An AI loudness readout in Bangkok, Thailand: Bow checks whether soft and loud phrases really differ and plans the first silence
Bow is thirteen and had a problem that is common among young pianists. The score asked for a soft opening, a louder middle and a gentle ending, and Bow was sure all three were there. Yet the teacher kept saying that everything sounded the same, as if every bar had been played at one comfortable level.
Bow’s home is a high-rise flat in Bangkok, Thailand, and practice happens at a digital keyboard near a balcony door. Traffic and air-conditioning hum in the background, so Bow had to think carefully about how to test the teacher’s comment without fooling the result.
First, Bow touches the opening rest on the page with a fingertip and counts its length under a breath, so the recording will begin with a planned silence instead of a nervous rush. A clean silence at the start also gives the recording a quiet reference point, which turned out to be useful later.
Bow tried a practice tool that uses AI to analyse recordings. After a short take, it drew a line across the screen showing how loud each moment was and tried to mark where each note began. The idea was simple: if the soft opening and the loud middle were truly different, the line should rise and fall.
What the readout can and cannot tell you
The first readout surprised everyone. The line barely moved. The soft opening and the louder middle sat almost at the same height, and only the final chord dipped a little. Bow had felt the difference in the fingers, but the recording showed that the sound had stayed nearly flat.
At home Bow works with TIGA ALPHA, an AI piano-teaching app from a Thai team. Its CEO, Teacher Aum, has 20 years of teaching experience, and the app is built to sit alongside traditional music schools, so a readout like this becomes one clue among many rather than a final judgement.
Tools like this sit inside wider music teaching. Bow still takes piano lessons and music lessons, follows a piano course and a piano learning course, and likes to play piano and play music with friends at school. Parents who want their children to learn piano online, or who are learning piano online themselves, compare an online piano teaching application, a piano teaching app, an online piano lessons app and an AI piano teaching app. Piano teaching with AI, an AI online piano course, online piano learning, an online piano learning app, AI piano lessons and an AI piano teacher can help people learn music, but careful piano teaching still decides how to learn piano with musical sense.
With the evidence on the screen, Bow changed the practice plan. The soft opening was played with the fingers closer to the keys and less arm weight. The louder middle used more weight from the shoulder down, not faster fingers. After each change, Bow recorded again in the same spot with the device in the same place.
The second readout showed a clear valley at the start and a hill in the middle. Listening back, Bow could finally hear what the teacher had meant. The difference had been in the plan all along, but the hands had not been delivering it.
Frequently asked questions
Q: What does an AI loudness readout show? A: It analyses a recording and draws how loud each moment was, often as a line or a set of bars across the piece. Some tools also try to mark where each note starts, so a learner can compare one phrase with another.
Q: Are the numbers exact? A: No. The microphone, its distance from the instrument, the room and background noise all change the reading. The readout is useful for comparing phrases inside one recording made in the same place, not for comparing different rooms or devices.
Q: Should learners trust the readout over their ears? A: The ears come first. A readout can point to a spot worth listening to again, but the final choice about how soft or how loud a phrase should be belongs to the player and the teacher.
In the next lesson with a live teacher, Bow played the piece and then showed both readouts on a tablet. The teacher pointed out that the hill in the middle was now too sudden and asked for a gradual climb across four bars. The screen had helped, but the musical shape still needed a human ear.
Bow also learned the limits of the tool. When the balcony door was open, traffic noise lifted the quiet sections, and the readout looked flatter than the playing really was. Moving the device or the room changed the line too, so Bow kept every comparison inside a single recording session.
A parent can help by asking simple questions after each take: which part was meant to be soft, and does the line agree? If the answer is no, the next step is to listen again, not to chase a number.
There was a bonus lesson in the numbers, too. Bow realised that a crescendo marked in the score is a promise to the listener, and a promise needs room to grow. Starting the loud middle a little softer than planned left space for a real climb instead of a jump.
With the balcony door shut and the city humming behind the glass, Bow records one more take, starts with a planned silence, and watches a gentle valley open before the middle climbs.
Want to hear whether your soft and loud passages really sound different? Start with TIGA ALPHA at https://tigaalpha.github.io/landing-en/ — Everyone gets the Premium package free for 7 days, then 799 THB/month. Prefer live lessons with Teacher Aum? See https://www.tigastudio.net/, where live teaching runs up to 30 hours per week. Read about TIGA ALPHA, the company and its technology, at https://www.tigaalpha.com.



Comments