Aumixys VERSION 2.0

Batch (Album)

Login

🎵 Aumixys (Single Track)

Upload your finished track to run a detailed audio analysis and check if it's ready for distribution on platforms like Spotify or Apple Music. The analyzer generates a comprehensive report evaluating loudness (LUFS), dynamics, spectrum profile, phase correlation, anomalies (AI), and provides optimization suggestions (Auto-QC).

🛡️
Login required: Aumixys measures Integrated LUFS from the full file, not from a 30-second sample. Log in to run a full analysis and avoid misleading results.
Log in / join free
📊 Full LUFS & TruePeak Report 🤖 AI Recommendations & Auto-QC 🕵️ AI Detection (Anomalies) 🎧 Platform Simulator (Spotify)
🎵

Drop audio file here

or click to select

WAV · MP3 · FLAC · OGG · AIFF · M4A · AAC (max 150 MB)

Not ready to upload yet? Open a demo report and see the kind of conclusions you get after analysis.
Open live demo
No AI training Your file is processed in real-time and automatically deleted after analysis. No audio is stored or used for AI training.
Badge: saves your CPU All analysis is performed on our powerful servers, saving your local CPU and memory for your creative work.

Professional Audio Analysis and Streaming Compliance

Aumixys Analyzer is a cloud-based precision tool designed for deep audio diagnostics before distribution. We use algorithms compliant with ITU-R BS.1770-4 and EBU R 128 standards, ensuring industry-standard measurement accuracy.

Why is LUFS analysis vital? Platforms like Spotify, YouTube, and Apple Music use loudness normalization. If your master is too loud, it will be turned down, potentially damaging your track's perceived dynamics. Our tool checks:

  • Integrated LUFS: Overall perceived loudness.
  • True Peak (dBTP): Inter-sample peak measurement to avoid distortion during lossy encoding.
  • LRA (Loudness Range): Statistical dynamic range distribution.
📊
LUFS & Loudness
Integrated, Short-Term, Momentary — which meter matters most?
Read article →
📈
True Peak & Safety
Why can Spotify distort your master after transcoding?
Read article →
🥁
Kick & Bass (Low-End)
Masking, phase & sidechain — how to tame your low end?
Read article →
✅ Ready for distribution
0 errors 0 warnings 0 info
Auto-Fix Beta

FIX AUDIO

Choose a mastering preset and Aumixys prepares a Before/After A/B preview. Free gets preview without download; Normal and Premium can download the fixed file within the Audio Fix limit.

File —
Limit —
Choose a mastering preset and run the fix.

What Auto-Fix checks and improves

🛡️ Signal Integrity Report

Phase
—
—
: —
Spectrum
—
—
Probability of generative processes (AI): —%
—
Integrated LUFS ?
—
LUFS
True Peak (analog ISP) ?
—
dBTP
Sample Peak (digital DAW)
—
dBFS
Dynamic Range ?
—
dB
Loudness Range ?
—
LU
Key
—
Tempo
—
BPM
Low-End Beta ?
—
Risk
Vocal Fit Beta ?
—
Vocal placement
Stereo Correlation ?
—
Clip Count
—
clips
📊 Integrated LUFS LUFS
Overall, averaged loudness of the entire recording.
✅ Ideal: −16 to −9 LUFS Streaming target: −14 LUFS
📈 True Peak dBTP
Highest signal peak accounting for analog conversion.
✅ Safe: below −1.0 dBTP ⚠️ Dangerous: above 0 dBTP
⚡ Dynamic Range dB
Difference between the loudest and quietest fragments.
✅ Good music: 8–14 dB ⚠️ Blocked: below 5 dB
🎚️ Loudness Range (LRA) LU
Measures loudness fluctuations over time — how much the music 'breathes'.
✅ Pop/EDM music: 3–8 LU
🎵 Tempo BPM
Recording speed expressed in beats per minute.
ℹ️ Higher BPM = faster rhythm
🎹 Key
Musical key of the recording — the 'scale' around which the melody revolves.
E.g. C-major, A-minor, F#-major
🔊 Stereo Correlation
How similar the left and right channels are to each other.
✅ Safe: 0.3 to 1.0 ⚠️ Risky: below 0
✂️ Clipping instances
Number of instances where the signal exceeded the maximum digital level.
✅ Ideal: 0 instances ⚠️ Problem: above 10 instances

🎯 Streaming Platform Compliance

📊 What is LUFS?

LUFS (Loudness Units Full Scale) is a unit measuring subjective loudness as perceived by the human ear — not the signal peak, but the actual sensation of loudness over time. Streaming platforms normalize music to approx. −14 LUFS.

