At least theoretically, there is no reason to assume that short-phase creative bursts are necessarily different than the longer processes that you mention, but I imagine it is substantially harder to study because- and I'm obviously ignoring the most practical concerns of "can someone be truly creative while undergoing fMRI", etc- the longer process of creating a song often calls upon multiple brain functions. The creation of a simple melody might involve mental memory (thinking of songs that sound similar, running multiple ways to evolve the idea in parallel, associated memories that the piece either recalls or draws him), muscle memory (thinking about how you might play it), emotions, etc etc. Each of those is not a single part of the brain, too, so you might see a LOT of areas lit up over time as the person sits and creates. It doesn't necessarily generate useful information as it is such an active process.
They've tried this here at Stanford and it wasn't a particularly useful endeavor. It doesn't lend itself to a useful control/baseline (like Dr. Limb had between playback of the tune's head and the traded soloing).
They've tried this here at Stanford and it wasn't a particularly useful endeavor. It doesn't lend itself to a useful control/baseline (like Dr. Limb had between playback of the tune's head and the traded soloing).

