coherence: share of signal without distortion and noise
Dynamics (DR)
9.6 dB
(average)
Spectral similarity
77.0 %
(medium)
Analysis reliability
High
Recording qualityGood
Phone AGCCan't tell
Clipping0.0 %
Synchronization91 %
FR (frequency response)
Reference (original)Your recording
Deviation (Your recording − Reference)
Key findings
Tonal balance noticeably off
Dynamic range preserved
Treble roll-off 4 – 20 kHz (up to −115.0 dB)
Strong bass excess 160 – 315 Hz (up to +14.0 dB)
Strong bass excess 63 – 160 Hz (up to +11.7 dB)
Bands (deviation from the original)
Range
Deviation
Rating
Sub 40–60 Hz
−12.3 dB
strong dip
Mid-bass 60–250 Hz
+6.5 dB
strong excess
Mids 250–4 kHz
+1.8 dB
slightly higher
Highs 4k–18 kHz
−79.9 dB
strong dip
Additional metrics
Band similarity30.2 %
Clarity (1–5 kHz)20.1 %
SPL (loudness contour)60.6 %
Dynamic similarity89.1 %
Signal cleanness (coherence)48.1 %
Stereo stage width–
Overall score489 / 1000
DSP tuning recommendations
High: 4 – 20 kHz
Add brightness if needed:
PEQ 8 kHz, +58…59 dB, Q = 0.5 – 0.7
Mid: 160 – 315 Hz
Reduce the peak:
PEQ 250 Hz, −6…11 dB, Q = 1.1 – 1.7
Low: 63 – 160 Hz
Lower the level:
PEQ 100 Hz, −5…10 dB, Q = 1.0 – 1.5
Sub: below 63 Hz
Raise the level:
LSF 63 Hz, +6…7 dB, Q = 0.7
Check the crossovers and phase at 1.6 – 2 kHz. Make sure mids and highs sum correctly; adjust the delays if needed.
A dip at 400 – 630 Hz is often a phase issue at the mid-bass/midrange crossover: try inverting the mid-bass polarity or adjusting the delays.
A narrow deep dip is usually caused by acoustic cancellation (reflections, phase). EQ can't fully fill it, so don't boost more than +3…4 dB.
Summary
The sound differs substantially from the original. Check speaker wiring and polarity, crossover and EQ settings, then measure again.
Recommended correction4 PEQ
Detailed report
Synchronization
Position in the original0:00.250
Track start in the recording1.40 s
Compared segment0:43
Correlation0.230
Peak sharpness (PSR)81.1
Clock drift+9.1 ppm
Confidenceshare of confirmed segments, weighted by PSR91 %
Recording integrity
Clippingshare of samples at full scale0.00 %
Noise floorestimated from quiet passages−19.5 dB
SNRneeds a 1–2 s pause before the track starts–
Phone AGCthe track has few loudness changesCan't tell
Level alignmentLUFS + FR median+9.0 dB
Recording file20261110_132252.amr
Loudness metrics
Metric
Original
Recording
LUFSperceived loudness
−11.3
−16.6
Peakmax sample level
−0.3 dB
−5.1 dB
True Peakpeak after reconstruction
0.1 dBTP
−5.0 dBTP
ST-LUFS maxloudest 3 s window
−11.2
−16.1
RMSaverage power
−14.5 dB
−18.9 dB
Crest factorpeak / average
14.2 dB
13.8 dB
LRAloudness range
0.1 dB
0.6 dB
DRdynamic range
11.0 dB
9.6 dB
Similarity scores
Spectralcloseness of frequency balance77.0 %
By bandsmatch in the main ranges30.2 %
Claritypresence bands 1–5 kHz20.1 %
DynamicLRA and crest factor89.1 %
SPLloudness contour over time60.6 %
Signal cleannesscoherence 200 Hz–5 kHz48.1 %
Overall scoreweighted combination489 / 1000
What you may hear
Sub 40–60 Hz: −12.3 dB · significant, match 13% The bass lacks depth and “pressure”.
Mid-bass 60–250 Hz: +6.5 dB · significant, match 34% Mid-bass may sound thick or “boxy”.
Mids 250–4 kHz: +1.8 dB · small, match 74% Close to the original.
Highs 4k–18 kHz: −79.9 dB · significant, match 0% Cymbals and “air” lack sparkle; less detail.
Stereo analysis
The recording is mono, so stereo analysis is unavailable. This is normal for most voice recorders.
1/3-octave detail
Excellent · |Δ| < 1 dBSmall · 1–3 dBNoticeable · 3–6 dBSignificant · > 6 dBOutside the analysis range
Frequency
Region
Δ
25 Hz
Infra / very deep sub
−31.5 dB
32 Hz
Deep sub extension
−24.2 dB
40 Hz
Deep sub body
−14.9 dB
50 Hz
Sub foundation
−9.7 dB
63 Hz
Upper sub / boom
−0.5 dB
80 Hz
Lower mid-bass, warmth
+5.3 dB
100 Hz
Mid-bass weight
+10.1 dB
125 Hz
Main mid-bass punch
+10.6 dB
160 Hz
Upper mid-bass body
−0.0 dB
200 Hz
Transition to lower mids
+13.5 dB
250 Hz
Low vocals / “boxiness”
+9.4 dB
315 Hz
Male vocal body
+3.1 dB
400 Hz
Lower presence
−2.1 dB
500 Hz
Vocal / guitar body
−4.1 dB
630 Hz
Speech intelligibility
−0.9 dB
800 Hz
Upper vocal body
+9.0 dB
1 kHz
Vocal presence
+7.6 dB
1.25 kHz
Presence / attack
+2.7 dB
1.6 kHz
Edge of clarity
−2.9 dB
2 kHz
Presence peak
−3.2 dB
2.5 kHz
Consonant clarity
+1.0 dB
3.15 kHz
Hard presence
+2.0 dB
4 kHz
Attack and clarity
−8.6 dB
5 kHz
Definition / start of sibilance
−48.8 dB
6.3 kHz
Brilliance
−93.9 dB
8 kHz
Air / cymbal shimmer
−86.8 dB
10 kHz
Air and fine detail
−104.9 dB
12 kHz
Ultra air
−114.6 dB
16 kHz
Top-end edge / often the mic limit
−102.0 dB
20 kHz
Ultrasonic edge / often limited
−109.2 dB
A relative comparison of a phone recording after synchronization and level alignment, not a laboratory frequency response measurement. The phone microphone and its processing affect the result, so the analysis covers only 40 Hz – 18 kHz. EQ recommendations are a starting point: verify changes by ear and with a repeat measurement.
