Advanced Generation Tips
Get more out of the Studio by combining parameters intentionally
You already know how to generate a worship track. These are the small adjustments that move a song from "good enough" to "this is the one we'll use Sunday."
Stack parameters with intent#
The Studio exposes style, mood, theme, tempo, key, vocal style, instrumentation, theme, and duration. The mistake most people make is leaving everything on the default and only changing the prompt. The opposite works better.
Pick three parameters that matter for the moment you're writing for, set them firmly, and leave the rest. A few combinations that produce predictable results:
| Goal | Style | Mood | Tempo | Vocal |
|---|---|---|---|---|
| Sunday opener | Contemporary | Joyful | 110–120 BPM | Mixed |
| Communion | Acoustic | Reverent | 70–80 BPM | Female solo |
| Youth night | Gospel | Celebratory | 120+ BPM | Mixed |
| Reflection / prayer | Instrumental | Peaceful | 60–70 BPM | None |
| Hymn modernisation | Hymn | Reflective | 80–90 BPM | Choir |
Match tempo to congregational singing#
Anything over 130 BPM is hard to sing congregationally without rushing. Anything under 60 starts to drag. Most modern worship sits between 70–120 BPM for a reason — that range matches comfortable group breathing.
If you set a tempo and the result feels off, regenerate with a 10 BPM swing rather than re-prompting. The model behaves more predictably to tempo nudges than to "make it faster" in the prompt.
Use the theme field, not the prompt, for occasion#
"Christmas," "Easter," "Wedding," "Communion" all behave better as explicit Theme selections than as adjectives in the prompt. The theme dropdown maps to internal templates that have been tuned for the occasion; prompt adjectives compete with the rest of your prompt for attention.
Be specific about scripture rather than thematic#
A prompt like "a song based on Psalm 23, focused on the shepherd metaphor" produces better lyrics than "a peaceful song about God's care." Specific references give the lyric generator concrete imagery to anchor to.
Generate three takes, then pick#
The fastest path to a track you love is to generate the same prompt three times with the same settings and audition all three. Generation is not deterministic — each take makes different musical choices in the bridge, the chord voicings, the vocal phrasing.
This is faster than re-prompting and burns roughly the same quota.
When a take is "almost right"#
Use Section Regeneration. Don't re-roll the whole song to fix a weak bridge — regenerate just the bridge and keep the verses you liked.
Save the prompt that worked#
Every track records the prompt + settings that produced it. When you find a combination that consistently delivers what you want for a moment in your service (offering, communion, response), save it as a Custom Style and reuse it.
Related Articles:
Was this article helpful?
Your feedback helps us improve our documentation.