The SonaLab Manifesto

In 2025, SonaLab was finished. But I sat on it for six months before letting it say hello to the world.

I could have made a flashy promo video and thrown money at ads. Instead, I decided to write this first. Because the world doesn't need more marketing — it needs more honesty. Honesty in the product, honesty in learning, honesty in how you treat your users.

I'm MiaoGeGe — a full-stack architect who's spent twenty years crossing boundaries: from Adobe Certified Instructor to FileMaker specialist, from brand management to indie software development, from polymer science to I Ching algorithms.

But SonaLab is the most challenging project I've ever taken on, because it's one of those rare things that demands you understand vocal pedagogy, signal processing, and software engineering all at once. It's not just a tool — it's a translator, turning the art of singing into data and charts that anyone can understand.

Let me start from where it all began — my relationship with music.

I've loved singing for as long as I can remember. I never had the chance to study music formally, but singing has always been one of the most important parts of my life.

In 2021, scrolling through Douyin's algorithm-driven feed, I stumbled upon Justin. Back then, my understanding of singing was stuck in the "just go by feel" era — can't hit the high note? Push harder. Throat tired? Take a break. It was all mysticism.

But when I got on a live call with Justin, everything was different. He didn't talk about "vocal cord closure," "breath support," or "resonance placement" — concepts I'd heard a thousand times but never truly understood. Justin simply had me follow his instructions and sing. I had no idea what was happening, but suddenly I sailed right past my passaggio at B4 with ease. Something told me he was onto something real. For the first time, I felt like singing could actually be understood.

That winter, I traveled to Changsha for Justin's master class for singers. I still remember the name: "Enlightment."

After the master class, I didn't just rediscover my passion for music — more importantly, I began to truly understand the voice. I started studying music theory, mixing, and arranging. I realized music isn't about feeling — it's an entire system that can be deconstructed, quantified, and mastered.

SonaLab traces back to that moment — Enlightment 2021, that winter, that classroom.

At the gym you have mirrors, body-fat scales, and a coach correcting you in real time. You can clearly feel every muscle doing its job.

Isn't singing just another kind of workout? Except you're training your vocal cords, laryngeal muscles, facial muscles, and breathing muscles — developing awareness and control of the space inside your body.

But singing has no mirror. Students can only rely on their ears to "feel" it, and on their teacher to "hear" it. Sound is fleeting — your teacher can't be at your ear every second. And trying to describe something that flows through time with words alone is inherently clumsy and imprecise.

What if there were a "mirror for your voice" — one that lets everyone see how their vocal cords are working in real time?

With that question in mind, I started digging.

I discovered that sound can actually be "seen":

• FFT analysis reveals the frequency makeup of a sound — fundamental, harmonics, formants
• Spectrogram shows how sound changes across time and frequency
• Normalized Amplitude Quotient (NAQ) quantifies how well the vocal cords close
• Singing Power Ratio (SPR) tells us whether a voice has "carrying power"

The first time I opened Logic and saw the spectrum of my own singing, I was stunned.

What I thought was "natural" was actually full of instability, breathiness, and throat tension…

The voice that felt so "powerful" when I was singing turned out to be painfully weak on the recording…

Voice science let me truly hear my own voice for the first time.

I started doubting whether I could really sing at all. I'm not conservatory-trained, but I'd always gotten compliments from people around me — and now I was drowning in self-doubt.

I started booking lessons constantly. Jasper, Billy, and Justin kept encouraging me. Listening back to those recordings now, I'm impressed the teachers managed to keep a straight face.

But what truly pulled me out of that spiral was a counterintuitive discovery — the auditory illusion.

When we sing, two versions of our voice reach our eardrums simultaneously: one conducted through the skull, the other reflected through the air. Our brain "mixes" these two into what we think is "our voice."

In other words, what you think you hear and what everyone else hears are not the same sound at all.

Only a recording reveals the truth. That's why so many people find their own recorded voice unfamiliar — that's what the audience actually hears.

From that day on, I hit record every time I practiced. Because I stopped trusting my own ears.

Studying mixing alongside vocal training taught me which problems can be fixed in post-production — and which can't.

The point is: singing is source control. If the source isn't right, every downstream step becomes exponentially harder, and the outcome becomes unpredictable.

Mixing can patch things up, but it's tedious and limited. To truly solve vocal problems, you have to get it right at the moment of phonation.

That got me thinking: what if there were a tool that could give you feedback the instant you produce sound — so you don't have to wait until you're in the studio to discover the issues?

So I decided to build one myself.

I started trying every vocal analysis app on the market.

Voice Vista is very professional — its FFT analysis and formant tracking are solid. But for a vocal beginner, those charts might as well be encrypted messages. Jargon everywhere, yet nobody telling you "here's what you should actually do."

What I wanted was a vocal teaching tool that serves everyday enthusiasts, not just a set of technical metrics for professionals.

An EQ lets us see sound from a frequency-vs-amplitude perspective. But singing needs far more than an EQ.

I needed a tool that could take all those "feelings" from the teacher's classroom and turn them into data anyone can read.

The hardest part wasn't the technology — it was figuring out how to translate "mysticism" into "science."

When Justin talked about "vocal balance," I went and studied what depth and compression look like on a spectrum. When he said "dynamic balance," I looked for the relationship between loudness and register. When he said "airflow balance," I researched how NAQ distinguishes pressed, flowing, and breathy phonation.

All of this was to make SonaLab the first mirror for vocal learners — real-time, quantifiable, and directional.

Over the two years of building SonaLab, I thought about giving up more times than I can count.

C++ memory management, real-time threading, SwiftUI performance tuning, account security, AI integration… every pitfall made me question my life choices.

But every time I opened SonaLab, heard my own voice, and watched those numbers dance, I kept going.

Because it makes vocal learning visible, quantifiable, and trackable.

When a complete beginner sings a stable resonance for the first time, what they see in SonaLab is a beautiful flash of light — a "fingerprint" of sound, the art of vocal cord vibration.

When a professional singer wants to level up, SonaLab can tell them: your passaggio was silky smooth today, your compression score improved by 0.3 since last week — that's the measure of progress, the evidence of deliberate practice.

When friends got their hands on a SonaLab beta, the first question was always: "Is this thing going to replace voice teachers?"

My answer: SonaLab won't replace voice teachers — just like a calculator doesn't replace math teachers. But a good tool makes a good teacher even better.

Think of it this way: when we weigh ourselves, different teachers might use different brands of scales, but they all use "kilograms" as the unit. With that standard, you can tell whether you're making progress.

SonaLab is that "kilogram" — an objective standard of measurement. It tells you what's happening with your voice, but it doesn't choose your training method for you.

If a teacher's theory doesn't hold up — today they say you sound great, and SonaLab confirms it; tomorrow you do the same thing, produce the same sound, but they say you've gotten worse — then you have every reason to ask: did the "kilogram" change? No, the kilogram is right there. It's the "scale" that changed.

The ultimate purpose of singing is freedom.

The freedom to express whatever emotion you want, unconstrained by your vocal cords, your throat, or your body — that's the state every person who loves to sing is chasing.

And SonaLab is a light on that road to freedom. It won't walk the path for you, but it lets you see every step beneath your feet.