Six Songs. One Identity
A Creator DNA proof showing how authorship persists across radically different musical forms.
Different genres. Different sound.
Same underlying pattern.
In a world of AI-generated music and voice, identity won’t be defined by similarity.
It will be defined by behavior.
Identity is not a snapshot.
It is a pattern under constraint.
As AI-generated music and voice become indistinguishable from human output, the problem is no longer creation. It's verification.
Scot Sier
Scot Sier has been making music for as long as he can remember.
A violin prodigy through elementary school, following in the footsteps of his Indiana state champion violinist father and grandfather, he competed and won — until the day he heard Jimi Hendrix play guitar. That moment didn’t end his musical career. It expanded the boundaries of creative exploration.
What followed was a lifetime of work spanning from power-pop to experimental noise collaborations with musicians across the globe — not as phases, but as genuine, fully-realized creative expressions, each rooted in a different world of sound. The classical foundation never left. It went underground, showing up in places most listeners wouldn’t expect to find it.
That’s the nature of a deep musical gift. It doesn’t stay in one genre. It travels.
His unique musical style and use of non-standard guitar tunings is what inspired Creator DNA. The loss of his mother when he was a teenager deeply affected his creative work, where he focused more on collaboration and human rights in his body of work. That is the core of his invention — it redefines and empowers the artist by creating a sonic footprint that, even if copied by AI, can stand up against deepfake replication by showing each person has a unique style over time that can’t be accurately reproduced by artificial means.
The six-song test across six radically different genres — each individually analyzed by Creator DNA biometrics — shows how Scot’s sonic biometric fingerprint is consistent in every song. The structural instincts, compositional behavior, and creative logic that define his work are an artist proof of identity that establishes a baseline of authentication against deepfakes, an infrastructure system for music distribution and ownership rights in the AI era.
Most systems ask: “Does this sound like X?”
Creator DNA asks: “Does this behave like X over time?”
Current AI classification is built on similarity — comparing surface features against known catalogs. Creator DNA operates at a different layer, measuring how a signal is organized across time, not how it appears at any single moment.
Six songs. Six genres. One creator.
A controlled dataset designed to stress-test identity. Six original compositions pushing into completely different stylistic spaces. If authorship identity were based on similarity, these tracks should appear unrelated. They are not.
Surface Difference, Structural Continuity
Listen to each track. The recordings are genuinely different in style, energy, and arrangement. Then look at the spectral data below each one.
Real divergence on the surface.
Tight convergence underneath.
Across all six tracks, one signal remained tightly consistent: chroma entropy — how harmonic energy is distributed across the chromatic scale.
The coexistence of high divergence and low entropy variance indicates a constraint-based generative system rather than genre-dependent composition. A similarity-based system cannot detect this constraint, because it evaluates surface features rather than the distribution rules governing those features.
The strongest measurable signal in this dataset is the constrained chroma entropy band. Style changes. Production changes. Genre changes. But the underlying behavioral system does not.
AI can copy what something sounds like.
Creator DNA measures what something is.
A voice can be cloned. A style can be imitated. But origin cannot be verified by similarity alone. Creator DNA introduces longitudinal identity verification — evaluating whether a signal stays within the same behavioral constraints across time.
- Tone and timbre
- Phrasing and rhythm
- Genre conventions
- Stylistic surface features
- Constraint-based behavioral identity
- Longitudinal harmonic distribution
- Authorship signal across genres
- Origin — not just appearance
- Music authorship verification
- Voice authentication
- AI-generated content detection
- Content provenance
- Larger datasets
- Cross-creator comparison baseline
- Blind testing protocols
- Independent replication
A Second Independent Signal.
Compositional Identity.
Chroma entropy measures compositional identity. This layer measures something different — the harmonic behavioral signature extracted from each recording: how melodic energy is voiced, how timbral character distributes across frequency, how pitch behavior organizes over time.
Two independent analytical layers. Six different genres. One consistent origin signal.
Across six radically different genres, two independent signal layers show the same result: one behavioral origin.
The harmonic identity analysis — extracted independently from the compositional chroma analysis — shows consistent timbral character (14.5% CV), consistent fundamental pitch behavior (17.3% CV), and consistent melodic presence across five of six recordings. These are not the same measurement as chroma entropy. They are measuring different properties of the signal. They are converging on the same conclusion.
When two independent analytical layers both show tight consistency across six different genres, the probability of that being coincidental collapses. This is what a behavioral identity signal looks like.
This layer uses harmonic-percussive source separation (HPSS) to extract the melodic/harmonic component of each recording, then applies MFCC timbral analysis, pYIN fundamental frequency estimation, and voiced ratio detection. This measures the combined harmonic behavioral signature of voice and melodic instrumentation. True isolated vocal biometrics (formant analysis F1/F2/F3) would constitute an additional third identity layer. Analysis performed using librosa 0.11.0.
A Third Independent Signal.
Your Vocal Tract. Your Identity.
Formant frequencies are determined by the physical shape and length of your vocal tract — your throat, mouth, and nasal cavity. They are anatomical constants. They do not change with genre, style, emotion, or intent. They cannot be trained away or imitated with precision.
Four isolated vocal tracks across four different genres were analyzed using Praat acoustic phonetics software — the same tool used in forensic voice analysis and clinical speech research. The results speak for themselves.
| TRACK | F0 | F1 | F2 | F3 | HNR |
| The Clown | 192.7 | 743 | 1830 | 2835 | 16.8dB |
| Time Out | 218.9 | 702 | 1697 | 2788 | 11.3dB |
| Rusty Cadillac | 243.4 | 884 | 1893 | 2953 | 17.8dB |
| Colonizer | 197.9 | 781 | 1943 | 3087 | 15.1dB |
F3 — the third formant — is primarily determined by the length of your vocal tract from glottis to lips. It is the most anatomically fixed of all formants. It does not change based on what you are singing, how you are singing it, or what genre surrounds it.
Across Glam Rock, Acid-Jazz-Punk, Roots-Pop, and Space Rock, F3 ranges from 2788 to 3087 Hz — a 4.0% window. This is a vocal tract signature. It belongs to one person.
Three separate analyses. Three different measurement systems. One origin.
Chroma entropy measures compositional behavior. Harmonic MFCC analysis measures timbral identity. Formant analysis measures vocal tract anatomy. These are not variations of the same test. They are independent signals drawn from fundamentally different properties of the recordings. All three converge on the same conclusion.
This is what Creator DNA is built on — not one signal, but the convergence of multiple independent behavioral and anatomical layers into a single verifiable origin point. That convergence is what AI cannot replicate. That convergence is the future of authentication.
Vocal analysis performed on four isolated vocal tracks using Praat acoustic phonetics software via parselmouth 0.4.x (Python). Formant extraction used Burg method, 5 formants, 5500Hz ceiling, 25ms window. Pitch analysis: 75–600Hz range. HNR: harmonic-to-noise ratio (cc method). Analysis covers The Clown (Glam Rock), Time Out (Acid-Jazz-Punk), Rusty Cadillac (Roots-Pop), Colonizer (Space Rock). Isolated vocal tracks provided by artist.
These six tracks are different. But they are not from different creators.
Current AI systems would get that wrong. They listen for similarity. They compare surfaces. In a world where any voice can be cloned and any style can be imitated, that approach has already failed.
Creator DNA Biometric behavioral analysis asks how a signal behaves over time, and the answer to that question is, origin.
This is the future of authentication. Not just in music. In every creative field where human identity must be distinguished from artificial reproduction.
© Scot Sier. All compositions, performances, and recordings by Scot Sier.