Mix tips for mastering

Export file formats

  • Stereo mixes in standard lossless formats (WAV/AIFF/FLAC)
  • Export at the native sample rate of your DAW project (i.e. 48khz project generates a 48khz mix file)
  • Export at 24-bit resolution or greater
  • Disable any mastering limiter on your master bus (see below)
  • Don’t edit, normalize or convert your mix
  • Export your mix or stem from a known start point (e.g. beat 1, bar 1), so if you send additional files later, we’ll be in sync

Mix levels

  • Please mix for a maximum of -1db true peak or less (no red lights)
  • If you are aiming for streaming platforms, check your integrated loudness (LUFS) measurement is no more than -14db LUFS. Hot mixes are okay, but they will be turned down by Spotify and YouTube. We can help you review causes of excessive volume. If you don’t have a loudness meter, check out the free Youlean Loudness Meter
  • Mixes for CD can be hotter, say -10db LUFS (loud), but please observe less-than -0.1db true peak and aim for zero overs. Note that, in general, streaming, CD and vinyl masters generally have similar profiles today.
  • If that’s not doable, we’ll do our best!

DAW mix export checklist

If you’re an engineer, you don’t need to read the next bit (but they’re mistakes we’ve all made when exporting from a DAW)

  • Export at the project sample rate: When exporting from your DAW, make sure you export to the sample rate at which the song is recorded (otherwise Bounce Mix will perform a software sample-rate conversion)
  • Export at 24 or 32-bit, Stereo Interleaved: Export at the native DAW file bit depth, usually 24-bit or greater
  • Don’t add dither to your master bus when exporting 24 bit files: For more information, see Dither, below
  • No red lights please!: Make sure your gain structure is clean throughout your digital sum - you shouldn’t be seeing any red lights in a DAW, an occasional one is not the end of the world, but remember that you don’t hear ‘overs’, because digital overload protection kicks in to prevent audible cracks! This means your transient will be getting splattered! Don’t worry about loudness, keep it well-balanced, clean - dynamic and musical (big contrasts!)
  • Don’t over-compress or limit stems: Modern streaming loudness standards mean that you shouldn’t over-use ‘power’ mixing techniques like parallel compression and limiting, so the emphasis has shifted to natural loudness, and more musical dynamics, rather than continuous, limited stem volume
  • Use monitor calibration, but disable it when exporting: We recommend monitor and room calibration tools. Sonarworks can be particularly valuable to give you confidence in your use of EQ, or when referencing low-end in headphones. But *remember to disable monitor correction when exporting your mix (or bounce from a ‘mix’ bus, rather than your main output, so the problem is completely avoided)
  • Use mix references: use benchmarks for your genre, when gauging timbre, fx, low end, etc. A useful resource is the iZotope Spotify Master Reference Playlist
  • Leave headroom for transients: In 24-bit audio, there is sufficient small-signal resolution to aim for an occasional peak touching -6db. So, if your master bus clips (with any master bus plugins disabled), bring all the channels feeding into your mix down by 6db (e.g. create a fader group), then, turn your monitors up to compensate, and keep an eye on your meters. In doing this, you should find that other sounds from your computer are way too loud - it’s normal to need to turn monitors down by 12db before playing mastered material, because it’s usually at least 12db (4x) louder than the mix level you’re working at in your DAW. This 12db difference in level equates to your mix headroom - so, make sure you play reference mixes at the same volume as your mix (cut them by 12db)

“I’m sorry, Mr Sinatra, can we run through that one again? The “My Way” line seemed to crackle a little in here….” (Anon., Ex-engineer, USA)

Master-bus processing

If you have processing on your master bus -

  • Print a mix with this enabled (your reference mix, something we know that you’re happy with)
  • Also print a mix without master bus compression enabled (do not worry about volume, or your mix sounding quiet)
  • Remember to disable any monitor correction software, e.g. Sonarworks.

We add the equivalent processing within an analogue chain, as we need to consider peak limiting carefully - we might use automation to reduce ‘outlier’ peaks from a track, and then apply a more gentle analogue limiting or compression to soften peaks more consistently.

