ASK JACK Member Resource Library · AI Music Repair Vault

Fix My Generation

100 AI music problems + correction directions. Start with what you can actually hear, identify the most likely cause, make the smallest credible correction, and protect what already works.

Vocal Problems Lyric, Syllable & Phrasing Problems Hook & Chorus Problems Structure & Arrangement Problems Genre Drift Problems Drums, Groove & Pocket Problems Bass, 808 & Low-End Problems Instrumentation, Density & Balance Problems Energy, Mood & Emotional Problems Prompt, Consistency & Generation Problems Decision Tree Protected Traits Repair Log A/B Repair Test When Prompting Stops Completion

How to Use Jack Righteous Prompt Vault Hub Find Your Sound

Included with JR access

Available through JR Creator Library — One-Time, Creator Pro, or Complete Access.

This is an APPLY repair layer. It helps you diagnose a generation quickly. When the issue requires deeper system control, follow the linked BUILD resources instead of endlessly rewriting prompts.

Before You Repair

Prerequisite

Start with a survivor

Use this vault when you already have a generation worth improving. If nothing in the version is worth protecting, return to the Prompt Vault Hub and rebuild from the right lane instead of repairing noise.

Input

Name one audible problem

Bring the current version, the problem you can hear, 3–5 protected traits and the smallest variable you believe may solve it.

Output

One tested repair decision

Leave with an A/B comparison, evidence of what improved or degraded, and a Keep / Revise / Reject decision.

Limitation: correction language does not guarantee a platform will preserve every requested trait. When prompting cannot isolate the repair, move to lyric editing, section replacement, stems, mixing or deeper BUILD work instead of repeatedly regenerating.

How to Use This Vault

1 · Hear

Describe the symptom

Use plain language first. Do not jump from “the chorus is muddy” to an assumed technical cause.

2 · Protect

Name what must survive

Write down the keeper traits before changing anything so the repair has boundaries.

3 · Test

Change one variable

Run the smallest credible correction, compare against the survivor, then decide whether to keep, revise or reject the repair.

Diagnosis Decision Tree — Fix the Right Layer

1 · Hear

Name the problem without guessing the cause

“The chorus is muddy” is better than “the prompt is bad.” Describe the audible symptom first.

2 · Locate

Which layer owns the problem?

Vocal, lyric, hook, structure, genre, groove, low end, instrumentation, energy/mood or prompt/generation behavior.

3 · Repair

Change one variable

Choose the smallest correction likely to address the symptom while preserving everything that already works.

Repair rule: if you cannot say exactly what variable you are changing, you are probably changing too much.

Protected Traits Box — Do Not Fix the Song by Breaking the Song

Before revising, write down what survives

Groove, tempo feel, lead-vocal character, central motif, hook phrase, emotional tone, structure, bass role, genre center or any other trait that already works.

Use explicit preservation language

preserve [SURVIVOR TRAITS]; change only [TARGET VARIABLE]

This will not guarantee platform compliance, but it creates a clearer controlled test.

01 · Vocal Problems — 10 Problems

The singer over-sings

Likely cause: Too much pressure, melisma, range or ad-lib activity.

Smallest correction: Reduce one expressive variable before changing the voice itself.

Copy / Adapt

restrained lead, selective runs only, shorter phrase endings, no ad-libs under important lyric words

Listen for: Clearer lyric focus with emotion still present.

The vocal is too breathy

Likely cause: Breathiness is masking consonants or weakening the hook.

Smallest correction: Keep intimacy but increase articulation.

Copy / Adapt

intimate close lead, soft onset, clearer consonants, reduced breath noise on title words

Listen for: Words should read clearly without losing closeness.

The voice is too theatrical

Likely cause: Vibrato, diction and dynamics are exaggerated.

Smallest correction: Move toward conversational delivery.

Copy / Adapt

natural conversational vocal, restrained vibrato, dry phrasing, minimal dramatic pauses

Listen for: Performance should sound less staged and more direct.

The vocal is too flat emotionally

Likely cause: Timing and phrase shape are static.

Smallest correction: Change phrasing before adding more runs.

Copy / Adapt

controlled emotional arc, delayed phrase entries, stronger emphasis on key words, selective sustained notes

Listen for: Emotion should come from timing and pressure, not constant ornamentation.