Short-term LUFS (3s window) — loudness measured in 3-second windows. Shows fluctuations over time — a chorus is usually louder than a verse.
Momentary LUFS (400ms window) — instantaneous measurement every 400 milliseconds. The most sensitive indicator — reacts to a single drum hit.
Integrated LUFS — a single collective average for the entire file. This is the value Spotify compares with its −14 LUFS standard.
💡 Tip: A streaming-friendly track usually sits around −14 to −12 LUFS Integrated. If your track is, for example, −8 LUFS, Spotify will turn it down automatically and it may lose impact.

📏 The loudness histogram shows how often the recording stayed at each loudness level. One tall bar = little dynamic variation. A wide spread = the music breathes.

📐 Crest Factor is the difference between peak and RMS loudness. High values (15+ dB) = natural punch. Low values (below 6 dB) = heavily compressed audio.

Loudness over time (LUFS)

Loudness Histogram (LUFS distribution)

Crest Factor over time

📈 What is True Peak and Peak Level?

Peak Level is the basic peak of the signal amplitude. True Peak (dBTP) is an advanced measurement that checks what peaks will appear after converting the file to MP3 or AAC. True Peak is responsible for distortion on platforms.

Peak Level — momentary signal peak measured in short time windows (50 ms) on the timeline. It shows where the track has the highest spikes.
Sample Peak (digital, DAW) — highest PCM sample value (dBFS), i.e. what a typical DAW peak meter shows. It does not include analog reconstruction between samples.
True Peak (analog ISP) — peak after oversampling (inter-sample peak), closer to what can occur after D/A reconstruction or lossy encoding. If it exceeds 0 dBTP, clipping risk increases even when DAW Sample Peak still looks safe.
True Peak − Sample Peak is the ISP margin. The larger it is, the bigger the gap between analog and digital peak readings.
💡 Practical tip: Watch both meters together: Sample Peak (as in DAW) and True Peak (post-reconstruction). For safer distribution, set limiter ceiling to −1.0 dBTP or lower.

True Peak over time (dBTP)

⚠️ Clipping positions

🌊 What is Waveform?

Waveform is a visual representation of the recording over time. It shows how the signal's loudness changed from beginning to end.

Positive peak (Max) — the highest signal point in each window. The higher it is, the louder it sounds.
RMS (energy average) — the actual signal power. A good indicator of perceived density and fullness.
Negative peak (Min) — the lowest point. A symmetric waveform usually means healthy audio. Strong asymmetry may indicate DC offset.
💡 Fade In is a gradual increase in loudness at the beginning. Fade Out is a fade at the end. The analyzer detects both automatically and shows their duration in seconds.

🔎 Lossy compression detection checks whether a WAV or FLAC file was actually converted from MP3. MP3 files often have a characteristic high-frequency cutoff above 16 kHz, still visible even after conversion to WAV.

Waveform overview

🔉 Fade In / Fade Out detection

🔎 Lossy compression detection

🌈 🌈 What is Frequency Spectrum?

The spectrum is an 'X-ray' of sound — it breaks down music into individual frequencies. X-axis is frequencies (bass to treble), Y-axis is loudness in dB.

Sub (0–60 Hz) — subwoofers, kick drum. Minimally audible on small speakers, but consume a lot of energy.
Bass (60–250 Hz) — bass guitar, kick drum, vocal body. The foundation of the sound.
Low-Mid (250–500 Hz) — guitar warmth, snarl. Too much = muddy/dull sound.
Mid (500 Hz–2 kHz) — vocal clarity, instrument presence. Key to mix intelligibility.
High-Mid (2–6 kHz) — sharpness, guitar presence. Too much = ear fatigue.
High (6–20 kHz) — "air", cymbals, brilliance. Too little = dull. Too much = sibilant.

Average Frequency Spectrum

Tonal profile of the entire recording. Ideal spectrum for pop/EDM is relatively flat or slopes slightly to the right (more bass than treble).

Spectrum Balance — band division

Percentage share of each band in the total energy of the recording. E.g. bass above 30% = excess of low frequencies.

🎙️ Vocal Fit Beta

A heuristic analysis of whether the vocal region in the finished mix is clear, stable and anchored in the center. The module checks Mid/Side, body, presence, sibilance, masking and changes over time.

Center — stability of Mid energy across the vocal-related range.
Clarity and body — relationship between low-mids, intelligibility and presence.
Masking and sibilance — competition risk around 1-5 kHz and excess energy in the upper vocal region.
Important: This is not vocal separation. Snare, guitar, lead synth and other centered sources can affect the result. Always confirm the readout by listening.

