LOGO

Songscription: AI Shazam for Sheet Music - Identify Music Instantly

June 30, 2025
Songscription: AI Shazam for Sheet Music - Identify Music Instantly

Songscription Launches AI-Powered Music Transcription Service

Last week marked the debut of Songscription, a new company utilizing AI models to automate the process of music transcription. The service swiftly converts audio files of songs into corresponding sheet music.

The product operates on a freemium model, designed to appeal to both professional musicians and those pursuing music as a hobby.

Vision for Accessible Music Education

Andrew Carlins, CEO of Songscription and a student within Stanford’s combined MBA/MA in Education program, articulated the company’s goals to TechCrunch. “Our aim is to enhance the enjoyment derived from playing music,” he stated.

Carlins envisions a future where educators in resource-constrained environments, such as a high school band teacher in rural Nebraska, can readily obtain sheet music for requested songs.

Furthermore, this music could be specifically arranged for the instruments available within the band and tailored to the individual skill levels of each student.

Current Capabilities and Future Development

Currently, Songscription supports music transcription for a variety of instruments. However, the piano model demonstrates the highest level of accuracy.

Looking ahead, the company intends to expand its offerings to include alternative transcription outputs, like guitar tabs.

They also plan to facilitate arrangements for entire bands, rather than focusing solely on single-instrument transcriptions.

Streamlining the Musician's Workflow

This technology proves beneficial for musicians who record their compositions and then require sheet music.

It eliminates the time-consuming task of manual transcription, allowing them to focus on creative aspects of their work.

For individuals unfamiliar with reading or writing sheet music, Songscription generates a piano roll, providing a visual representation of the music played on a virtual piano.

YouTube Integration and Copyright Considerations

The platform allows for automated music transcription directly from YouTube links.

While users are required to acknowledge they possess the rights to transcribe uploaded files, the ease of confirming this could potentially facilitate access to sheet music for copyrighted material.

Navigating the Legal Landscape

Carlins addressed the legal complexities surrounding AI-assisted music creation. “Given the allowance to listen to a song, notate it by ear, and perform it privately, it’s debatable whether utilizing a platform to accelerate this process infringes on any legal boundaries.”

He acknowledged the evolving nature of the legal field and the potential for operating within a gray area.

Recent court rulings appear to be leaning in favor of technology companies in similar disputes.

Augmented Notation, Not Creation

Songscription distinguishes itself by not generating entirely new, AI-composed music.

Instead, it functions as a tool to expedite the process of creating guitar tabs or sheet music for musicians.

“We position ourselves as an augmented music notation software, designed to accelerate the transcription process,” Carlins clarified, emphasizing the platform’s role in assisting, rather than replacing, human musicians.

Underlying Technology and Data Sources

The AI model powering Songscription is built upon research published in a paper co-authored by the company’s co-founder, Tim Beyer, and researcher Angela Dai.

To acquire the necessary training data, Songscription collaborated with musicians who shared or sold their piano performances and sheet music.

Public domain sheet music was also utilized, but the majority of the training data was synthetically generated.

This involved converting sheet music into audio and then introducing alterations to mimic real-world conditions, such as background noise and reverb.

Funding and Acceleration

Within just seven months of its founding, Songscription secured pre-seed funding from Reach Capital.

The company is also participating in the Stanford’s StartX accelerator program, further bolstering its development and growth.

#sheet music#ai#music identification#song recognition#songscription#shazam