The register is wrong

Likely cause: The melody sits too high or low for the intended character.

Smallest correction: Protect melody shape but redirect the range center.

Copy / Adapt

lower-register lead, compact verse range, selective lift only in the chorus

Listen for: The voice should feel more natural without rewriting the song.

The words are hard to understand

Likely cause: Dense runs, doubles or breathiness are masking diction.

Smallest correction: Simplify support and protect consonants.

Copy / Adapt

clear lyric-first delivery, crisp consonants, fewer runs, doubles only on emphasis words

Listen for: Important words should remain understandable at normal listening volume.

The chorus vocal is not bigger

Likely cause: Verse and chorus use the same range and support behavior.

Smallest correction: Change one lift mechanism.

Copy / Adapt

chorus moves slightly higher, wider doubles on title phrase, one upper harmony on sustained words

Listen for: The hook should feel larger without becoming louder everywhere.

Ad-libs are everywhere

Likely cause: The model is filling all negative space.

Smallest correction: Restrict ad-lib placement by section.

Copy / Adapt

ad-libs only after complete lead phrases and mainly in final chorus or outro

Listen for: The main melody should stay readable.

The singer identity changes too much

Likely cause: Too many vocal traits are changing between tests.

Smallest correction: Lock the protected traits.

Copy / Adapt

keep tone, register, phrasing and support behavior stable; change only one vocal variable

Listen for: New generations should vary less in the core identity.

The duet blends into one voice

Likely cause: Both voices share register, texture and timing.

Smallest correction: Assign distinct roles.

Copy / Adapt

two-voice duet, contrasting registers and tones, alternate lines, overlap only in hook harmony

Listen for: Each voice should remain identifiable.

02 · Lyric, Syllable & Phrasing Problems — 10 Problems

The words are rushed

Likely cause: Too many syllables are packed into the available phrase.

Smallest correction: Reduce lyric density or extend the phrase.

Copy / Adapt

fewer syllables per line, clearer rests, longer phrase space, no compressed delivery

Listen for: The singer should stop racing the beat.

Stress lands on the wrong words

Likely cause: Natural speech emphasis conflicts with melodic emphasis.

Smallest correction: Protect key stressed syllables.

Copy / Adapt

emphasize content words, avoid stretching weak syllables, place strong words on strong beats

Listen for: The lyric should sound more natural when spoken and sung.

Lines run together

Likely cause: Phrase endings are not clearly separated.

Smallest correction: Create rests and shorter line shapes.

Copy / Adapt

shorter lyric phrases, clear breath points, deliberate rests between lines

Listen for: Each line should read as a separate thought.

The chorus is too wordy

Likely cause: The hook carries too much information.

Smallest correction: Compress around one anchor phrase.

Copy / Adapt

simpler chorus lyric, repeat title phrase, fewer supporting words, stronger melodic repetition

Listen for: The hook should be easier to remember after one listen.

The verse feels robotic

Likely cause: Line lengths and accents are too uniform.

Smallest correction: Introduce controlled variation.

Copy / Adapt

mix short and medium phrase lengths, conversational contractions, selective offbeat entries

Listen for: The verse should feel spoken through the groove rather than mechanically counted.

The rhyme feels forced

Likely cause: Rhyme placement is driving unnatural wording.

Smallest correction: Protect meaning first.

Copy / Adapt

natural phrasing, fewer end-rhyme obligations, use internal rhyme selectively

Listen for: Lyrics should sound less constructed around rhyme.

The singer mispronounces key words

Likely cause: Dense phrasing or unusual spellings are creating ambiguity.

Smallest correction: Simplify and clarify the lyric surface.

Copy / Adapt

clear standard spelling, isolate important names or terms, slower phrase around difficult words

Listen for: Critical words should become intelligible.

The melody fights the lyric

Likely cause: Syllable shape and melodic contour are mismatched.

Smallest correction: Adjust phrase length before rewriting the entire melody.

Copy / Adapt

match stressed syllables to melodic peaks, shorten crowded lines, preserve core message

Listen for: The lyric should sit inside the melody more comfortably.

