Decentralized Sound Ownership & Monetization Protocol

Abstract

Since the relatively recent inception of recording technology and the distribution of recording art following shortly after, there are fundamental flaws in the way sound creators around the world have been able to distribute, monetize, and protect their creations. In a 2018 report from CitiGroup, it was stated that a significant majority of revenue from sound art—as high as 88%—never reaches the original creator.

SoundMint solves this through a comprehensive platform for sound creators to:

• Collaborate while receiving verifiable credit for contributions
• Publish work directly to fans through blockchain-based ownership
• Monetize through a novel three-token system with transparent revenue distribution
• Protect intellectual property with cryptographic provenance seals and audio watermarking
• Resist censorship through decentralized storage and on-chain provenance

This is powered by SoundMint's Audio Provenance system (cryptographic Provenance Seals today, with Meta AudioSeal watermarking on the delivery path), a Three-Token Architecture (SongIPAsset, OwnershipToken, CollectorToken), Quadratic Bonding Curves for fair price discovery, and decentralized storage of release metadata and encrypted masters on IPFS.

1. The Problem

The music industry's current infrastructure fails sound creators in four critical ways:

1.1 Revenue Inequality

• Artists receive only 12-18% of revenue from their work
• Complex intermediary chains (labels, distributors, streaming platforms) extract value
• Payment delays of 90+ days are standard
• Opaque royalty calculations prevent accountability

1.2 Ownership Ambiguity

• No universal system for proving creative contribution
• Collaborator disputes resolved through costly litigation
• Unauthorized sampling and theft are rampant
• Master recordings often controlled by third parties

1.3 Monetization Limitations

• Streaming pays fractions of a cent per play
• Fans have no way to invest in artists they believe in
• No mechanism for sharing success with early supporters
• Geographic and platform lock-in restricts reach

1.4 Protection Gaps

• Content ID systems favor large rights holders
• Independent artists lack enforcement resources
• Audio manipulation defeats traditional fingerprinting
• No chain of custody for creative works

2. Fundamental Features

SoundMint addresses these challenges through five integrated systems:

3. Audio Provenance & Watermarking

SoundMint protects sound in two layers: a Provenance Seal that cryptographically binds every pressed master to its content and creator (live today), and neural audio watermarking applied to the audio that actually leaves the vault (in development).

3.1 Evolution from Sequenced Sound Signatures

The original SoundMint concept proposed "Sequenced Sound Signatures"—audio watermarks built on steganography concepts using ultrasonic frequencies to embed data. While innovative, this approach faced limitations:

• Frequency-based watermarks vulnerable to compression
• Limited to ~100 alphanumeric characters
• Easily defeated by pitch shifting or time stretching

3.2 The Provenance Seal

When a master recording is pressed on-chain, SoundMint encrypts it (AES-GCM) and pins the ciphertext to IPFS. At encryption time, the platform computes a Provenance Seal—a deterministic authenticity hash that binds together:

• The exact audio content (content hash of the master)
• The creator's verified identity
• The moment of sealing

Because the seal is deterministic, anyone holding the original audio and the creator's identity can independently recompute and verify it. The pristine master itself never circulates: what exists publicly is an encrypted artifact whose authenticity is provable, and whose decryption is gated by token ownership.

3.3 Meta AudioSeal on the Delivery Path

For the audio that does leave the vault—public previews and per-listener streams—SoundMint is integrating Meta's AudioSeal, a state-of-the-art neural audio watermarking system:

Key Capabilities:

• Imperceptibility: Watermarks inaudible to human hearing
• Robustness: Survives compression and re-encoding
• Localization: Sample-level detection identifies exact watermarked regions
• Speed: Real-time embedding and detection
• Payload: Embeds creator ID, timestamp, and provenance data

The division of labor is deliberate: the master stays pristine and encrypted with its Provenance Seal, while every preview or stream that reaches the open internet carries a watermark tracing it back to its origin. Longer term, this enables per-listener forensic watermarking—each delivered stream individually marked.

