READY
🇺🇸
English
English
🇪🇸
Spanish
Español
🇫🇷
French
Français
🇧🇷
Portuguese
Português
🇨🇳
Chinese
中文
🇯🇵
Japanese
日本語
🇩🇪
German
Deutsch
🇳🇬
Yoruba
Yorùbá
🇳🇬
Pidgin
Naijá Pidgin
🇸🇦
Arabic
العربية
🇰🇷
Korean
한국어
🇮🇹
Italian
Italiano
🇰🇪
Swahili
Kiswahili
P
![]()
Profile Name
🎯 Best mix levels for Auto-Analyze
• Peak level: −6 to −3 dBFS
• Integrated LUFS: −14 to −18 LUFS
• True Peak: −3 dBTP minimum
• Remove limiters/maximizers from master bus before exporting
• Peak level: −6 to −3 dBFS
• Integrated LUFS: −14 to −18 LUFS
• True Peak: −3 dBTP minimum
• Remove limiters/maximizers from master bus before exporting
AI Track Guide
AI tracks output at −8 to −10 LUFS — already hot. Use AI Track for gentle Red Book polish. Light EQ, minimal compression, target −14 LUFS.
AI tracks output at −8 to −10 LUFS — already hot. Use AI Track for gentle Red Book polish. Light EQ, minimal compression, target −14 LUFS.
Navigation
Compare
Multiband Spectrum
LIVE
LOW
MID
HIGH
🎵
Drop your audio file here
WAV · MP3 · FLAC · AIFF — no install needed
📂 Tap to browse your files
WAVEFORM
0:00
No file loaded
0:00 / 0:00
FADE IN 0.0s
FADE OUT 0.0s
Gain & Limiting
Makeup Gain
+0.0 dB
Limiter Ceiling
−1.0 dBTP
Real-Time LUFS
LIVE
Integrated
— LUFS
True Peak
— dBTP
Loudness Analysis
Integrated
— LUFS
True Peak
— dBTP
Dynamic Rng
— LU
Spectrum Analyzer
AUTO
5-Band Parametric EQ
LIVE
Band Energy Analyzer
LIVE
80Hz
250Hz
1kHz
5kHz
12kHz
Low Shelf
0.0 dB
80 Hz
Low Mid
0.0 dB
250 Hz
Mid
0.0 dB
1 kHz
Upper Mid
0.0 dB
5 kHz
High Shelf
0.0 dB
12 kHz
0:00
/
0:00
Level Meter
LIVE
Compression
LIVE
Threshold
−18 dB
Ratio
2:1
Attack
10 ms
Release
120 ms
💡 Gain & Limiting controls are in the Studio tab
Clarity Tools
LIVE
Noise Gate
Silences hiss & noise below threshold
Threshold
−50 dB
Release
100 ms
Rumble Filter
Cuts sub-noise that compressors amplify
Cutoff Frequency
30 Hz
Parallel Compression
Blend dry + compressed for punch without noise
DRY
WET
50% wet blend
Auto Gain Compensation
Smart makeup gain — no noise floor boost
When ON: calculates exact gain needed to hit your platform target without over-amplification. Keeps noise floor clean.
Calculated gain: —
Calculated gain: —
Compression Character
Dithering
Dithering prevents quantization noise when exporting to 16-bit. Recommended for all 16-bit exports.
Multiband Spectrum Analyzer
LIVE
LOW 20–250Hz
MID 250Hz–4kHz
HIGH 4k–20kHz
0:00
/
0:00
Multiband Compressor
LIVE
Low
20–250Hz
−18 dB
Mid
250Hz–4kHz
−18 dB
High
4kHz–20kHz
−18 dB
Ratio
2:1
Ratio
2:1
Ratio
2:1
De-Esser
Frequency
7.5 kHz
Threshold
−20 dB
Amount
6 dB
Targets harsh sibilance
Stereo Width
Width
100%
Phase Correlation MONO SAFE
−1 Phase Cancel0+1 Mono
+1.00
History
No history
Reference Track Analyzer
Noise Floor Monitor
LIVE
QUIETNOISE FLOORLOUD
−∞ dB
No signal
Drop a reference track here
Compare your master against a pro release
REFERENCE SPECTRUM
REFERENCE TRACK
0:00
/
0:00
YOUR MIX SPECTRUM
0:00
/
0:00
Overlay Comparison
— Your Mix
— Reference
Your Mix
—
Reference
—
Mastering Chain
Click any node to jump to its controls
🎵
INPUT
›
〰
EQ
›
✦
EXCITER
›
⊞
M-BAND
›
◈
M/S EQ
›
⚡
TRANS
›
🔴
LIMITER
›
⬇
OUTPUT
Soft Clipper
Rounds peaks before the limiter using a smooth sigmoid curve — adds 1-2 dB of perceived loudness with zero pumping artifacts. Used on virtually every commercial master.
Drive
Gentle
TransparentGentleWarmPunchyHeavy
Harmonic Exciter
LIVE
Adds warmth & analog character
Drive
3
Tone
4.0 kHz
Mix
20%
Mode: Tape — warm 2nd harmonic saturation
Mid / Side EQ
LIVE
MID = center/mono · SIDE = stereo edges. Affects only that channel.