The hook phrase is buried

Likely cause: The title or anchor phrase lacks placement priority.

Smallest correction: Give it space and repetition.

Copy / Adapt

isolate the title phrase, repeat it, place it at phrase ending, reduce surrounding lyric density

Listen for: The listener should recognize the hook phrase immediately.

The second verse feels like a copy

Likely cause: Lyric cadence and line lengths mirror verse one too closely.

Smallest correction: Preserve pocket but vary phrasing.

Copy / Adapt

same core pocket, different phrase lengths, one response line, one intentional pause

Listen for: Verse two should advance without sounding disconnected.

03 · Hook & Chorus Problems — 10 Problems

The chorus does not lift

Likely cause: Verse and hook share the same density, register and width.

Smallest correction: Change one or two contrast variables.

Copy / Adapt

raise hook register slightly, widen backing vocals, simplify lyric rhythm

Listen for: The chorus should feel like a destination.

The hook is forgettable

Likely cause: There is no clear repeated anchor.

Smallest correction: Create one memory signal.

Copy / Adapt

repeat one short title phrase, simple melodic contour, consistent cadence

Listen for: You should be able to recall one phrase after the song ends.

The hook arrives too late

Likely cause: Too much setup happens before payoff.

Smallest correction: Move recognition earlier.

Copy / Adapt

short intro, hook tease in opening, first chorus earlier

Listen for: The listener should encounter the central idea sooner.

The hook repeats too much

Likely cause: Every chorus is nearly identical in arrangement.

Smallest correction: Keep the hook but change support.

Copy / Adapt

preserve main hook, vary backing vocals, width or post-hook layer on later repeats

Listen for: Repetition should feel intentional, not static.

The chorus sounds like another verse

Likely cause: Lyric density and phrase behavior do not change.

Smallest correction: Simplify and widen.

Copy / Adapt

fewer words, longer notes, stronger repetition, wider support

Listen for: The section should clearly change function.

The hook sounds too poppy

Likely cause: Melodic lift is overpowering the intended genre.

Smallest correction: Use rhythm or chant instead of bigger melody.

Copy / Adapt

narrow melodic range, stronger rhythmic repetition, call-and-response hook

Listen for: Genre identity should stay intact.

The hook is too busy

Likely cause: Too many counterlines and layers compete with the title.

Smallest correction: Remove support before adding anything.

Copy / Adapt

one lead hook, one backing response, reduce secondary melodies

Listen for: The title should become the focal point.

The final chorus is not stronger

Likely cause: No new payoff is reserved for the end.

Smallest correction: Save one major expansion.

Copy / Adapt

final chorus adds octave double or choir response, wider harmony, one new rhythmic layer

Listen for: The final return should feel earned.

The post-chorus steals focus

Likely cause: Secondary motif is stronger than the main hook.

Smallest correction: Reduce complexity or shorten it.

Copy / Adapt

short post-chorus motif, lower lyric density, keep chorus title dominant

Listen for: The main hook should remain the song’s memory center.

The chorus starts weak

Likely cause: The first beat or phrase does not signal arrival.

Smallest correction: Strengthen entry behavior.

Copy / Adapt

clean downbeat entry, brief pre-hook space, strongest phrase first

Listen for: The chorus should announce itself immediately.

04 · Structure & Arrangement Problems — 10 Problems

The AI ignores section labels

Likely cause: Too many competing structural instructions are present.

Smallest correction: Simplify to the essential form.

Copy / Adapt

Verse 1 → Chorus → Verse 2 → Chorus → Bridge → Final Chorus; preserve section order

Listen for: Structure adherence should improve before adding detail.

Every section sounds the same

Likely cause: No explicit contrast rule exists.

Smallest correction: Define one variable per section.

Copy / Adapt

verse sparse, pre-chorus rising, chorus wide, bridge stripped

Listen for: Sections should become distinguishable.

The bridge feels unrelated

Likely cause: Too many identity traits change at once.

Smallest correction: Keep one anchor stable.

Copy / Adapt

bridge changes harmony but keeps vocal identity and groove pulse

Listen for: The bridge should refresh, not replace, the song.

The bridge kills momentum

Likely cause: Reset lasts too long or removes too much pulse.

