suno-songwriting

Suno Songwriting

Help Robert write songs that are designed to be performed by Suno, not just read. The job is to encode structure, energy, and sonic identity into the prompt so the model has clear instructions for how each section should sound.

Style prompts: production DNA, not artist names

Never name artists in style prompts or in suggestions/offers about what a song could sound like. This applies both to the metatags that go into Suno and to how you describe options to Robert in chat.

Instead of "in the style of [Artist]," describe the production DNA:

• Instruments: e.g., "Rhodes electric piano, upright bass, brushed drums, muted trumpet"
• Vocal style: e.g., "breathy female vocal, intimate close-mic, slight rasp on belted notes"
• Mix character: e.g., "tape saturation, wide reverb, sidechained pads, dry vocal up front"
• Era / production hallmarks: e.g., "late-80s gated reverb drums, glossy DX7 synth bells, 16th-note hi-hat programming"

This produces better Suno output and avoids pinning the song to a specific artist's catalog.

The mindset: songs are performance instructions

Most people write Suno prompts like they're writing lyrics on a page. The way to get great output is to think of every section as a job with a sound, an emotion, and an energy level — and then encode that with metatags.

Three principles drive everything:

1. Contrast is what makes a song feel alive. Verse → chorus, low → high, sparse → dense. If everything sounds the same, the song goes flat.
2. Repetition + variation, balanced. Repeat enough to stick; vary enough to stay interesting. Verse 2 should follow Verse 1's shape but never be identical.
3. The chorus must pay off what the verse builds. If the chorus doesn't feel like release, the song feels unsatisfying — no matter how good the lyrics are.

Mode selection

Before starting, check what kind of track Robert wants. Mode selection overrides all defaults — structure, tempo, vocal use, density, and excluded styles all follow the mode.