4. Studio Workspace

4.1 Overview

The SoundMint Studio provides a home for sound creators to manage their entire catalog:

4.2 Features

4.3 Supported File Types

• Audio: WAV, FLAC, MP3, AAC, OGG, AIFF
• Projects: Stems, multitracks, session files
• Assets: Artwork, lyrics, metadata

5. Track Sources (Version Control)

5.1 Git-Inspired Audio Version Control

Track Sources brings software development's version control principles to audio production:

5.2 Collaborator Roles

5.3 Version Metadata

Each version captures:

• Contributor identity (verified via authentication)
• Timestamp of contribution
• Audio watermark with version hash
• Optional description/notes
• Approval status and approver

5.4 Collaboration Invitations

Track Sources supports first-come-first-serve invitations:

• Owner invites up to 50 potential collaborators
• First to accept claims the collaboration slot
• Others have invitations automatically voided
• Prevents permission conflicts and disputes

6. Three-Token Architecture

6.1 Overview

SoundMint's monetization system uses three distinct token types:

6.2 SongIPAsset (ERC-721)

The SongIPAsset is a 1/1 NFT representing the master recording intellectual property:

// Core Properties
uint256 public constant TOKEN_ID = 1;  // Always 1 (1/1 NFT)
bytes32 public ipfsHash;               // IPFS metadata hash
address public originalArtist;         // Creator address
uint256 public mintedAt;               // Mint timestamp
bool public isStaked;                  // Monetization status

Key Features:

• Proof of Ownership: Immutable on-chain record of creation
• Staking: A single stake() call transfers the NFT into its own MonetizationVault and deploys the song's Collector and Ownership tokens
• Unstaking: The NFT can be retrieved only while nothing has been sold—zero collector tokens minted and zero ownership tokens purchased. Once anything sells, the stake is permanently locked; unstaking also permanently retires monetization for that song
• ERC-2981 Royalties: 5% royalty to the artist on secondary sales, set at mint

6.3 OwnershipToken (ERC-20)

Revenue-sharing tokens for fans who want to invest in a song's success:

Configuration:

Rights Granted:

• Pro-rata share of collector token revenue
• Pro-rata share of licensing/sync and other external revenue
• Governance (future implementation)

(Primary bonding-curve proceeds go entirely to the creator collective—see Section 7.)

Rights NOT Granted:

• Copyright ownership
• Creative control
• Veto rights

6.4 CollectorToken (ERC-1155)

Access tokens for fans who want to support and access content:

Configuration:

Rights Granted:

• Streaming access in SoundMint player
• Download rights (if enabled by artist)
• Proof of support/early fandom
• Potential airdrops/perks from artist

Rights NOT Granted:

• Revenue sharing
• Ownership stake
• Governance rights

7. Monetization & Revenue Distribution

7.1 Revenue Flow Architecture

Creator/holder percentages of net are set per release by creatorCollectorShareBps (default 90%, bounded 51–95% on-chain); ownership primary-sale proceeds always go entirely to the creator collective — holders participate in collector and external revenue only.

7.2 Protocol Fee Structure

With the default configuration, a solo artist holding 100% of the ownership token receives 95% of gross collector revenue (85.5% via the creator bucket + 9.5% via the holder bucket). This is an identity, not a coincidence of the defaults — Appendix C derives it and shows it holds for any configured split.

7.3 Vault-Internal Split Enforcement

Revenue distribution is enforced inside the MonetizationVault — collaborator splits are committed as creatorAllocations (basis points per wallet, up to 50 wallets, validated to sum to exactly 100%, immutable after initialization) and settled as pull-based pending balances. Ownership-holder revenue accrues via a checkpointed per-token accumulator on the OwnershipToken contract:

Benefits:

• No custodial risk to third-party split infrastructure (funds never leave the protocol's own contracts)
• Pull-based claims — each payee can only ever withdraw their own balance
• Transparent on-chain accounting, exact to the USDC unit
• Splits are immutable once committed at stake time

0xSplits (SplitsWarehouse + PullSplitFactory) remains integrated as a legacy/secondary path only; it is not on the primary money path for v2 releases.

7.4 Quadratic Bonding Curve

Ownership tokens are priced using a quadratic bonding curve:

price(n) = a × n² + b × n + c

Where:
  n = token number in sequence (1 to tokensForSale)
  a = quadratic coefficient (price acceleration)
  b = linear coefficient (base growth rate)
  c = base price (minimum floor)

Artists configure the curve per release (the coefficients are committed on-chain as fixed-point integers at stake time and are immutable afterward).

Example Configuration (Standard preset):

a = 0.10 USD   (acceleration)
b = 0.50 USD   (linear growth)
c = 10.00 USD  (base price)

Token #1:   $10.60 USDC
Token #10:  $25.00 USDC
Token #25:  $85.00 USDC
Token #49:  $274.60 USDC

Economic Properties:

• Early supporters get lower prices
• Price increases as demand grows
• Natural price discovery without market manipulation
• Predictable total cost for any quantity

On-chain safeguards:

• The curve must be strictly increasing (a flat or decreasing curve is rejected at deploy)
• Deploy-time guardrails sanity-check the total raise, final token price, and worst-case basket cost
• Per-transaction and per-block purchase caps blunt sniping and flash-loan games

The curve's closed-form mathematics — exact basket costs, the sold-out raise, and a worked example — are given in Appendix C.