Smallest correction: Shorten and rebuild earlier.

Copy / Adapt

short bridge, retain light pulse, begin rebuild halfway through

Listen for: Momentum should dip without collapsing.

The intro wastes time

Likely cause: It has no functional job.

Smallest correction: Establish motif or cut it.

Copy / Adapt

two-bar motif intro, immediate groove, no long ambient setup

Listen for: The intro should justify every second.

The outro drags

Likely cause: A full chorus is repeated after the emotional peak.

Smallest correction: Use a shorter ending.

Copy / Adapt

short hook tag, stripped reprise, clean final button

Listen for: The song should end after the payoff, not after momentum fades.

Verse two is identical

Likely cause: No progression is built into the repeat.

Smallest correction: Add one controlled variation.

Copy / Adapt

verse two adds response vocal or counterline, core groove unchanged

Listen for: The song should progress without resetting.

The song feels too long

Likely cause: Redundant setup or repeat sections are present.

Smallest correction: Cut repetition before cutting identity.

Copy / Adapt

shorter intro, fewer repeated pre-choruses, preserve strongest hook and bridge

Listen for: Length should fall without damaging the core arc.

The drop or chorus comes too early

Likely cause: There is no anticipation.

Smallest correction: Add one concise build mechanism.

Copy / Adapt

short build, rising tension, one-bar reduction before payoff

Listen for: The payoff should feel prepared but not delayed.

The final section feels pasted on

Likely cause: No earlier element prepares it.

Smallest correction: Seed the returning element earlier.

Copy / Adapt

introduce hook motif or harmony cue earlier, expand it only in final section

Listen for: The ending should feel connected to the song’s established language.

05 · Genre Drift Problems — 10 Problems

R&B drifts into Pop

Likely cause: Pocket and harmony looseness are being simplified.

Smallest correction: Reassert R&B relationships.

Copy / Adapt

syncopated pocket, warmer harmony, conversational phrasing, less straight pop chorus rhythm

Listen for: The result should regain R&B center without changing the entire song.

Hip-Hop drifts into Trap

Likely cause: 808 and hat behavior are taking over.

Smallest correction: Reduce trap-specific grammar.

Copy / Adapt

stronger kick-snare pocket, reduced hat density, shorter 808 sustain, sample-inspired harmony

Listen for: The beat should return toward hip-hop center.

Trap drifts into Drill

Likely cause: Sliding bass and cold syncopation dominate.

Smallest correction: Stabilize low-end movement.

Copy / Adapt

less bass sliding, more stable 808 roots, reduce cold syncopated pressure

Listen for: The groove should feel trap-centered again.

Reggae drifts into generic Pop

Likely cause: Skank, bass role and pocket are weakened.

Smallest correction: Protect groove grammar.

Copy / Adapt

one-drop or steppers pocket, skank guitar, organ bubble, bass-led movement

Listen for: The seasonal or pop layer should sit on top of reggae, not replace it.

Country drifts into Pop-Rock

Likely cause: Storytelling space and acoustic identity are reduced.

Smallest correction: Restore country anchors.

Copy / Adapt

acoustic guitar, steel or fiddle accent, conversational vocal, restrained drums

Listen for: The song should read country before production polish.

Gospel drifts into generic inspirational Pop

Likely cause: Call-and-response and church-rooted harmony are missing.

Smallest correction: Restore communal vocal grammar.

Copy / Adapt

piano-organ support, choir response, congregational hook, live-feeling rhythm section

Listen for: The worship/gospel center should become clear again.

EDM drifts into Pop

Likely cause: Drop/build logic is too weak.

Smallest correction: Restore dance structure.

Copy / Adapt

clear build, drop contrast, club-focused low end, reduced lyric density in payoff

Listen for: The dance grammar should lead.

Pop drifts into EDM

Likely cause: Production spectacle overtakes the song.

Smallest correction: Restore lyric-led structure.

Copy / Adapt

strong lead vocal center, chorus melody priority, reduce drop emphasis

Listen for: The song should feel like Pop with electronic production.

Lo-fi drifts into Jazz

Likely cause: Improvisational complexity becomes foreground.

Smallest correction: Simplify movement.

