Automatic Audio Mastering (Match a Reference Track)

About the Automatic Audio Mastering (Match a Reference Track)

Choose a finished mix — WAV, FLAC, AIFF, MP3 or M4A — and, if you like, a reference track whose sound you want to get close to. The page measures your mix’s loudness, range, peaks and tone, then masters it the way a mastering chain does: an EQ that moves its tone towards the reference (or towards a starting curve when you have none), gentle three-band compression, optional stereo width with the bass kept in mono, and a true-peak limiter that brings the song to the loudness you choose — −14 LUFS for streaming, −16, −23 LUFS for broadcast (EBU R 128), or −11 LUFS for a loud master.

Every setting is heard at once: the loudest 20 seconds play as mix and master at the same loudness, so you judge the tone and the punch, not just the extra level, and a chart shows your mix’s tone, the target and the EQ. Everything runs in your browser on your own device; the songs are never uploaded. It is MySmartCoPilot’s own signal processing — no AI model — and an alternative to the AI mastering services you pay for by the song or by subscription.

The free preview on this page is the first half of the master (up to 30 seconds) with a short MySmartCoPilot chime every 10 seconds, and the page tells you the full master’s length, loudness and true peak. With a Pro pass, or after unlocking this one master, you get the whole file.

How it works

• Tone. The long-term average spectrum of your mix (and of the reference) is measured in bands a twelfth of an octave apart, leaving out quiet passages. The EQ is the difference to the target after both are levelled over 100 Hz–8 kHz, smoothed over half an octave so it follows the overall balance rather than single notes, held within ±6 dB (boosts of at most 2 dB below 40 Hz and 3 dB above 12 kHz) and scaled by the tone match. It is applied as a linear-phase filter, so the tone changes without smearing the timing.
• Dynamics. The song is split at 150 Hz and 2.5 kHz with Linkwitz–Riley crossovers that add back flat. Each band has a soft-knee compressor whose threshold is set from that band’s own peak levels in your song, with a gentle ratio (1.5:1 light, 2:1 medium, 2.8:1 strong) and slower timing for the bass than for the treble; the level the band has most of the time is given back.
• Width. The side signal (the difference between left and right) can be made a fifth wider or narrower, and below 120 Hz it is removed, so the bass stays centred and works on mono speakers and in clubs.
• Loudness. The processed song is measured with the ITU-R BS.1770 method, raised or lowered to the target, and held under the ceiling by a look-ahead true-peak limiter; trial runs correct the level until the result is within 0.3 LU of the target, and when the ceiling keeps a loud target out of reach, the report says how close it came. 16-bit files get TPDF dither.

Which loudness to choose

• −14 LUFS is where Spotify adjusts tracks to by default, with true peaks below −1 dB; it also asks louder masters to keep peaks below −2 dB, which the automatic ceiling does.
• −23 LUFS with peaks under −1 dBTP is the EBU R 128 level for broadcast programmes.
• −16 LUFS is a quieter master with more room for peaks, for spoken word and dynamic music.
• −11 LUFS is loud: the limiter has to work harder, and the report says how hard.

Getting a good master

• Master from the best file you have: a WAV or FLAC mix, not an MP3 of it.
• Pick a reference that is close to your song in genre and arrangement; the tone match moves the overall balance, so a very different song pulls the EQ the wrong way.
• If the report says the limiter works hard, choose a quieter target or stronger dynamics, or fix the loudest moments in the mix.
• Listen to the master on headphones and on speakers before you release it.

Limitations

• Automatic mastering is a starting point, not a mastering engineer: it cannot fix a muddy mix, a clashing arrangement or distortion that is already in the file. Listen on headphones and speakers before you release.
• The tone match follows the overall balance of the reference, not its individual instruments.
• Up to 60 minutes on a computer and 15 minutes on a phone, at up to 96 kHz; the original sample rate is kept (MP3 is written at 44.1 or 48 kHz). Phones take noticeably longer.
• Files with a picture are mastered as sound only; 5.1 mixes are mixed down to stereo.

Privacy

On this device; its files download once

Everything happens in your browser. What you enter or open here is not uploaded or stored by MySmartCoPilot. The audio engine (and, when your file needs them, the MP3 or AAC decoder and the MP3 encoder) downloads from MySmartCoPilot the first time you open a file. Your songs and reference tracks are never uploaded.

Open-source components

This page can load these open-source components, each only when a task needs it:

• It uses FFmpeg under the LGPLv2.1 and the LAME MP3 encoder under the LGPL, through libav.js: licence and source code.
• It uses FFmpeg under the LGPLv2.1, through the Mediabunny add-ons: licence and source code.
• It uses the LAME MP3 encoder under the LGPL, through the Mediabunny MP3 add-on: licence and source code.
• It uses Mediabunny under the MPL 2.0: licence and source code.

You may replace them with your own builds. Their licences and the credits of every library the site uses are on the open-source credits page.