Original songs, now shared

Traditional singer-songwriting and modern digital production

About the use of AI to Self-Produce My Songs

Over the past 40 years, I recorded original songs by singing and playing guitar or piano along with selected drum tracks. My lyrics and base compositions remain 100% without AI. Now retired, I use AI to produce and share my songs. I input my recordings to train AI, create, curate and edit better music, and keep full artistic control over the final music and style. I also input an hour of my live singing into AI tools to create a studio session voice (a clone of my singing) for my vocals with faster, better, and more flexible recording. Many fully AI-generated songs and singers (not real) today can sound "scary great," and they should, as they are a mashup of the greatest live singers, songs, and musicians in the world. I could do the same. Instead, I try to use my voice and original songs, as they were and still are, now re-created.

My guiding principle: The most important part of any new song is you, the listener.

AI as a tool

Live Performances, AI, and the Production Middle Ground

Apart from rare "one take" recordings, producers and music artists have long spent many hours in studio sessions creating and refining songs. Clips below from a live performance and a raw song track demonstrate how AI helped me — after much mixing and editing work — produce better songs. I've found vocal cloning to have pros and cons. It can enhance a voice artistically, reflect well or poorly, and it often slips up. Any music producer must select and build from raw materials.