Copy / Adapt

short loop, dusty drums, stable harmony, restrained instrumental variation

Listen for: Texture and repetition should lead again.

Genre fusion sounds confused

Likely cause: Too many genre grammars are equally dominant.

Smallest correction: Choose a primary and secondary role.

Copy / Adapt

primary genre owns groove and structure; secondary genre contributes one texture or harmony trait

Listen for: The listener should know what the song fundamentally is.

06 · Drums, Groove & Pocket Problems — 10 Problems

The hats sound robotic

Likely cause: Subdivision is too constant.

Smallest correction: Use rolls as accents.

Copy / Adapt

simple hat grid, occasional short rolls, clear reset points

Listen for: Movement should feel intentional rather than machine-gunned.

The groove feels stiff

Likely cause: Everything lands too exactly on grid-like accents.

Smallest correction: Introduce pocket variation.

Copy / Adapt

slightly relaxed timing, human-feeling drum pocket, selective syncopation

Listen for: The groove should breathe without losing stability.

The drums are too busy

Likely cause: Too many fills and subdivisions compete with vocals.

Smallest correction: Remove decorative activity.

Copy / Adapt

simpler core groove, fills only at section transitions, fewer percussion layers

Listen for: The vocal should regain space.

The kick feels weak

Likely cause: It lacks role clarity against bass.

Smallest correction: Define transient ownership.

Copy / Adapt

firm kick on selected downbeats, bass handles sustain

Listen for: Impact should improve without low-end clutter.

The snare/clap feels wrong for the genre

Likely cause: Backbeat character conflicts with lane expectations.

Smallest correction: Change the anchor before changing the whole beat.

Copy / Adapt

genre-appropriate snare/clap placement and texture, preserve tempo

Listen for: The groove should read closer to target genre.

The groove has no bounce

Likely cause: Bass and drums are not interacting rhythmically.

Smallest correction: Create syncopated space.

Copy / Adapt

bass and kick alternate accents, intentional rests, stronger pocket

Listen for: The groove should move without adding speed.

Percussion masks the vocal

Likely cause: High-frequency rhythm density is too high.

Smallest correction: Reduce foreground percussion.

Copy / Adapt

fewer bright percussion hits, reserve accents for transitions

Listen for: Lyrics should become easier to follow.

The beat feels too slow

Likely cause: Sustained layers and sparse subdivisions reduce perceived motion.

Smallest correction: Increase rhythmic activity selectively.

Copy / Adapt

more active percussion or bass rhythm, shorter sustained textures

Listen for: Energy should rise without changing tempo.

The beat feels rushed

Likely cause: Too many events happen in each bar.

Smallest correction: Create breathing room.

Copy / Adapt

reduce hat density, longer bass notes, fewer fills

Listen for: The same tempo should feel calmer.

The second verse groove feels stale

Likely cause: The beat repeats with no controlled change.

Smallest correction: Add one variation.

Copy / Adapt

one added percussion layer or drum dropout in verse two, core pocket unchanged

Listen for: Progression should be audible but identity preserved.

07 · Bass, 808 & Low-End Problems — 10 Problems

The 808 swallows the mix

Likely cause: Sustain and overlap are excessive.

Smallest correction: Shorten low-end events.

Copy / Adapt

shorter 808 decay, fewer overlapping notes, leave gaps between phrases

Listen for: Vocals and kick should regain space.

Kick and bass fight each other

Likely cause: Both are trying to own the same transient and sustain.

Smallest correction: Assign separate roles.

Copy / Adapt

kick owns punch, bass owns sustain, reduce simultaneous emphasis

Listen for: Low end should feel clearer and stronger.

The bass feels weak

Likely cause: It lacks consistent role or audible movement.

Smallest correction: Clarify bass behavior.

Copy / Adapt

stronger root support, selective rhythmic movement, stable low-end presence

Listen for: The groove should feel anchored.

The bass is too busy

Likely cause: It follows every melodic or vocal event.

Smallest correction: Simplify the line.

Copy / Adapt

fewer bass notes, longer sustain, respond only to key rhythmic phrases

Listen for: The low end should support instead of narrating everything.

808 slides are excessive

Likely cause: Glides have become the main event.