Tonal balance up to 300 points
What it is
How well the levels of four frequency bands — sub (40–60 Hz), mid-bass (60–250 Hz), mids (250 Hz–4 kHz) and highs (4–18 kHz) — match the original. Based on the average deviation in each band: the closer to 0 dB, the more points.
What it affects
The overall character of the sound. Too much sub and mid-bass means boom and muddiness, and the music loses detail. Too little means a thin sound without body. Unbalanced mids make the sound nasal or hollow; unbalanced highs make it harsh and fatiguing, or dull.
How to fix
First balance the amplifier or DSP channel levels: sub, mid-bass, mids and highs relative to each other.
Then apply the PEQ recommendations from the report; they are calculated from exactly these deviations.
Turn off Loudness, Bass Boost, EQ presets and “enhancers” on the head unit.
Excess sub-bass is often caused by phase rather than level: try switching phase 0/180° and adjusting the delay.
Clarity up to 200 points
What it is
Frequency response accuracy in the 1–5 kHz presence region. This is where vocal intelligibility, guitar and snare attack, and the “definition” of the sound live.
What it affects
Vocal intelligibility and the sense of detail. Dips make the voice muffled and coming “from your feet”, and the stage collapses. Peaks make the sound shouty, “in your face”, and fatiguing at volume.
How to fix
Check the midrange/tweeter crossover: frequency, slope and tweeter polarity.
Adjust the channel time delays: dips in this region are often caused by out-of-phase summing.
Don't boost narrow deep dips by more than 3–4 dB with EQ; better shift the crossover frequency or change the phase.
Try changing the tweeter aiming: towards the listener or off-axis.
FR shape up to 150 points
What it is
How similar the shape of the whole frequency response curve is to the original: correlation of levels across 27 one-third-octave bands from 40 Hz to 16 kHz. Overall loudness does not affect this score, only the shape matters.
What it affects
The overall tonal tilt: whether the sound is “bright” or “dark” and whether the big picture matches. This score is usually high; a low value points to a serious problem such as a dead speaker, a disabled channel or a major crossover mistake.
How to fix
If the score is below 75 %, make sure all speakers are playing and all amplifier channels are on.
Check the crossover settings: make sure LPF/HPF filters are not mixed up between channels.
Then fix the tilt: if the sound is dark, raise the top end with a shelf (HSF); if too bright, lower it.
Fix small irregularities after the big ones, following the PEQ recommendations.
Dynamics up to 150 points
What it is
Whether the dynamics of the original are preserved. Two parameters are compared: LRA, the difference between loud and quiet parts of the track, and crest factor, how far peaks (hits, attacks) rise above the average level.
What it affects
How “alive” the sound is: drum punch, attack, contrast between quiet and loud passages. Compressed dynamics make the sound flat and fatiguing even with a flat frequency response.
How to fix
Turn off compressors, limiters, DSP effects and speed-dependent volume on the head unit.
Don't drive the amplifier into clipping: if the sound “compresses” at volume, lower the gain and check the power supply.
Record with a voice recorder app without automatic gain control (AGC) and noise suppression — check the “Analysis reliability” block.
A strong bass boost also lowers this score, so fix the tonal balance first.
Loudness contour up to 100 points
What it is
How closely the recording's loudness over time follows the original: short-term loudness (0.4 s blocks) compared across the whole track.
What it affects
Shows whether the system reproduces loudness changes consistently. A low value means the loudness drifts regardless of the music: phone or head-unit automation is at work, noise interferes, or panels rattle.
How to fix
Turn off automatic volume control in the head unit (SVC/AVC, speed-dependent volume).
Record in silence: engine and climate control off, windows and doors closed.
Use a recorder mode without AGC and do not move the phone while recording.
If something rattles at volume, the doors and panels need sound deadening.
Signal cleanness up to 100 points
What it is
Coherence is the share of the recorded sound that is clearly related to the original (200 Hz–5 kHz range). Everything else is noise, distortion, rattles and cabin reflections.
What it affects
Transparency and “cleanness” of the sound, no muddiness at volume. In a car cabin 50–70 % is a normal result: some loss is unavoidable due to reflections from glass and panels.
How to fix
If the sound gets “dirty” at volume, lower the amplifier gains and source volume, and check the speakers for distortion.
Eliminate rattles: door sound deadening, securing trim panels and wiring.
Eliminate background noise while recording: engine, climate control, street.
Soft materials and proper speaker installation reduce reflections and raise this score.