How Đúng Chưa? practices honesty.

Recording check before tone feedback

Before any tone judgment, Đúng Chưa? runs a recording check — whether the audio is clear enough to evaluate. Noisy, clipped, or unstable audio gets a retry prompt — no fake feedback that would train the wrong habit.

What is Measured

Tone calibration relies on **vocal pitch contour tracing** rather than simple spelling check calculations. When you record a phrase, our processing core extracts the fundamental frequency (F0) trace of your voice over the duration of the syllable.

Operational Parameters

• Contour Shape (F0 Trace): Evaluates the literal direction of the sound. Does your pitch stay flat (ngang), dip and rise (hỏi), fall steadily (huyền), or break suddenly and climb (ngã)?
• Slope & Velocity: Measures how fast your pitch transitions. A sluggish rise on the rising *sắc* tone will fail calibration because it can sound flat to Vietnamese listeners.
• Duration & Cutoff: Evaluates timing. Tonal vowels that end with final consonant cutoffs (e.g. *t, c, p*) must be short and crisp.

What is NOT Measured

We purposefully exclude **volume/amplitude** and **grammatical composition** from the core evaluation algorithm. Speaking louder will not fix a flat pitch shape, and vocabulary checks are handled by your written coursework, not our sound processor.

Signal vs. AI Role

A key technical and ethical boundary is established between mathematical signal matching and language intelligence helpers. **AI is never the judge of your speech attempt.**

This prevents chatbot-style hallucinated judgments. When audio is scoreable, one correction is surfaced. When it is not, the app abstains — no fake certainty.

The "One Correction" Rule

Adult language acquisition research demonstrates that learners suffer from **cognitive overload** when presented with a wall of detailed pronunciation corrections simultaneously. If an app tells you that your vowels, timing, volume, and tones are all incorrect, your speaking trace will lock up.

Đúng Chưa? enforces a strict single-focus practice structure. If your recording contains three distinct pitch contour errors, the calibration core will select **the single most critical correction cue** for your next attempt, ignoring the others until the core shape is calibrated.

This allows you to relax, focus on one correction cue (e.g. "dip lower before the break"), and retry the phrase with clear, singular intent.

Dialect & Accent Boundaries

Vietnamese pronunciation features significant regional variations between Northern (Hanoi), Central (Danang), and Southern (Ho Chi Minh City) dialects. For example, the *hỏi* (dipping-rising) and *ngã* (broken-rising) tones are pronounced distinctly in the North, but merge into a single dipping shape in the South.

[!IMPORTANT] Our early pre-launch training decks utilize strict **Northern dialect target baselines** to ensure baseline calibration consistency. Learners practicing for Southern speech must be aware that the app will coach them toward standard Northern contour shapes. Southern target baselines are in active validation and will be rolled out systematically once finalized.

Known Technical Limitations

Calibrating vocal contours requires capturing high-quality voice signals. We communicate our pre-launch boundaries with absolute transparency:

• Ambient Background Noise: High levels of environmental noise (e.g. coffee shop crowd static, wind, or street traffic) introduce frequency anomalies that interfere with pitch contour tracing. Calibration works best in reasonably quiet environments.
• Microphone Quality: Standard smartphone microphones are optimized for voice calls, but cheap headset mics can sometimes clip frequencies, affecting trace accuracy.
• Controlled Practice Deck: Our waitlist version utilizes a curated baseline practice deck of high-frequency phrases. It does not support arbitrary free-form speech input.