8. Enforcement & Permissions

8.1 Enforcement API

SoundMint provides an API for external platforms to detect and validate watermarked audio:

8.2 Permission Keys

Artists can grant usage permissions to non-owners:

8.3 Permission Key Structure

{
  "key_id": "pkey_xyz789",
  "song_id": "song_abc123",
  "grantee": "0x1234...5678",
  "permissions": ["stream", "embed", "monetize"],
  "platforms": ["youtube", "twitch"],
  "expires_at": "2026-01-15T00:00:00Z",
  "revenue_share_bps": 0
}

9. Sharing & Distribution

9.1 Share Links

SoundMint provides secure, granular sharing for audio files:

https://app.soundmint.com/s/[token]

Share Configuration:

9.2 Access Levels

9.3 Share Analytics

Artists receive analytics on shared content:

• View/play counts
• Geographic distribution
• Referral sources
• Download metrics (if enabled)

10. Technical Architecture

10.1 System Overview

10.2 Backend Services

10.3 Audio Processing Pipeline

When a track is pressed on-chain, the pipeline extends: the master is encrypted (AES-GCM), sealed with its Provenance Seal, and pinned to IPFS alongside the release metadata. AudioSeal watermarking is applied on the delivery path (previews and streams) rather than to the stored master.

10.4 Blockchain Integration

Payment Token: USDC (6 decimals) on Base

11. Smart Contract Specifications

11.1 Contract Architecture

Two supporting contracts complete the system: AlbumGroupFactory groups released songs into albums (a metadata-level grouping—albums are not themselves NFTs), and SoundMintRegistry maintains the global index of songs and authorizes which factories may register them.

11.2 Gas Optimization

EIP-1167 Minimal Proxy Pattern:

• Implementation contracts deployed once
• Each song uses ~$0.01 gas for deployment
• 99% gas savings vs full contract deployment

11.3 Security Features

11.4 Contract Addresses

The protocol is live on Base Mainnet (chain 8453) as Protocol v1.0, with ownership held by a 2-of-3 Gnosis Safe. Base Sepolia remains available as the public testnet. On the whitepaper page the block below is an interactive network switcher (Mainnet / Sepolia); the fence lists the mainnet deployment for GitHub readers.

Per-song contracts are deployed as EIP-1167 minimal-proxy clones of the implementation contracts above, so the addresses listed are the shared implementations plus the factories and registry. All are source-verified on Blockscout (canonical list in contracts/MAINNET.md); the Base Sepolia deployment mirrors this topology (contracts/TESTNET.md). 0xSplits is referenced only as an optional path for album-level splits; per-song revenue is distributed entirely inside the MonetizationVault (Section 7.3).

12. Conclusion

SoundMint's ultimate purpose is embodied in its name—a triple entendre:

1. Mint information to sound files through audio watermarking
2. Mint ownership to the blockchain through tokenization
3. Mint authenticity as an authority of provenance

By combining cutting-edge audio watermarking, blockchain-based ownership, and fair economic mechanisms, SoundMint creates a new paradigm where:

• Artists retain control and earn fair compensation
• Fans can invest in music they believe in
• Collaborators receive verifiable credit
• The ecosystem operates transparently on-chain

The vision from 2022 has evolved into a comprehensive protocol ready to transform how sound is created, owned, and monetized.

13. References

1. Sound Recording History - Wikipedia https://en.wikipedia.org/wiki/Sound_recording_and_reproduction

2. Putting the Band Back Together: Remastering the World of Music - CitiGroup (2018) https://ir.citi.com/

3. DAO - Decentralized Autonomous Organization https://ethereum.org/en/dao/

4. NFT - Non-Fungible Token https://ethereum.org/en/nft/

5. Interplanetary File System (IPFS) https://ipfs.io/

6. Meta AudioSeal: Proactive Detection of AI-Generated Audio https://github.com/facebookresearch/audioseal

7. EIP-1167: Minimal Proxy Contract https://eips.ethereum.org/EIPS/eip-1167

8. 0xSplits Protocol https://splits.org/

9. Base L2 (Coinbase) https://base.org/

10. Git - Distributed Version Control System https://git-scm.com/

Appendix A: Glossary

Appendix B: Token Economics Example

Scenario: Solo artist releases a song with the standard curve preset and the minimum creator retention.

