Sonar

Case studies

Before and after: what the recommendations actually change

Four worked examples of a report finding, the specific change it prompted, and the score movement that followed — with every dimension shown, not just the ones that improved.

How to read these

Each study pairs one report finding with one round of revisions and re-analysis. The score tables show all six dimensions so you can see the trade-offs, including where a fix moved one number a lot and left the rest flat.

These are composite examples, not published customer records. The revisions and the before/after scores come from running both versions through the same analyzer; artist identities and customer audio are never shared. Treat the numbers as representative of the pattern, not as a promise about your track.

The studies

Bedroom pop single

Four-minute single with a 22-second ambient intro

Strong song, flat performance. Listeners were leaving before the first vocal entry, so the record never accumulated qualified plays.

What the report said

Hook strength scored 58 with the note that the central melodic idea did not appear until 0:47. Streaming readiness scored 61, flagging the intro against the 30-second play threshold.

What changed

  • Cut the ambient intro from 22 seconds to 6
  • Cold-opened on a two-bar instrumental statement of the chorus melody
  • Moved the first vocal entry from 0:23 to 0:08

Before

62

First hook at 0:47

Two thirds of the 30-second play window elapses before the song starts.

After

78

First hook at 0:04

The idea is stated before the listener has decided anything.

Score comparison before and after the recommended changes for Bedroom pop single
DimensionBeforeAfterChange
Hook5881+23
Mix clarity7172+1
Vocal balance6970+1
Arrangement6077+17
Streaming readiness6184+23
Commercial appeal6472+8
Overall6278+16

Nothing was rewritten. The only change was where the existing material was placed, and it moved two scores by more than 20 points.

Hip-hop feature record

Dense 808-driven beat with a guest verse in the second half

The record sounded huge on studio monitors and disappeared on phone speakers — the lead vocal was inaudible on the playback system most listeners actually use.

What the report said

Vocal balance scored 54 with the note that the lead sat roughly 4 dB below the instrumental in the 1–4 kHz band, where speech intelligibility lives. Mix clarity scored 63, citing masking between the 808 and the vocal low-mids.

What changed

  • Sidechained a narrow 200–400 Hz dip on the 808 to the lead vocal
  • Added 2.5 dB of presence at 3 kHz on the lead rather than raising the whole fader
  • Pulled the ad-libs down 3 dB so they supported rather than competed
  • Checked the result in mono at phone-speaker volume before re-exporting

Before

64

Lyrics unreadable on a phone speaker

Masking in the low-mids buried consonants after codec compression.

After

79

Lead intelligible in mono at low volume

Space was carved for the vocal instead of volume being added to it.

Score comparison before and after the recommended changes for Hip-hop feature record
DimensionBeforeAfterChange
Hook7476+2
Mix clarity6380+17
Vocal balance5483+29
Arrangement7071+1
Streaming readiness6678+12
Commercial appeal6877+9
Overall6479+15

Vocal balance problems are usually mix-clarity problems. Subtracting from the element that masks beats adding to the element that is masked.

Indie rock master

Self-mastered at -6.5 LUFS with the ceiling at 0.0 dBFS

The master measured clean in the DAW but had audible crackle on the loudest choruses once it was uploaded for distribution.

What the report said

Streaming readiness scored 47, flagging inter-sample peaks above 0 dBTP and integrated loudness far past the normalization target — meaning the record would be turned down on playback while keeping every limiting artifact.

What changed

  • Re-mastered to -9.5 LUFS integrated instead of -6.5
  • Set the limiter ceiling to -1.0 dBTP with true-peak detection enabled
  • Reduced average limiter gain reduction from 6 dB to 2 dB and let the mix breathe
  • Verified the transcoded AAC preview before delivering to the distributor

Before

58

Distortion on every chorus after transcode

Loudness the platform removes; artifacts it does not.

After

76

Clean transcode, larger perceived dynamics

Loudness-matched against the reference, the quieter master sounds bigger.

Score comparison before and after the recommended changes for Indie rock master
DimensionBeforeAfterChange
Hook7273+1
Mix clarity5578+23
Vocal balance6170+9
Arrangement6870+2
Streaming readiness4789+42
Commercial appeal6371+8
Overall5876+18

Mastering louder than the normalization target trades transients for nothing. Headroom is the cheapest score improvement available.

Dance-pop record

Three and a half minutes, identical first and second chorus

The first chorus landed; the second one lost people. Full-listen rate fell off sharply around the two-minute mark.

What the report said

Arrangement scored 56, noting that choruses two and three were bit-identical to chorus one and that energy accumulated with no removal anywhere in the record.

What changed

  • Stripped the second chorus to vocal, bass and one percussion layer for its first four bars
  • Added an octave-up vocal double and a counter-melody only in the final chorus
  • Cut eight bars from the bridge so the last chorus arrives earlier
  • Rewrote the final line of the last chorus so the outro pays off the full listen

Before

66

Drop-off concentrated at 2:05

Three identical choruses teach the listener there is nothing left to hear.

After

81

Contrast at 2:05, peak moved to 2:48

Energy is removed before it is added, so the last chorus reads as the biggest.

Score comparison before and after the recommended changes for Dance-pop record
DimensionBeforeAfterChange
Hook7984+5
Mix clarity7476+2
Vocal balance7274+2
Arrangement5685+29
Streaming readiness7079+9
Commercial appeal6681+15
Overall6681+15

Perceived lift comes from contrast, not mass. Subtracting from the second chorus is what makes the third one land.

The pattern across all four

Three of the four gains came from removing something — an intro, a masking frequency, limiter gain, a layer in the second chorus. Only one required writing anything new. The scores that move most reliably are the ones tied to structure and delivery, because those are the parts of a record that fail for measurable reasons.

Common questions

Are these real customer tracks?
They are composite examples built from the recommendation patterns Sonar issues most often, with the before/after scores produced by running each version through the analyzer. Artist identities are not published and no customer audio is shared.
Will my score move this much?
It depends on how much of the problem was structural. Placement, headroom and arrangement-contrast fixes move scores the most because they are what the model measures most directly. A songwriting problem will not be fixed by an edit.
Do higher scores guarantee more streams?
No. Scoring measures whether a record is technically and structurally ready to compete. Marketing, timing and audience still decide the outcome — the score removes the avoidable reasons a good song fails.
How do I compare my own before and after?
Analyze the current version, apply the recommendations, then analyze the revision. Both reports are saved to your history so you can compare each dimension side by side, and shared workspace members see them too.

Run your own before and after

Analyze a trackRead the scoring guides