About

Built by Ranjan in Canberra for my own learning, which began in September 2026.

It began with listening, and with tones. Those are the two things a self-taught learner gets the least honest feedback on: nobody is there to tell you that you missed a word, or that your second tone didn't rise. So the lab writes short clips from the words I'm actually learning and reads them in natural voices, drills the four tones by ear, and, when I say a syllable back, draws my pitch over the voice's and judges it.

Dictation, handwriting, a conversation partner and my own texts followed, each because I wanted it for my own practice. A few native-speaker friends tested it, recording on their own phones, and their ears and feedback shaped the tone judge and much else. It is one person's project, and it says so wherever it matters: the voices are synthetic, the partner is an AI, and the judge is a guide, not a ruling.

Free accounts with daily limits are open now, and a paid version with a larger allowance may follow if there is demand; the Accounts page says what that would mean, and that nothing available here requires payment at this time. The free part of the lab, tone practice with the judge, dictation and handwriting, is at tinglilab.org for anyone, with no account.

Thanks

To the founding testers: Bangle Wu (Bernice, 吴帮乐), Kanghong Shi (史康宏) and Yongfu (Deven) Huang.

Contact

Write to hello@tinglilab.com. I read every email, and feedback from native speakers and teachers is especially welcome.