Purchase prices along the curve (USDC):

If the sale sells out, the raise and its split — primary curve proceeds go to creators, not holders:

Collector tokens are priced at $5.00 USDC each — artist-set, and locked after the first mint. Each sale takes the 5% protocol fee, then the net splits 90% creator collective / 10% ownership holders (the default). A solo artist holding 100% of the ownership supply therefore nets 95% of gross collector revenue.

Appendix C: Protocol Mathematics

The protocol's economics reduce to three pieces of arithmetic — each simple enough to verify by hand, and each enforced exactly, in integer math, by the deployed contracts. This appendix states them formally. Amounts are USDC throughout; one basis point (bps) is 0.01%.

C.1 The Bonding Curve in Closed Form

The marginal price of the $n$-th ownership token is quadratic:

Because the price is polynomial in $n$, any basket of tokens has an exact closed-form cost — no iteration, no approximation. Using the standard power sums

the cost of buying $q$ tokens when $s$ have already been sold is

and the total raise if the sale sells out is

This is precisely the computation the OwnershipToken contract performs on every purchase — a single O(1) evaluation, accumulating before dividing so that rounding is applied exactly once. The deploy-time guardrails evaluate the same closed forms to sanity-check a curve's total raise, final token price, and worst-case basket cost before it can go live.

Worked example — the standard preset from Appendix B, with $a = 0.10$, $b = 0.50$, $c = 10.00$ and $N_{\text{sale}} = 49$:

Supply bookkeeping: of a maximum supply $N$, the creator collective retains $\lceil \beta_c N \rceil$ tokens — retention rounds up, so the configured majority is never under-minted on small supplies — and $N_{\text{sale}} = N - \lceil \beta_c N \rceil$ tokens are offered on the curve.

C.2 The Distribution Algebra

Fix the protocol fee $\varphi = 0.05$ (hard-coded at 500 bps). Each release commits, immutably at stake time:

• $\beta \in [0.51,\, 0.95]$ — the creator collective's share of net collector and external revenue (default $0.90$)
• creator allocations $\alpha_1, \dots, \alpha_k$ with $\alpha_1 + \cdots + \alpha_k = 1$ and $k \le 50$

Let $\omega_h$ be holder $h$'s fraction of the ownership token supply. A collector-sale or external inflow of $P$ USDC is distributed:

An ownership primary-sale inflow of $P$ pays $(1-\varphi)\,\alpha_i\,P$ to each creator wallet and nothing to holders.

Conservation. Because $\sum_i \alpha_i = 1$ and $\sum_h \omega_h = 1$, the branches always sum back to the inflow:

Nothing is created, nothing is lost, and no balance is stranded — by construction, not by policy.

The 95% identity. A solo artist ($k = 1$, $\alpha_1 = 1$) who retains the full ownership supply ($\omega = 1$) receives, from every collector dollar:

— independent of $\beta$. The creator and holder buckets always recombine to the whole of the net. With the default $\beta = 0.90$, the composition is 85.5% via the creator bucket plus 9.5% via the holder bucket; selling ownership tokens moves value between the buckets, never out of them. This is why the protocol can guarantee the artist's floor rather than merely promise it.

C.3 The Holder Accumulator and Its Invariants

Holder revenue uses a per-token revenue index (the Synthetix staking pattern). Let $T$ be the ownership token supply. On each holder-bound distribution of $\Delta$:

and a holder $h$ with balance $b_h$ may claim

On-chain, $R$ is kept in $10^{18}$ fixed point and checkpointed on every transfer, so revenue accrues to whoever held the tokens at the moment of each distribution: buying later does not dilute earlier distributions, and selling does not forfeit accrued claims. Sub-unit dust that cannot divide evenly is buffered into the next distribution rather than lost.

Invariants. The following properties are not aspirations — they are machine-checked against the contract code with randomized fuzz and invariant testing:

1. Global USDC conservation — treasury, vault, and token-contract balances plus everything already claimed equal everything ever deposited.
2. Creator majority — creatorCollectiveBps ≥ 5100, $\beta \in [0.51,\, 0.95]$, and allocations sum to exactly 10,000 bps.
3. Monotone index — $R$ never decreases.
4. Solvency — the sum of all claimable balances never exceeds the contract balances backing them, for creators and holders alike.
5. Supply bounds — tokensSold ≤ N_sale, and total supply never exceeds the configured maximum.

This whitepaper is a living document and will be updated as the protocol evolves.

Last Updated: July 10, 2026 Version: 1.0