Smallest correction: Limit them to phrase endings.

Copy / Adapt

stable roots, glide only at selected transitions

Listen for: Slides should regain impact.

Sub bass disappears on small speakers

Likely cause: Too much energy sits only in very low frequencies.

Smallest correction: Add clearer upper-bass definition conceptually.

Copy / Adapt

rounded bass with audible mid-bass presence, preserve sub support

Listen for: The bass role should remain perceptible on more systems.

Low end changes every generation

Likely cause: Bass behavior is under-specified.

Smallest correction: Protect it as a trait.

Copy / Adapt

keep 808 length, root behavior and kick relationship stable across tests

Listen for: Versions should differ less in the low-end identity.

Bass muds the chorus

Likely cause: Hook adds too many sustained low layers.

Smallest correction: Simplify chorus low end.

Copy / Adapt

one primary bass layer, shorter sustain under dense harmony

Listen for: The hook should widen without getting cloudy.

Bass does not match genre

Likely cause: Behavior conflicts with the target pocket.

Smallest correction: Use genre-specific role language.

Copy / Adapt

define bass as walking, sub-sustained, syncopated, one-drop, bouncing or root-supporting

Listen for: The groove should communicate genre more clearly.

Bass entrance has no impact

Likely cause: It is present from the beginning with no contrast.

Smallest correction: Delay or thin the opening low end.

Copy / Adapt

reduced bass in intro, full bass enters at hook or first major section

Listen for: The low-end entrance should create a noticeable lift.

08 · Instrumentation, Density & Balance Problems — 10 Problems

The track is overcrowded

Likely cause: Too many foreground layers compete.

Smallest correction: Choose one focal element per range.

Copy / Adapt

lead vocal foreground, one main motif, reduce secondary midrange layers

Listen for: The hierarchy should become obvious.

The wrong instrument dominates

Likely cause: Foreground priority does not match the song job.

Smallest correction: Reassign its role.

Copy / Adapt

main guitar or synth becomes supportive, lead vocal or hook motif moves forward conceptually

Listen for: Attention should shift to the intended focal point.

Everything is wide

Likely cause: No center/edge contrast remains.

Smallest correction: Narrow key elements.

Copy / Adapt

lead and bass centered, width reserved for pads, harmonies or chorus expansion

Listen for: The mix should feel more dimensional.

The arrangement sounds thin

Likely cause: Too few supportive roles exist.

Smallest correction: Add one support layer, not five.

Copy / Adapt

one pad, harmony bed or counterline added beneath main elements

Listen for: Fullness should improve without clutter.

The chorus gets muddy

Likely cause: Too many new elements enter at once.

Smallest correction: Choose one expansion mechanism.

Copy / Adapt

add width or harmony, not multiple competing instruments

Listen for: The hook should feel larger and cleaner.

The verse has too much going on

Likely cause: No negative space remains for lyrics.

Smallest correction: Strip secondary layers.

Copy / Adapt

remove counterline or pad, keep groove and one motif

Listen for: The verse should feel more intimate and readable.

The instrumental hook fights the vocal hook

Likely cause: Two memorable ideas occupy the same space.

Smallest correction: Separate their timing.

Copy / Adapt

instrument motif answers after vocal phrase instead of playing through it

Listen for: Both hooks should become clearer.

The guitar/piano/synth sounds generic

Likely cause: Instrument role is under-described.

Smallest correction: Specify behavior instead of instrument name only.

Copy / Adapt

short syncopated guitar answers / sparse piano motif / pulsing synth pattern

Listen for: The part should gain a more specific function.

The track lacks movement

Likely cause: Layers enter and exit without purpose.

Smallest correction: Tie changes to section jobs.

Copy / Adapt

verse sparse, pre adds one layer, chorus widens, bridge strips

Listen for: Arrangement movement should support form.

The final section is overloaded

Likely cause: Every available layer has been added.

Smallest correction: Protect the lead hierarchy.

Copy / Adapt

final section adds only one major new layer plus wider support

Listen for: Peak energy should remain intelligible.

09 · Energy, Mood & Emotional Problems — 10 Problems

Everything is the same intensity

Likely cause: No section has a distinct energy job.

Smallest correction: Map low/medium/high before regenerating.