Low
0 dB
80Hz
Lo-Mid
0 dB
250Hz
Mid
0 dB
1kHz
Hi-Mid
0 dB
5kHz
High
0 dB
12kHz
Low
0 dB
80Hz
Lo-Mid
0 dB
250Hz
Mid
0 dB
1kHz
Hi-Mid
0 dB
5kHz
High
0 dB
12kHz
Transient Shaper
LIVE
Attack
0
Sustain
0
Negative attack = softer. Positive attack = punchier.
True Brickwall Limiter
LIVE
Hard ceiling — zero inter-sample peaks
Ceiling
−0.3 dBTP
Lookahead
2 ms
Release
50 ms
GR: 0.0 dB
Peak: — dBTP
Clips: 0
AI Recommendations
Based on your mix analysis
Load and analyze a track to get processing suggestions.
Batch Mastering
Drop multiple audio files. Current settings will be applied to all.
📂
Drop multiple audio files here
WAV · MP3 · FLAC · AIFF — applies current settings to all
Processing...
0%
Export Settings
Bit Depth
Sample Rate
Normalize to platform
Filename suffix
Queue
0 filesMastering Reports
🤖 AI Mastering Report
Analyze a track then click Generate to get a full AI assessment.
📊 Loudness Graph
StartLoad a file to see loudness over timeEnd
— Your Mix
- - Target
📈 Before / After Summary
📋 Session History
No sessions recorded yet.
🎙 VOICE CLEANUP
Clean up podcast, voiceover and vocal recordings — remove background noise, hum, rumble and hiss.
🎙
Drop your voice recording here
Podcast · Voiceover · Vocals — WAV, MP3, M4A, FLAC
0:00 / 0:00
1 · Learn the Noise
Capture a sample of the background noise (a quiet moment with no talking). Oni learns its fingerprint, then removes it from the whole recording — even while you speak.
No profile captured yet
2 · Cleanup Tools
🔇 Noise Reduction
Strength60%
Removes constant background noise (fan, AC, hiss) using the learned profile.
⚡ De-Hum
Removes electrical mains hum from lights, cables and outlets.
🌊 Rumble Filter
Low Cut85 Hz
Cuts low rumble — traffic, AC, mic handling, plosives.
🐍 De-Ess
Amount40%
Tames harsh "sss" sounds and sibilance in vocals.
✨ Voice Enhance
Presence / Clarity3 dB
Adds presence and intelligibility — makes voices clear and upfront.
Quick Presets
After cleaning, switch to the Studio tab to master your cleaned vocal, then Download.
Theme
Switch between dark and light mode
Help & Reference
⭐ Quick Start — Read This First
1
Load your audio file
Drag and drop or tap the Studio drop zone
2
Choose your settings
Auto-Analyze, AI Track, or adjust manually
3
Tap Render
This actually processes the audio
4
Compare A / B
Press play, switch A↔B to hear the difference
5
Download your master
Tap the yellow Download button
💡 Think of it this way:
Auto-Analyze / AI Track = choosing your recipe
Render = actually cooking the food
Download = putting it on the plate
Auto-Analyze / AI Track = choosing your recipe
Render = actually cooking the food
Download = putting it on the plate
What each button does
✦ Auto-Analyze
AI sets everything for you
Sends your track measurements to the AI. Claude analyzes your LUFS, frequency balance and dynamic range, then automatically adjusts all EQ, compression and gain sliders to the best values for your specific track. The AI explains every decision in the chat.
⚠️ Use when: You want AI to handle everything or need a professional starting point fast.
Requires API key.
⚠️ Use when: You want AI to handle everything or need a professional starting point fast.
Requires API key.
🎵 AI Track
For AI-generated tracks only
A preset specifically tuned for tracks from Suno, Udio or other AI music generators. These tracks arrive already hot at −8 to −10 LUFS — they don't need heavy processing. This applies gentle EQ (cuts mud at 250Hz, adds air at 12kHz), very light compression, and reduces gain to hit −14 LUFS Red Book standard.
⚠️ Use when: Your track was made by Suno or an AI music tool. Do NOT use on live recordings — it's too subtle.
No API key needed.
⚠️ Use when: Your track was made by Suno or an AI music tool. Do NOT use on live recordings — it's too subtle.
No API key needed.
⚙️ Render
The most important button
This is what actually processes your audio. It runs your entire file through the chain: EQ → Compression → Gain → Limiting → Normalization. Creates a new mastered version in memory and updates all the loudness meters.
After pressing it: press Play and switch A (original) ↔ B (mastered) to hear the difference.
⚠️ Always press this after adjusting any settings — nothing is applied to the audio until you do.
No API key needed.
After pressing it: press Play and switch A (original) ↔ B (mastered) to hear the difference.
⚠️ Always press this after adjusting any settings — nothing is applied to the audio until you do.
No API key needed.
Mastering Chain Guide
🎵
1. Input / Studio
Load your mix. Analyze LUFS, true peak, dynamic range. Use Auto-Analyze for AI mastering.
