Try it — no account needed
Record a standard read-aloud passage and see acoustic feedback on pitch and resonance. This is a limited preview (3 analyses per day from your network): audio is deleted on our servers right after analysis, scores use population benchmarks only and stay in this browser session for about 30 minutes, then expire. Local replay can repeat in the same tab until you refresh or record again.
How the demo works
- Server discards audio — your clip is analyzed in memory and deleted on our side. Nothing is stored on our servers.
- Local replay (optional) — right after analyzing, this tab can replay your clip as many times as you like from browser memory. Refreshing, closing the tab, or recording again removes it.
- Scores are temporary — kept in this browser session for about 30 minutes (refreshing the page is fine), then gone. Nothing is stored on our servers.
- Population benchmarks only — with an account you can save your own starting voice and training target for a personal practice index.
- 3 of 3 tries left today from this network.
Why sign up? Accounts keep recordings private on your account, let you compare takes over time, set personal baselines, and revisit past clips—not just a one-off snapshot.
What these scores mean Acoustic cues for practice — use with your ears, not instead of them
VoxAccalia is not a passability detector. It does not judge how others will perceive you, whether you “pass,” or whether a voice sounds convincing. No percentage here can answer that—and we will not pretend it can.
This does not replace listening. Scores are no substitute for hearing your own recordings, getting feedback from a trusted listener, or working with a speech clinician or voice coach when you have one. Use the numbers to track practice—not instead of real listening.
This app is an acoustic measurement tool. It extracts pitch (how high/low), formants F1–F3 (vowel resonance / mouth shape), and intonation (pitch movement) from a clip, then compares those numbers to reference ranges or your own baselines. Practice index is a weighted blend of those measurements—useful for tracking your practice over time, not for labeling you.
Pitch and formants move independently. A very high, cartoony falsetto can raise pitch scores while formants stay unchanged—so the overall % may rise even when the voice does not sound natural or aligned with your goal. That is expected math, not a bug. Sustainable training usually shifts pitch and resonance together; watch the breakdown lines, not just practice index.
Garbage in, garbage out. Whispering, mumbling, background noise, or clips so quiet the analyzer barely finds speech will produce unreliable numbers. Record a few seconds of clear, natural speech at a normal volume—or upload a clean file from the Upload menu (same as Record in the navigation bar).
How to use VoxAccalia → — recording, uploading, reading the breakdown, and getting trustworthy numbers.
Short, simple sentences—easy to read clearly.
Read aloud
Speak clearly at a natural pace for about 30–60 seconds.
Do you like honey? Well, honey is sweet, and so are you. You can have a little or you can have a lot. You can have some now and the rest later. Or you can have it all now. Whatever you want, whenever you want. But don't eat too much or you might get sick. And remember, honey is made by bees. Bees buzz around flowers and collect nectar. Then they turn the nectar into honey in their hives.
Microphone
Clips around 30–120 seconds work best. Recording auto-stops at 120 seconds.
Live level and pitch are rough estimates—not your analyzed scores.
Preview (not saved)
Check your clip before analyzing