Copy / Adapt

verse low, pre medium, chorus high, bridge reset, final high-plus

Listen for: The song should gain an obvious arc.

The song is too happy

Likely cause: Harmony, tempo feel or timbre is brighter than intended.

Smallest correction: Darken one dimension at a time.

Copy / Adapt

darker chord color, restrained brightness, lower vocal energy, less celebratory percussion

Listen for: Mood should shift without destroying genre.

The song is too dark

Likely cause: Low register and minor-color cues dominate.

Smallest correction: Introduce warmth selectively.

Copy / Adapt

brighter upper harmony, warmer vocal tone, less ominous low texture

Listen for: The result should feel lighter without becoming cheerful if that is not desired.

The track sounds cheesy

Likely cause: Too many obvious genre or mood signals are stacked.

Smallest correction: Remove clichés and keep one cue.

Copy / Adapt

reduce novelty effects, keep one defining texture, rely on groove and harmony

Listen for: The style should feel more credible.

The track feels generic

Likely cause: The prompt names mood but not behavior.

Smallest correction: Specify musical relationships.

Copy / Adapt

define groove, bass role, vocal phrasing, section contrast and one signature texture

Listen for: The song should gain a clearer identity.

The bridge loses emotion

Likely cause: The reset removes the core message.

Smallest correction: Keep one emotional anchor.

Copy / Adapt

bridge strips arrangement but preserves vocal tone or motif

Listen for: Contrast should not erase the song’s emotional center.

The final chorus has no payoff

Likely cause: Peak resources were used too early.

Smallest correction: Reserve one expansion.

Copy / Adapt

save choir, octave, counterline or widest harmony for final section

Listen for: The ending should feel earned.

The song is too cinematic

Likely cause: Large swells and orchestral cues dominate intimacy.

Smallest correction: Reduce scale.

Copy / Adapt

smaller arrangement, closer vocal, fewer impact hits, simpler harmonic motion

Listen for: The song should feel personal rather than trailer-like.

The song is too intimate

Likely cause: It never expands enough for the intended payoff.

Smallest correction: Increase one dimension.

Copy / Adapt

wider chorus support, stronger low end, longer sustained hook notes

Listen for: The song should grow without becoming overproduced.

The emotional tone changes randomly

Likely cause: Prompt traits conflict across sections.

Smallest correction: Protect a single intent.

Copy / Adapt

keep core emotional intent stable; vary intensity, not emotional identity

Listen for: Sections should feel like one song.

10 · Prompt, Consistency & Generation Problems — 10 Problems

The model ignores the most important instruction

Likely cause: Too many priorities compete in the prompt.

Smallest correction: Reduce to one primary correction.

Copy / Adapt

keep core song identity; change only [single priority]

Listen for: The targeted change should become easier to hear.

Every generation changes too much

Likely cause: Too many variables are unconstrained.

Smallest correction: Protect survivor traits.

Copy / Adapt

preserve groove, vocal tone, key motif and structure; change one variable only

Listen for: Versions should become comparable.

The prompt is too long

Likely cause: Priority is buried under descriptive language.

Smallest correction: Compress to role-based instructions.

Copy / Adapt

genre center + groove + vocal behavior + section job + one correction

Listen for: The model should receive a clearer hierarchy.

Negative instructions are not respected

Likely cause: Exclusion wording is fighting positive instructions.

Smallest correction: Replace exclusions with desired alternatives.

Copy / Adapt

use restrained percussion and simple fills instead of constant rolls

Listen for: Positive target behavior should improve compliance.

One fix creates two new problems

Likely cause: The revision changes too many connected variables.

Smallest correction: Return to smallest intervention.

Copy / Adapt

keep all protected traits; revise only bass sustain / vocal phrasing / chorus width

Listen for: Collateral drift should decrease.

The result is better but loses the original magic

Likely cause: The repair removed a survivor trait.

Smallest correction: Name and protect what must stay.

Copy / Adapt

preserve original groove, motif, vocal character and emotional tone; fix only [problem]

Listen for: The repaired version should retain recognizable strengths.

The model keeps adding unwanted elements

Likely cause: Prompt leaves instrumentation roles open.

