Finding the sound, and why the first method failed for songs

The first musical choice was between a historical reconstruction of fifteenth-century music and a contemporary interpretation of a fictional medieval world. We chose the second.

The palette includes hurdy-gurdy, lute, fiddle, bagpipe, wooden flutes, shawm, drums, handclaps and stamping rhythms. Heavier guitars are allowed when they serve the identity. The working rule: an electric guitar that raises the tension can work, but a generic rock production with a bagpipe laid over it is something else.

The first production method kept a cyclical, repetitive structure. That suited dance music. It did not suit songs that need verses, a chorus, a build and a clear vocal line. The identity was interesting, but the songs did not develop.

So the method changed. For the narrative songs, the style is now described directly, with the emphasis on song structure, vocal dynamics, refrains and dramatic shape. The earlier material was kept, but it no longer had to be the basis for every track.

The lesson: source material that works for an instrumental piece is not necessarily a good starting point for a vocal song. The tool does not remove that design question. It makes picking the right method more important.

Two further points:

• Genre labels. The internal identity is medieval folk with folk-rock influence. When a distributor only offers broad options, rock/folk fits better than pop/folk because of the physical energy. Not every climax needs guitars, though: drums, hurdy-gurdy, group vocals and repetition can carry it.
• Open problems. Vocals vary from track to track in timbre, apparent age and pronunciation, and we have no reproducible fix yet. The options are stricter selection, a consistent human vocalist (more cost, stronger identity) or accepting some variation. Dutch lyrics add their own issue: a line can be correct on paper and still put the stress in the wrong place when sung, so lyrics are judged as performances, not as text.