Session: 00:00
PAUSED
--
softness (H1–H2): --
sing a steady note to read it
Target (click to edit)
F#3–A#3 (185–233 Hz)
Spectrogram settings
Practice session complete
Nice work. Here's what you did this session.
Duration
00:00
Stages completed
0 / 0
Progress
Everything stays on this device. Trend lines are relative to your own history — never a score.
Day streak
0
Sessions recorded
0
How practice has felt (last sessions)
| Date | Comfort | Fatigue | Target | Effort | Pain |
|---|
Minutes per day (last 3 weeks)
This session by stage
| Stage | Time | Target |
|---|
Recording archive (on this device)
Reference voices
Collect voices you like (links or notes) to widen your sense of what a feminine voice can be. Observe them objectively — many valid fem voices are lower in pitch or larger in resonance than the stereotype suggests. Avoid judging your own voice as good/bad; compare qualities, not worth.
✓ saved
Recent sessions
| Date | Time | Stages | Brightness |
|---|
Save this session to your progress?
Your voice profile
Shapes how the routine talks to you and how targets are framed. Stays on this device; edit anytime.
How this works — read me once (~2 min)
Session flow, how often to train, which graph to use, what the app refuses to score, and where your data lives.
Skip to the questions →
What actually happens in a session
About 15–20 minutes, in fixed stages: Check-in (30 s — how your voice feels today, and a hard stop if you have pain) → Warm-up (free, easy sounds) → Resonance (tongue-root & cavity size — the piece that changes how your voice reads) → Register Blend (smooth connection between the two textures) → Volume (loud without getting heavier) → Sentences (real speech, read aloud) → Prosody & Emotion (melody, stress) → Fluidity (moving between your everyday and target voice on purpose). A Benchmark stage appears once a week: the same locked List 1 every time, so week-to-week comparison is honest.
Why the order is what it is
Every later stage is built on the setup from the earlier ones. If resonance collapses, sentences collapse with it — so the app makes you establish the sound before you use it in speech.
How often you should train
15–20 focused minutes a day, 5–6 days a week is the sweet spot. Short and consistent beats long and rare — your voice adapts through repeated gentle exposure, not through marathon sessions. Take a rest day when your throat feels tired; skip days are fine, this app deliberately has no streak counter to punish you. Do the benchmark once a week (same day each week if you can). Every few weeks, re-listen to your old recordings — that's where progress becomes audible, because day-to-day change is too small to hear.
Pitch view vs Spectrogram — which one do I use?
Pitch view (default): one line showing how high or low you are over the last 10 seconds. Use it for anything about notes, glides, and the blend — and to catch yourself drifting upward when you get excited. Spectrogram: a heat map of energy across all frequencies. You can't read a note off it, but you can see resonance move: brighter/forward speech lights up energy on the right of the picture. Use it when you're working on resonance or brightness and want to see the change.
Rule of thumb: start on Pitch and stay there for stage work; switch to Spectrogram when you're specifically hunting resonance. You can change the default in the splash screen, and toggle any time from the header.
Neither view gives you a score or a "sounds female" verdict.
What the app deliberately does not do
No gender score, no "male/female %" reading, no leaderboard, no streak pressure — those either lie or push you into strain. Targets are bands, not single notes, and brightness is shown as a trend against your own history, never as an absolute number.
Where your data lives
Entirely in this browser on this device: recordings, progress, profile. Nothing is uploaded and there is no account. Moving to another device means using Export backup in the Progress panel and importing it on the other side.
Safety, plain and simple
Burning, pain, or a strained feeling = stop for today. Hoarseness past a day or two = rest; past two weeks = see an ENT. This is a self-practice tool, not therapy or medical care.
My goal
What you want out of this. It does not lock anything — it picks which routine you get and how the app frames targets.
How do you want to work?
Two honest routes to the same place. Practice-first starts from real speech on day one (faster to sounding natural, less theory). Foundation-first trains each piece of the instrument alone first — play, pitch control, tongue root, softness, brightness — then combines them (the order VAMethod teaches in; slower start, fewer entrenched habits). You can switch later in this panel.
Which language(s) do you want to practice in? (multi-select)
The first selected language is the one sessions start in. Sentences, prosody and emotion prompts follow it; technique stages (tongue root, resonance, blend) are the same in every language — that is physics, not language. Add more later; each language has its own sentence lists and prompts.
Experience
So the app knows how much explaining to do. Nothing is assumed — "completely new" is a perfectly good answer.
Timbre I'm aiming for
Timbre (say TAM-ber) is the color of your voice — what makes two people on the same note sound different. It is not pitch (high/low) and not loudness. It comes mainly from resonance: how big the space in your throat and mouth is, and where the sound is focused. "Brighter" = smaller, more forward space; "softer" = less pressed, more relaxed texture. Pick whatever attracts you — nothing here is a commitment.
Things I'd rather avoid
Common traps. Tick any that worry you and the routine will warn when an exercise tends to cause that — e.g. "do not chase high pitch" or "keep it non-nasal".
Where I'll use it
Where your voice has to work in real life. This decides which contexts the practice prompts imitate (phone calls, home, friends, work).
Backup & restore
Your progress, profile, and recordings never leave this device automatically. Export everything into a single file when switching devices; importing config/json keeps this app device-local.