AudioCipher V3

Categories: Software Prototypes/Non-Commercial Products

Submitted by:AudioCipher Technologies

24 May 2023

Elevator Pitch

AudioCipher V3

AudioCipher is a MIDI Generator Plugin that turns words into melodies and chord progressions, in your key signature and rhythmic style of choice.

Product Description

AudioCipher V3

AudioCipher is the first DAW plugin to offer a text-to-MIDI converter for creative inspiration. Users type in words and transform them into melodies and chord progressions. Take control of parameters like key signature, scale, chord extensions, and rhythm automation. Leverage the randomization features to audition unlimited variations. Find a starting point and build from there!

How It's Innovative

> Innovation #1: Built a JUCE plugin from a lost musical code, popular among European composers 100 years ago AudioCipher's algorithm was inspired by a century-old musical code from France, known as the French musical cryptogram. Composers used this cipher to convert their names into short musical phrases. They used the musical phrase as a motif in their songs. We created an app that makes these same transformations instantly, as a source of creative inspiration for producers and composers. > Innovation #2: Expanded on the original cipher, to cover 12 notes, nine scales and six chord extensions. The original cipher featured one scale only: A minor. We extended this basic model across the twelve root notes, seven diatonic modes, harmonic minor and the 12-note chromatic scale. We also introduced the idea of chord extensions and rhythm automation, with randomization features to create unlimited variations. > Innovation #3: Positioning ourselves for an emergent Artificial Intelligence revolution (text-to-music APIs) In the past three months, three major AI text-to-music services have emerged: Google MusicLM, Riffusion, and Mubert. I believe that this is an indicator that a wave of music generators will soon emerge, supporting text-to-music prompts like ChatGPT and Midjourney. There is currently a plugin called Neutone that acts as the host for multiple AI music APIs, but they perform tone transfer only (not MIDI generation). There is no text-to-music functionality anywhere in Neutone. AudioCipher is positioning itself to play that role in the near future. We're creating a user interface for something that doesn't exist yet: a set of text-to-music APIs linked up to neural networks, to generate MIDI and audio on command. I believe that these APIs will become available within 1-5 years. Our app will act as the interface that takes text prompts from users, key signature, chord and rhythm information. When the user hits a "generate" button, AudioCipher will send that packet to a neural network in the cloud and retrieve a musical segment to drag and drop onto a track in their DAW. While we wait for these APIs to become available, our team is working on other important functionality that's not dependent on AI. V4 will have a save button and a separate MIDI library plugin to help producers keep their ideas organized.

See MIDI Innovation In Action

Most Inspiring Use Cases

AudioCipher lets music producers instantly generate variations on a musical idea, using an unconventional text-to-MIDI technique. One customer bought AudioCipher and used it to write his wife a song for their anniversary. I thought that was a sweet and inspiring use of the app. Our team has used AudioCipher to create dozens of demo videos, across a variety of musical genres and styles. Each demo uses a different collection of words to spin up a unique song from scratch, without artificial intelligence! https://www.youtube.com/@audiocipher/playlists In the future, if we do pursue an integration with text-to-music APIs, we could be the first plugin to support AI text-to-MIDI generation in the DAW.

Expansion Plans

We are working on version 4.0 and hope to publish it by August 2023.


We went to market with V1 in early 2021, V2 in late 2021, and V3 in late 2022. Our company currently earns $2.5-5k/month depending on promotions and organic traffic performance.