How does sayit's shadowing trainer work?

Shadowing in sayit works as a listen-then-repeat loop: it plays a native-quality reading of a line, you record yourself saying that same line back, and it's scored the same way any other take is — per word, with your version compared against the model's. Pure listening exercises train your ear; this turns the same material into an actual speaking rep, one sentence at a time.

30-second version: Pick a difficulty level, and sayit generates a leveled passage, splits it into sentences, and works through them one at a time: play the model line, you record your version, get an immediate score with per-word tinting, then move to the next line. It's the same recorder and scoring engine as regular practice — shadowing is a structured way of feeding it material, not a separate scoring system.

What actually happens on each line?

Three steps, repeated: sayit plays the reference line first, at native quality, so you hear the target before you attempt it. Then you record your own version of that same line. Then it's scored — a per-word verdict and score, tinted the same way flagged words are tinted everywhere else in sayit — before the session moves on to the next sentence. The loop is designed to keep momentum: hear it, say it, see it, move on, rather than stopping to review every detail before continuing.

How is this different from just listening and repeating on my own?

You get an actual measurement of the gap, rather than your own impression of how close you got. It's easy to feel like you nailed a sentence in the moment and be wrong about it — shadowing without scoring relies entirely on your own ear, which is exactly the instrument that's usually least reliable at catching your own first-language habits. sayit's version replaces that guess with a real per-word comparison against the model line you just heard, using the same phoneme-level scoring as any other take.

Can I control how hard the material is?

Yes — there are three difficulty levels (roughly easy, medium, and hard), which control the complexity of the generated passage the shadowing session works through. Starting easier gives you room to focus purely on matching the model's pronunciation and pacing before the vocabulary and sentence structure add their own difficulty on top.

Does shadowing check my timing, or just my pronunciation?

The scoring itself is the standard per-word pronunciation comparison — the same accuracy pipeline used everywhere else. What shadowing adds is the discipline of the exercise: hearing the model immediately before attempting it trains you to match not just individual sounds but the rhythm and flow of a native reading, since you're producing your version right after hearing the real thing rather than reading cold from text.

Shadowing versus reading a passage cold

Where do I find it?

Shadowing sits inside the Listen section of the app, alongside dictation and the reading library — all three work with spoken material, and shadowing is specifically the one built around speaking the line back rather than only listening or typing it.

Try it

Open Listen in sayit, pick a level, and work through a few lines — hear each one, say it back, and watch the per-word score land before the next line loads.