Dither

Should you apply dither to your master bus when exporting files? In general, no, because you should never be sending us 16 bit exports. But, check your DAW’s manual before applying a dither plug-in to your master bus. If you print files by recording buses to audio tracks, and grab files from your Audio Files folder, those files have already been dithered (else they would be full of all kinds of nasty artefacts!) A useful read is the Pro Tools Reference Guide, ch 53, below - the concept applies to all DAW software, though some use 32-bit summing internally.

When to Use a Dither Plug-In

You should use a dither plug-in in any situation where you are reducing bit depth, for example, when mixing down to a 16-bit file with the Bounce Mix command, or when sending your final mix to an external digital device that records at 16-bit resolution.

A dither plug-in is necessary even when mixing down to 16-bit from a 16-bit session. Even though 16-bit sessions use 16-bit files, they are still processed at higher internal bit depths. Pro Tools systems use 24-bit audio input and output signal paths, and internal 64-bit floating point processing for mixing and audio processing.

For this reason, whether you are working with a 32-bit floating-point session, or a 24-bit session or a 16-bit fixed-point session, the following are recommended:

  • When mixing down to a 16-bit destination,use a dither plug-in on the main output
  • When bus recording a submix to a track in the same session, do not use a dither plug-in on the destination track
  • When mixing down to a 24-bit destination, do not use a dither plug-in on the main output
  • When mixing down to an analog destination with any 24-bit capable interface, do not use a dither plug-in on the main output
  • During normal recording and playback, bypass any dither plug-in on the main output

Bounce Mix and Dither

Bouncing to disk does not apply dither when reducing bit depth. To apply dither when bouncing to disk, you should insert a dither plug-in as the last processor in the signal path on a Master Fader assigned to the Mix Source audio output path.

The golden rule is to check your DAW manual, then choose an export depth where you don’t need to think about dither.

Top-and-tail mix noise

  • Listen to starts and ends of tracks for hiss, and other noise. Use mute / fader automation to keep intro sections noise free, but also narrow noise down to its source. Occasionally plug-ins contribute noise into the mix buss, even when their channel contains no audio. This is particularly the case with Waves hardware emulations - try turning off ‘Analogue’, or reduce it to a minimal level
  • Also check noise in middle sections and breaks!
  • If your main ‘intro’ section channel is noisy (e.g. a vocal mic with a lot of gain), don’t trim the start of your mix - if you can leave a period of noise with no other signal, we can grab a noise print and the results can be impressive. Other modes of noise reduction are not ideal
  • Don’t edit the tail of your mix, let instruments ring out if they sound natural, but use automation to fade each track before noise or clicks become apparent - the last thing ringing is often the bass, cymbals, piano
  • If the mix tail has a long fade, it might cross-fade into the next song, so don’t commit early to a particular fade length, it may be too abrupt. Fades have to make sense with regard to the song coming afterwards - track gaps are determined by the mood of song changes

Stem mastering checklist

  • Print stems as contiguous, unbroken files, from your project start (e.g. beat 1, bar 1), to the end of the audio. So, stems might be four stereo files, all the same length
  • Assemble the stems in a new DAW session, and check they play back as expected (especially, missing instruments)

Analogue summing from stems assumes you have performed mix-level automation on printed stems, i.e. set flat, your reference mix should emerge. Because stems are exported via a mix process, vocal stems should contain all vocal edits, de-essing and pitch correction

What kind of stems do we need?

Stems are stereo WAV or AIFF files, containing the elements of your mix, e.g. drums, bass, guitars, vocal, backing vocals, fx. A minimal mastering stem set might be -

  • music (stereo)
  • vocal (mono / stereo, no fx)
  • vocal fx (stereo)

This allows us to re-balance the vocal level quickly, and perform additional processing in a much less invasive manner, e.g. vinyl de-essing can be done unobtrusively, by applying automation to the dry vocal stem, where necessary.