Smallest correction: Define foreground and support.

Copy / Adapt

lead vocal + one main motif; background layers restrained; no competing counterlines

Listen for: Unexpected clutter should reduce.

Regenerating the whole song makes things worse

Likely cause: The problem is local, not global.

Smallest correction: Stop using global regeneration for a section-level issue.

Copy / Adapt

keep song identity and arrangement; revise only the problem section or behavior

Listen for: More of the survivor should remain intact.

Prompting is no longer solving the issue

Likely cause: The problem belongs to editing, mixing or lyric revision.

Smallest correction: Change tools, not words.

Copy / Adapt

stop regenerating; move to stem edit / mix balance / lyric rewrite / section replacement

Listen for: The workflow should become more efficient.

You do not know when to stop

Likely cause: No success criterion exists.

Smallest correction: Define the repair target before testing.

Copy / Adapt

success = problem reduced, protected traits preserved, no new major issue introduced

Listen for: You should be able to accept a result deliberately instead of endlessly regenerating.

Repair Log — Turn One Fix Into Reusable Evidence

Record these seven fields for every serious repair attempt:

Survivor version: which generation you are protecting.
Audible problem: what you can actually hear.
Protected traits: 3–5 things that must survive.
Single repair variable: the one thing you are changing.
Repair language: the exact copy/adapt direction used.
Result: what improved, what degraded, and any new issue introduced.
Decision: Keep / Revise / Reject, plus the next action.

Why this matters: after several repairs, your notes become a personal evidence base for what your tools, genres and creative preferences respond to reliably.

EDITABLE WORK AREA

Repair Record — Complete One Test

Click or tap inside each light box. Use the Print / Save PDF button above when you want a checkpoint copy.

1. Survivor version

[Version title / link / file name / generation ID]

2. Audible problem

[Describe what you hear without guessing the cause]

3. Protected traits

[List 3–5 traits that must survive]

4. Single repair variable + exact direction

[What are you changing? Paste the repair language used.]

5. A/B result

[What improved? What degraded? Any new problem?]

6. Decision

[Keep / Revise / Reject] — [Next action]

A/B Repair Test — Prove the Fix Helped

A · Survivor: save the current best version and name the problem.

B · Repair: keep protected traits stable and apply one correction only.

Compare: did the target problem improve? Did any protected trait degrade? Did a new major problem appear?

Decision: Keep / Revise / Reject. A repair only wins if it improves the target without creating a larger problem.

When Prompting Is No Longer the Right Fix

Move to lyric editing when…

The issue is syllable count, unnatural stress, confusing wording, weak rhyme logic or an overstuffed chorus.

Move to section replacement when…

One bridge, hook, verse or ending is wrong but the rest of the survivor is strong.

Move to stems / editing when…

The musical idea works but one instrument, balance issue or local arrangement problem needs surgical control.

Move to mixing when…

The parts are right but hierarchy, low-end clarity, width, vocal space or playback translation is the blocker.

Move to deeper BUILD training when…

The same failure keeps recurring across songs and you need a reusable control rule rather than another isolated repair.

Stop when…

The target problem is reduced, protected traits survive, and no new major issue was introduced. “Different” is not automatically “better.”

Completion Task & Success Criteria

Task: choose one flawed generation you still believe is worth saving. Name one audible problem, list 3–5 protected traits, choose one repair from this vault, and run one controlled A/B repair test.

Record: problem, likely cause, repair language, protected traits, what improved, what degraded, and Keep / Revise / Reject.

Success: you can explain what changed and why the repair was accepted or rejected without relying on “I just liked this generation more.”

Go Deeper When the Problem Is Bigger

Vocals: Vocal Direction Prompt Vault.

Structure: Song Structure & Hook Architecture Vault and Advanced Song Structure Engineering.

Lyrics: Lyric Engineering Checker, Syllable Density & Flow Physics, and Stress & Emphasis System.

Mix / production: Advanced Studio Mixing Lab.

Workflow discipline: Production Workflow VIP Manual.

Vault navigation: return to the AI Music Prompt Vault Hub for genre, vocal, lyric, structure, mood and creator-use-case vaults.

If the problem is really direction: return to Find Your Sound before doing more repair work.