›
〰
2. EQ & Dynamics
Shape tone with 5-band EQ. Set compression for glue and punch. Set limiter ceiling.
›
⊞
3. Dynamics
Multiband compression for low/mid/high. De-esser for sibilance. Stereo width and phase.
›
✦
4. Processing
Harmonic exciter for warmth. Mid/Side EQ. Transient shaping. Brickwall limiter.
›
🎧
5. Reference
Compare your master against a professional release. Match loudness and frequency balance.
›
⚙️
6. Render
Render your master. Compare A (original) vs B (mastered) in real time.
›
⬇
7. Export
Choose platform target. Export as 16-bit or 32-bit WAV. Batch export multiple files.
✓
Keyboard Shortcuts
Space
Play / Pause
Dbl-click ▶
Restart from beginning
Ctrl + Z
Undo last change
Ctrl + Y
Redo
Dbl-click slider
Reset to default
Click waveform
Seek to position
Drag waveform
Scrub audio
Hold ⏮ / ⏭
Fast rewind / forward
Pro Tips
AI generated tracks: These arrive at −8 to −10 LUFS — already hot. Use the 🎵 AI Track preset for gentle Red Book mastering. Avoid heavy compression.
Reference matching: Load a pro track on the Reference page. Match your spectrum and LUFS to theirs — this is the fastest way to a competitive master.
Mid/Side EQ: Cut low frequencies on the SIDE channel to tighten bass. Add air (12kHz+) on the SIDE to widen the top end without affecting the mono center.
Loudness for streaming: Target −14 LUFS for Spotify/YouTube/Tidal, −16 LUFS for Apple Music. Going louder just gets turned down by the platform — wasted effort.
Phase correlation: Keep the meter above 0 for mono-compatible masters. Below 0 means phase cancellation — sounds great in stereo but disappears on mono speakers.
About
Oni
AI Mastering Studio · Built with Web Audio API + Claude AI
All processing runs locally in your browser. No uploads. No data collection.
Stem Mastering
Load your drum, vocal and instrument stems separately. Adjust each volume and EQ, then master them together as one final track.
🥁
Drums
No file loaded
VOL
0 dB
📂 Drop or tap to load drums stem
🎤
Vocals
No file loaded
VOL
0 dB
📂 Drop or tap to load vocals stem
🎹
Instruments
No file loaded
VOL
0 dB
📂 Drop or tap to load instruments stem
Stem Mix Preview
0:00
/
0:00
MASTER VOL
0 dB
Mix Level
"Send Mix to Master Chain" merges your stems and loads them into the Studio page with all your EQ, compression and limiting applied.
Settings
🔑API Keys
Anthropic API Key (AI Chat & Mastering)
Get yours free at console.anthropic.com
OpenAI API Key (Human Voice / TTS)
Get yours at platform.openai.com — ~$5 for thousands of responses
🎙️Voice
AI Voice
Speak AI responses aloud
Voice Speed
Normal (1.0x)
Voice Volume
100%
Select Voice
Nova
Warm · Female
Alloy
Clear · Female
Echo ⭐
Warm · Male
Fable
British · Male
Onyx
Deep · Male
Shimmer
Gentle · Female
🌐Language
🇺🇸
English
English
🇪🇸
Spanish
Español
🇫🇷
French
Français
🇧🇷
Portuguese
Português
🇨🇳
Chinese
中文
🇯🇵
Japanese
日本語
🇩🇪
German
Deutsch
🇳🇬
Yoruba
Yorùbá
🇳🇬
Pidgin
Naijá
🇸🇦
Arabic
العربية
🇰🇷
Korean
한국어
🇮🇹
Italian
Italiano
🇰🇪
Swahili
Kiswahili
🎨Appearance
Theme
Dark or light interface
Zoom Level
Choose display size
⚙️Default Settings
Default Platform
Target LUFS on export
Default Bit Depth
Export quality
Default Sample Rate
Auto-Analyze on Load
Analyze file when dropped
💾Data & Privacy
🎵Project Summary
Tell the AI about your project. It will use this context in every mastering conversation.
Project Name
Project Description
Genre & Style
Target Platforms
👤About You
Personalize your AI engineer. The more it knows about you, the better it can assist.
Your Role
Experience Level
About Me (optional — AI uses this to personalize)
Oni
AI Mastering Studio · v2.0
All processing runs locally in your browser.
No uploads. No data collection.
No uploads. No data collection.
Spectrum Analyzer
Spectrum is shown on the Studio tab above the drop zone.
Band Energy
P
![]()
Profile Name
Factory Presets
Export Options
Bit Depth
Sample Rate
Normalize
My Presets
No presets saved.
Save Current
Platform Levels
Spotify −14
YouTube −14
Apple −16
Before / After
Before
—
— dBTP
After
—
— dBTP
Suggested Presets
Projects
Slot A
—
—
Slot B
—
—
AI ENGINEER
AI
Hi! Drop an audio file to start. Use 🎵 AI Track or ✦ Auto-Analyze for AI-generated tracks for full mastering.
SPEAKING
🎤 Listening...
🔑 API keys needed
AI ENGINEER
More Pages