Vocal-region fit over time

Fit score plus momentary masking and sibilance risk. Gaps indicate insufficient vocal-region activity.

Vocal Fit Beta readout

—

— — —
How to read the percentages: Center, Clarity, Body and Stability behave like quality scores: higher is usually better. Sibilance and Masking show risk: lower is better there. The dB value underneath is an internal band relationship that explains where the percentage comes from.
Center — — Checks whether the vocal-like energy is anchored in the Mid channel, the center of the stereo image. High = vocal/lead usually sits solidly in the center; low = it may drift sideways or disappear in the mix.
Clarity — — Relationship between intelligibility/presence bands and the surrounding mix. High = words and the main melody are more likely to cut through; low = check guitars, synths, snare and reverbs.
Body — — Balance of vocal body, roughly low-mid weight, against presence. High = the voice has weight and warmth; very low = it may feel thin; too much body can become muddy.
Sibilance — — Risk of sharp S, T and SH sounds in the upper vocal area. This is a risk metric: lower is better. High = check de-essing, EQ and bright reverbs.
Masking — — Risk that instruments and Side energy compete with the vocal in voice-related bands. This is also a risk metric: lower is better. High = check whether the vocal hides behind the arrangement.
Stability — — Checks whether the level and balance of the vocal region stay consistent over the track. IQR is the spread over time: a smaller spread usually means a more stable vocal.

What to check in the mix

—

🥁 Kick/Bass Beta

Heuristic analysis of kick and bass interaction in the finished master. Aumixys points to the harmonic low-end component, transient low-end component, overlap area, and a starting clue for dynamic EQ/sidechain. Overlap itself is not a problem; timing, alternation, and listening context matter too.

Harmonic low-end — approximation of the bass part or tonal low-end foundation.
Transient low-end — approximation of kick impulses and short low-end hits.
Overlap area — shared kick/bass energy. This is a candidate to check, not an automatic mix problem.
💡 Important: This is not stem separation. Small or controlled kick/bass overlap can be normal. Aumixys should point to moments worth checking in the DAW, not treat every overlap as a mistake.

Low-End Conflict Beta

Comparison of harmonic low-end, transient low-end, and overlap area. The red area means “check this”, not automatically “wrong”.

Kick/Bass readout

Risk score —
Status —

Technical suggestion

🔥 What is Heatmap?

A spectrogram is a three-dimensional representation of sound. It combines time (X), frequency (Y) and loudness (color).

Dark blue — none or very low level of the given frequency at this point
Blue / Cyan — moderate frequency level
Yellow — high level — dominant frequency
Red — very high level — may indicate an excess of a given band
💡 How to read the Heatmap?
• Long horizontal bright bands = frequency dominates throughout the track (e.g. steady bass or pad)
• Vertical bright columns = loud, transparent moments (e.g. drum, chord)
• Dark top of the chart = lack of treble above 16 kHz — classic sign of transcoding from MP3

Spectrogram — heatmap

🔊 🔊 What is Stereo Analysis?

Stereo analysis checks the safety of the mix on various devices. It checks phase correlations and Mid/Side balance.

L/R Correlation — scale from −1 to +1. Value +1 = identical channels (mono). Value 0 = wide stereo. Value below 0 = phase issues — something disappears on a mono speaker.
Stereo Width — how "vast" the sound space is. 0 = narrow/mono, 1 = maximum wide. Typical pop music: 0.3–0.6.
Mid — sounds common to L and R (center stage: vocal, kick, bass). Usually 60–80% energy.
Side — sounds differing L and R (effects, reverb, widening). Too much = risk of disappearing in mono.
Goniometer (stereo phase) — Each dot = one audio sample. Vertical ellipse shape = healthy stereo. Horizontal shape = antiphase!

Stereo correlation L/R over time

Values close to +1 = in-phase, safe mono. Negative values = channels cancel each other out.

Stereo width over time

How the spatial sweep of sound changes. Choruses often have higher width than verses.

Mid/Side Balance

Ratio of center energy (Mid) to sides (Side). Typical pop mastering: approx. 70% Mid / 30% Side.

Goniometer (stereo phase)

Each dot = one audio sample. Vertical ellipse shape = healthy stereo. Horizontal shape = antiphase!

⚡ What is Dynamics Analysis?

Dynamics is the difference between the loudest and quietest fragments. Music with high dynamics 'lives' — it has its breath and tension.

