Compare melody variations with saved examples
Nine public piano recordings let you compare three Motiflo controls against saved note data. The audio, settings and project files come from the same examples.
What stays the same
Samples S01–S09 all use piano, C major, 4/4, 96 BPM, two bars and random seed 104729. Within each group of three, one control moves from 0% through 50% to 100%; the other saved controls remain at 50%. The samples have no user-entered opening notes. These facts are visible under “Full generation settings” on the listening page.
Color: S01, S02 and S03
Listen to the first four notes. S01 begins E4–G4–F4–E4 at 0% color. S03 begins F4–B4–A4–A4 at 100%. This is a concrete change in the saved pitches, not a claim that one version is better. Compare how the two openings move, then listen for what happens at the end of each clip.
Play S01 · Play S02 · Play S03
Energy: S04, S05 and S06
These three start on E4, but their rhythms are not identical. At 0% energy, S04 starts with a quarter-note E4 and contains 11 note events. At 100%, S06 starts with an eighth-note E4 and contains 14 events. The saved project uses 480 ticks per quarter note. Try tapping a steady beat while switching between the recordings, then decide which spacing suits your own phrase.
Play S04 · Play S05 · Play S06
Tension: S07, S08 and S09
S07 starts E4–G4–G4–F4 at 0% tension; S09 starts F4–A4–A4–G4 at 100%. Listen for changes in pitch movement and the way each phrase resolves. The control affects the generation request, but neither the label nor the percentage can tell you whether a listener will experience tension.
Play S07 · Play S08 · Play S09
How to use the comparison
Choose one difference you heard and recreate the experiment with your own settings. Keep the instrument, key, meter and tempo fixed; change a single control; generate a fresh batch; then compare both the sound and score. Your results need not match these recordings. The violin samples S10–S12 are useful for hearing different meter presets, but they also have different random seeds, so they do not isolate meter as the only cause of a change.