Have you ever imagined customizing your favorite artist’s music to match your personal style? For example, adding a distorted guitar sound to Taylor Swift’s latest song or transforming NewJeans’ “Ditto” into a lullaby.

Unfortunately, several reasons have prevented us from doing this:

1. All music released nowadays is distributed in a mixed-down, single audio file format.
2. no clear standard exists for utilizing copyrighted works at the stem level.
3. The necessary technology has not been sufficiently developed.

But don’t be disheartened. It wasn’t impossible; we were confined to the world of Music 2.0. Thankfully, time has passed, and now we are beginning to discuss ways to customize music to our tastes. A prime example is when Kanye West, a prominent American hip-hop artist, introduced the ‘Stem Player’ in 2021, allowing listeners to remix his tenth studio album.

These innovative experiments have significantly raised awareness about music elements like ‘stems’ among listeners. Following this, with the emergence of Web 3.0 and particularly generative AI, music technology is undergoing significant transformations. We are now introducing MixAudio, a multimodal AI music generator optimized for the era of generative AI.

MixAudio is designed as a perfect creative playground for the generative AI era, where every creative idea can easily be turned into music. We have devised a new music unit called “block music,” allowing creators to connect music modules as effortlessly as connecting LEGO blocks. We have modularized artists’ rhythms, instruments, vocals, and more for easy assembly.

In the music creation process, MixAudio configures music blocks into six stem units: rhythm, bass, mid-range instruments, high-range instruments, melody (top line, vocals), and sound effects. It separates the structure of popular music into sections like intro, verse, chorus, bridge, and outro, ensuring smooth transitions no matter how they are combined.

MixAudio's AI engine handles the seamless assembly of these diverse music blocks. This engine rapidly collects, transforms, and generates music by gathering optimal blocks that meet various user requirements. Most notably, it boasts a rapid generation speed of up to four songs every two seconds and offers high convenience for easy editing according to preferences and purposes.

This adaptive music has been used in-game music and sound for years. It offers a familiar and convenient environment for game developers, video editors, and designers who need customizable music.

Like delicious food requires fresh ingredients, the MixAudio engine uses high-quality sound sources and samples (cleared of copyright issues) as its building blocks. This method has a significant advantage: it resolves copyright issues even with AI-generated music. Since it’s possible to track how much of each music block was used, it ensures fair revenue distribution to the producers/artists who provided the blocks.

Moreover, MixAudio’s unique music blocks and modular AI engine's future is limitless due to their high scalability. MixAudio can provide dynamic, adaptive music suitable for various media environments, such as games, mobile devices, autonomous vehicles, and home appliances.

MixAudio is taking steps towards implementing various platforms where anyone can joyfully explore the possibilities of music creation, from interactive/adaptive streaming music services and efficient custom sound generation studios to user-generated content publishing, covering the entire process from searching and distributing to streaming. This is the future ecosystem of Music 3.0!