From July 14 to 17, 2026, the AMORPH founding team developed custom conversion technology and a workflow that turn the vocals of a song made with Suno into a real person's voice. As a result, the client could work from the generated vocal files alone, without the hassle of extra recording.
Suno is an AI music generation service: describe the genre and mood of the song you want in text, or enter lyrics you wrote yourself, and it creates a song complete with vocals and instrumentation (Suno official site). This article does not disclose details of the client or the song. It covers, in order, the problem the client faced, how it was solved and the result.
Can Suno vocals be converted into a real person's voice?
In this project, yes. The client had struggled with metallic noise when overlaying a voice using commercial tools. After the AMORPH founding team analyzed the generated vocals and developed a custom conversion process, the client could work from the generated vocal files alone, with no extra recording.
This answer applies only to this project. This one case does not show that the same result would come from changing Suno vocals with other voices or other songs.
Problem: vocals overlaid with commercial tools had metallic noise and off-pitch notes
The client wanted to convert the vocals of a song made with Suno into a real person's voice. They trained a voice with existing commercial tools and overlaid it on the vocals separated by Suno, but noise such as metallic sounds and off-pitch notes crept in, making the work difficult.
- Voice cloning is a technology that trains AI on a person's voice in order to generate new speech in that voice or convert other speech into it. Training a voice with commercial tools, as the client did, is this kind of approach.
- A separated vocal is a file with only the vocals taken out of a finished song. In this project, Suno did the separating; what the AMORPH founding team developed was the conversion technology and workflow that turn generated vocals into a real person's voice.
- An off-pitch note is a sung note that drifts away from the intended pitch.
Solution: analyzing the generated vocals and developing a custom conversion process
The AMORPH founding team closely analyzed the characteristics of the generated vocals and developed custom conversion technology and a workflow so that the real person's tone and expression come through naturally.
Could the client work from generated vocals alone, with no extra recording?
Yes. The client now has an environment where they can work comfortably from the generated vocal files alone, without the hassle of extra recording.
This result came from the AMORPH founding team visiting the client in person, hearing where they were stuck and reviewing several possible solutions.
Project details
The work ran from to . The field is sound AI, specifically vocal conversion.
Details of the client and the song are not disclosed. This case has no measured figures; the result describes how the client's working environment changed.