Solo mode
Listen, imitate, get your score. Five sounds per run.
Sound pack
How the score is calculated
Your recording never leaves your device in solo mode. The game looks at the shape of the sound, not at your voice. It first measures the shape of your take: the rhythm (the volume curve and the attacks of the syllables) counts for 60 % of it and the length for 40 %. That shape is then multiplied by the content: the melody, the pitch contour compared relatively, so you can sing higher or lower than the original. For sounds without a clear pitch, such as noises and effects, the spectral content replaces the melody, and that part still depends on the voice.
So the right rhythm with the wrong words caps around 35, and the right melody unlocks the rest. Starting more than 250 ms before or after the expected moment costs points, down to a quarter of the score when you are 1.2 seconds off. Identical audio scores 100, a good impression 60 to 80, the right rhythm with the wrong words about 30, an unrelated sound 20 to 30, silence 0.
FAQ
Are my recordings uploaded?
No. In solo mode everything is analyzed locally in your browser and nothing is sent to a server.
Why is my score low?
The right rhythm with the wrong words caps around 35: the melody is what unlocks the rest, so reproduce the syllables, their attacks and the ups and downs of the pitch. Start exactly when the sound plays in the middle of the window (follow the grey waveform), match its length and avoid background noise. Headphones help a lot.