Home / Lab / Chorus lift test

Result in this run

All four A records kept the chorus near the verse scale. B1, B2, and B4 produced a clearly larger chorus; B3 did not. A2 and B4 were the closest musical pair, with corresponding chorus entrances at 0:40 and 0:39. This is a repeated field comparison, not a seed-matched controlled experiment.

Listen to the closest musical match

Start flat, then hear the chorus open

A2 and B4 were the closest pair in overall style, melody, singer character, and timing. Start just before each chorus entrance. Listen for section contrast rather than loudness alone: drum weight, layer density, stereo width, bass movement, brightness, vocal doubles, and backing harmonies can all contribute to the arrival.

A2 · intentionally flat chorus

Creator observation: the chorus does not feel clearly larger than the verse. Start at 0:40.

Cover artwork for Open Windows A2 Flat Chorus
Use the official player, or open A2 at 0:40.
B4 · clear chorus lift

Creator observation: the chorus becomes clearly larger than the verse. Start at 0:39.

Cover artwork for Open Windows B4 Chorus Lift
Use the official player, or open B4 at 0:39.

What stayed stable

The shared title was Open Windows. All eight generations used the same lyrics, Suno v5.5, warm female lead, electric piano, muted electric guitar, bass, acoustic drums, and clean mid-tempo indie-pop production. The comparison changed the requested section arc: A asked the verse, pre-chorus, and chorus to remain nearly equal; B held the verses back and asked the chorus to add several coordinated lift mechanisms.

A conditionNearly equal energy and loudness across sections; narrow stereo image, minimal layering, no chorus-only vocal doubles or harmonies, and intentionally flat chorus dynamics.B conditionRestrained verses followed by fuller drums, wider guitars, stronger bass movement, brighter tone, lead-vocal doubles, and stacked backing harmonies in the chorus.RepeatsFour generations per condition; every record was retained rather than selecting only successful outputs.Listening focusThe perceived size change at the verse-to-chorus boundary, not raw loudness alone.
Shared lyrics
[Verse 1]
Streetlights flicker on the glass
Another quiet evening passes
I hear the city breathe below
But there is somewhere else to go

[Pre-Chorus]
I have been waiting for a sign
Something to pull me through the night

[Chorus]
Open windows, let the daylight in
Let the air move underneath my skin
I was standing still, but now I know
Open windows, I am ready to go

[Verse 2]
Old words gather in the hall
Like faded pictures on the wall
I leave the silence where it lies
And watch the morning change the sky

[Pre-Chorus]
I have been waiting for a sign
Something to pull me through the night

[Chorus]
Open windows, let the daylight in
Let the air move underneath my skin
I was standing still, but now I know
Open windows, I am ready to go

[Bridge]
No more holding back the sound
No more keeping my feet on the ground

[Final Chorus]
Open windows, let the daylight in
Let the air move underneath my skin
I was standing still, but now I know
Open windows, I am ready to go

Shared Exclude: EDM drop, trap hi-hats, cinematic orchestra, choir, key change, tempo change, rap, spoken word, heavy distortion, lo-fi production.

All eight generation records

Retain B3 as part of the result

RecordRequested conditionCreator listening observationClassification
A1Flat chorusNo clear increase in chorus size.Clear condition match
A2 · 0:40Flat chorusNo clear increase in chorus size; closest musical match to B4.Clear condition match
A3Flat chorusNo clear increase in chorus size.Clear condition match
A4Flat chorusNo clear increase in chorus size.Clear condition match
B1Chorus liftThe chorus becomes clearly larger.Clear condition match
B2Chorus liftThe chorus becomes clearly larger.Clear condition match
B3Chorus liftNo clear increase in chorus size.Direction not clearly followed
B4 · 0:39Chorus liftStrongest verse-to-chorus contrast in the B set.Clear condition match

What the result supports

The selected pair supports one direct listening observation from this review: A2 keeps the chorus near the verse scale, while B4 creates a clearer arrival. The full set supports a second observation about consistency in this run: the flat condition appeared in four of four A generations, while the lift condition appeared in three of four B generations.

Evidence boundary: the B prompt changed a bundle of chorus-arrangement cues, so this comparison cannot tell us which individual mechanism was responsible. The generations also did not share a fixed seed. The records show what happened in this eight-generation run; they do not prove that the wording will transfer at the same rate in another session.

Back to the evidence collection: Browse all Lab studies