.5 seconds of every session.
When you start the game, the first choice is how pairs are selected. There are three modes — pick whichever matches what you're working on.
In Custom mode, every English phoneme is on the grid. Tap one or more on each side. The engine pairs from your selections.
If you pick 3 FROM sounds × 4 TO sounds, the engine has 12 possible pairs to draw from each round. Pick a single pair to drill the same transition every round.
Or skip the picker entirely with the Drill Library below — curated themed sets cover the most common stutter targets:
Two controls scale how hard the same pair becomes. Start where you're successful and tighten as you improve.
Difficulty Tier (in Random mode) filters pairs by inherent difficulty — Easy through Extreme, plus Auto. Auto starts you on near-vowel transitions and ramps toward fricative-fricative pairs as you succeed.
After a round, the results screen lists every pair you attempted. Tap any row — the full engine analysis opens inline.
Engine signals, phoneme-level Azure pronunciation accuracy, plain-language diagnosis, and articulator views — for every single attempt. Two more buttons inside the panel: drill that pair again, or try its reverse.
Every session feeds your profile. Over time it surfaces the pairs you struggle with most — and turns them into a targeted drill.
Real attempts. Real signal.
The first round walks you through a quick mic calibration (1.5 seconds of silence). After that, every session is yours.
Glide between sounds without breaking your voice
This game gives you two sounds and asks you to connect them without stopping your voice. You'll hold the first sound, smoothly slide into the second, and hold that one too — building a bridge of continuous voice between them.
The progress bar stretches to show which phase you're in. If your voice drops, the bridge breaks and you try again. The game scores how smooth your transition was and how well you held each sound.
Start with easy pairs like AH→M or EE→N. These have big differences between the sounds, so your voice naturally carries across. Harder pairs like S→TH are similar-sounding and require more precise control.