Peak Level — highest signal point in a short time window (50 ms). Shows momentary amplitude peaks — how high the "spikes" of the waveform are.
RMS Level — Root Mean Square — energy average of the signal, closer to what the ear hears. If Peak and RMS are very close throughout the track = heavily "blocked" music, without natural breaths.
DR (Dynamic Range) — amount of dB between RMS average and peak. DR 14+ = very good dynamics. DR 6-10 = typical pop mastering. DR below 5 = heavily compressed.
💡 Peak vs RMS differences — if they are large (above 12dB), it means a natural recording with transients. If close to zero — a brickwall limiter has crushed your audio.

⚡ Peak, True Peak & RMS

🎹 What is Key and Chroma?

Key is the 'musical scale' around which the melody is built. Crucial for DJs and harmonic mixing.

Major — major scale. Sounds bright, happy. Typical for pop, dance, anthems.
Minor — minor scale. Sounds serious, melancholic, dramatic. Typical for trap, dark EDM, ballads.
Confidence (%) — how certain the algorithm is. Below 50% = atonal recording, without a clear key (e.g. ambient, noise).
💡 Chroma Distribution — a chart of the 12 musical scale notes (C, C#, D, D#, E, F, F#, G, G#, A, A#, B). The note with the highest bar is the dominant note, and the algorithm fits a matching key (major or minor) to it.

Detected Key

—
—
—

Chroma Distribution (tonal profile)

Each bar = one note. A higher bar means the note is more "present" in the recording. The pattern is compared with major and minor profiles for each of the 12 notes.

🎵 What is Tempo (BPM)?

Tempo is the speed of the recording's rhythm, expressed in beats per minute. The algorithm tracks tempo throughout the track.

40–80 BPM — slow ballads, slow hip-hop, ambient music
80–130 BPM — pop, R&B, typical rap, nu-disco, reggaeton
128–145 BPM — house, techno, trance, progressive
160–180 BPM — drum&bass, jungle, hardstyle, speedcore
💡 Algorithmic note: The detector may sometimes show half or double the actual BPM (e.g. 75 instead of 150 BPM). This "half/double tempo" phenomenon is a normal artifact and it's worth checking the result manually. Stable line = electronics from a DAW. Jumping line = live recording or rubato tempo.

🎵 Tempo over time

🛡️ Signal Integrity Report

Technical analysis checking "cleanliness" and origin of the audio signal for phase errors and unnatural spectral artifacts. Translates technical data into specific conclusions.

=== SIGNAL INTEGRITY REPORT ===

[PHASE]: —

Loading data...

[SPECTRUM]: —

Loading data...

[STATISTICAL ANALYSIS]:
Probability of generative processes (AI):
0%
WARNING: This report is a technical analysis and is not definitive proof. Full interpretation requires human verification.

🔍 Quality Control Report (Auto-QC)

🤖 AI Mastering Recommendations

Our advanced AI analyzed your track against professional standards such as LUFS, True Peak, and spectral balance. Below you will find personalized suggestions to help you achieve release-ready, high-quality sound.

📋 Recommendation list

🎧 Platform Simulator

Normalization to -14 LUFS (Spotify/YouTube).

0.0 dB

✅ Before distribution

🎧 Intelligent comparative listening (Gain Matching)

Compare versions without the volume "cheat". Choose a normalization level to reliably assess tone and dynamics. What is louder always seems better — level the volumes to know the truth.

A: 0.0 dB
B: 0.0 dB
Normalization (EBU R128):
Original

No changes. You hear the files exactly as they were exported from your DAW.

Streaming (-14 LUFS)

Spotify/YouTube standard. Check how the track sounds after leveling to platform loudness.

Analiza (-18 LUFS)

Reference level. Allows your ears to rest and better catch details in timbre and space.

Match B to A

Fair comparison. Levels Master loudness to the Mix, to eliminate the "louder = better" illusion.

🛡️ Signal Integrity Report

Spectrum
—
—
—

🔀 Comparison Table

Parameter File A File B Difference

📊 Spectrum Comparison

The chart overlays the frequency profiles of both tracks. Pay close attention to the curve divergence. Conclusions you can draw: Where does the Master (File B) differ from the Mix (File A)? If curve B rises sharply in the 6-20kHz region, a lot of "air" and harshness was added. If curve B is lower than A in the bass regions, it means the limiter squashed the low-end. Look for unnatural bumps that weren't present in the original.

📦 Album Analysis

File LUFS True Peak S.Peak DR LRA Key BPM Time Loudness Status
LUFS

Normalized loudness level (standard is -14 LUFS).

True Peak

Peak value accounting for inter-sample peaks during DA conversion.

DR (Dynamic Range)

Dynamic Range — difference between peaks and average loudness.

LRA (Loudness Range)

Loudness Range — variation of loudness over the track duration.