I decided to give it a real workout on a song I’ve been tweaking for years. Out of ten recommendations that were different from what I thought, I rejected seven. Some of those were inadequate but not really wrong, just incomplete. I gave examples above. Two made me reconsider my work and, while OK, were not what I think best. But one was something I accepted.

I’ve made money for 50 years doing transcriptions. PG Music even paid me. I’m slow, but I’m good. This app is the first time I’ve ever changed my mind based on a computer program’s recommendation. I think the reason is that an algorithm has no preconceived ideas about what it hears. I see a chord progression and I think, sure, that’s common, expected, and fits here, so that’s done. Except, maybe there’s another note in there. And maybe that note is masked by the more obvious bass and melody tones so I overlooked it, but the computer program heard all notes as equally important in the calculation.



BIAB 2025 Win Audiophile. Software: Studio One 7 Pro, Swam horns, Acoustica-7, Notion 6, Song Master Pro, Win 11 Home. Hardware: Intel i9, 32 Gb; Roland Integra-7, Presonus 192 & Faderport 8, Royer 121, Adam Sub8 & Neumann 120 monitors.