ZI-NARASHI is a browser-based audio normalizer designed for game developers and video creators. Simply drag & drop audio files to perform LUFS analysis, optimization, and export all in one place.
No installation or account registration required. All processing is done in the browser, so audio data is never sent to a server.
Note: This is a simple normalizer that runs in the browser. While convenient, its precision and features differ from DAW plugins or professional tools. For professional production, we recommend using it alongside dedicated tools.
MP3, WAV, OGG, FLAC, M4A, AAC, WebM — Supports all audio formats the browser can decode
Integrated LUFS measurement using K-weighting filter compliant with ITU-R BS.1770. Also calculates EBU R128 compliant Short-term (3s window) / Momentary (400ms window) LUFS.
Graph display of Momentary/Short-term LUFS along the time axis. Easy comparison with target values at a glance.
Optimization by purpose: Game SE (-12 LUFS), Game BGM (-18 LUFS), YouTube (-14 LUFS), Podcast (-16 LUFS), Spotify (-14 LUFS), etc.
4 modes: LUFS / Peak / RMS / Manual. Auto gain calculation based on peak or RMS values in addition to LUFS.
Adjustable parameters: Threshold, Ratio, Attack, Release, Makeup Gain, Knee. Ideal for dynamics control of SE material.
2-band EQ with Low Shelf (200Hz) and High Shelf (8kHz). Corrects muddy lows and harsh highs.
Suppresses True Peak below the specified Ceiling value. Essential for clipping prevention.
Auto-detects DC offset during analysis. Removed with a 5Hz High-Pass filter.
Apply 0-5 second linear fade. Useful for creating BGM loop material.
Downmix stereo audio to mono. Mono is generally recommended for game SE.
Drop multiple files at once. Batch export all files with the same settings.
Switch between pre/post-processed audio in real-time. Includes real-time loudness meter.
Inter-sample peak value of the audio signal. The closer to 0 dBFS, the higher the risk of clipping. Generally -1 dBFS or below is recommended.
Root Mean Square. Average volume level of the entire audio. Unlike LUFS, it does not consider human hearing characteristics.
Average loudness of the entire audio. Calculated with K-weighting (high-frequency emphasis + low-frequency cut) based on ITU-R BS.1770. Major platform standards:
EBU R128 compliant. Short-term uses a 3s window, Momentary uses a 400ms window. Momentary LUFS is effective for volume evaluation of short SE material.
Difference between Peak and RMS values. An indicator of dynamic range. 6-15 dB is typical for music. Below 4 dB indicates excessive compression, above 20 dB is typical for SE material.
5-band (Sub-80Hz / Low-250Hz / Mid-2kHz / Presence-8kHz / High 8kHz+) RMS energy distribution. Measured by bandpass filter separation.
Audio is processed in the following order:
IN → DC Removal → Gain → EQ → Compressor → Makeup → Fade → Limiter → OUT
All processing uses offline rendering with Web Audio API OfflineAudioContext. Since it is not real-time processing, accurate results are obtained even for long audio.
Automatically optimizes the following parameters based on analysis results:
This tool is provided for the purpose of assisting with audio file analysis and processing. Please note the following.
Supports the latest versions of the following modern browsers:
Requires a browser that supports Web Audio API, OfflineAudioContext, and MediaRecorder API.