A Mastering Engineer’s Guide to Final Mixdown

“Garbage in, garbage out” is a common saying among mastering engineers. The quality of the source material limits the quality of the final product. Most of my clients have no problem following my simple preparation instructions, but they stop there. They figure once each mix sounds as good as they can get it, they’re done. In fact, there’s a higher level of refinement that pays huge dividends. I’ll break it down in this mastering engineer’s guide to final mixdown.

1. Choose a reference. Find a major label track with the tonal balance you’re looking for – ideally something that hasn’t been totally decimated in mastering, since you’ll be comparing it to your unmastered tracks. If you followed my earlier advice on using a reference during the mixing process, you’ll want to use the same track here.

2. Load in your tracks. I’m assuming that you’ve already rendered all the tracks for your release as stereo 24-bit or higher .wav or AIFF files, with no processing on the master bus, and that they peak under 0 dB. If they hit 0 dB, that means they’re clipped. Lower the gain on the master bus by 6 dB and try again. Once you’ve got clean mixdowns, fire up your DAW and put each of them and your reference on separate channels, like so:

Final Mixdown: Load-In
My reference track and six mixdowns

3. Trim each track down to a representative clip. We’re going to use the loudest section of each track as a stand-in for the entire mix. In most cases, this means trimming all but 15 seconds or so of the chorus. Be sure to solo each channel before you hit play so you don’t blow your ears out! In fact, turn down your reference track by 12 dB right off the bat, since it’s already mastered. You’ll end up with something like this:

Final Mixdown: Trimmed
A representative 15 seconds of each track (note the timeline at the bottom)

4. Match volumes. Bounce between the reference and your mixdowns, adjusting volume levels until everything matches. Be sure to make the gain adjustment at the clip level, not on the channel, so you won’t lose your settings when we…

5. Line up all the clips onto a single channel. Alternate between reference and mixdown, like so:

Final Mixdown: The Lineup
Reference, mixdown, reference, mixdown…

6. A/B compare your mixdowns and reference. Hit play and close your eyes. How does each mixdown sound immediately after the reference track? Bright? Dull? Muddy? Boomy? Take plenty of notes, and keep fine-tuning the volume of your clips.

7. Back to the drawing board. Use your notes to make adjustments to your mixes. Import your reference track into each of your projects if you haven’t already. If a mix is too bright, an easy fix is to lower the hi-hats. If a mix is too bassy, ensure the low end is rolled off of non-bass instruments and/or turn down the kick and bass (I detail the process here). If a mix is too muddy, look at cutting some 200-400 Hz, raising some rolloff frequencies, or thinning out the arrangement. Don’t forget to scan the entire track for consistency – not just the chorus.

8. Render, remix, repeat. Open back up our project, shuffle the clips around, take more notes, and keep adjusting your mixes. When they sound consistent, remove the internal reference track clips and shuffle them some more. Eventually you can eliminate the reference track completely. I can’t stress enough how important it is to take frequent breaks! Continue to fine tune your mixes until they match to the best of your abilities, preferably over the course of several days.

Final Mixdown: Continuous Mix
Reference, mixdown, mixdown, mixdown…

You may be wondering, “Did I just master my album?” No, but you made your mastering engineer’s job a lot easier (easier still if you passed along the final volume levels of each of your clips). You minimized the amount of EQ your ME needs to use to create a consistent tonal balance, which means less phase coloration. It means that instead of correcting problems in your mixes, your ME can focus on finding the density and punch that best serves your music on a broad range of playback systems. It means no nasty surprises when you hear your mastered release for the first time, because your ME didn’t have to cut 10 dB off the highs to tame that hi-hat you couldn’t get enough of.

It means better sound, and ultimately, better sales.

Photo by Adi Goldstein on Unsplash

32 Comments

  1. Effective and helpful article. I will forward this to all those b developing artists and engineers that seem to ask me about this process weekly.
    Love and happiness,

    Charley Burch

  2. I like the idea of a reference track, it's really good for getting a feel of what kind of sound you want to get, and also what kind of levels you will be needing.

    At the same time, some tracks need to be mixed down by themselves, especially if you're mixing down tracks that are in really niched genres.

  3. I suppose that depends on what you're referencing. I'd advise having some sort of reference, even if it's in a totally different genre, for the overall tonal balance – and especially the amount of sub bass.

  4. I agree, I do find though that having a reference track from a different genre can completely ruin the whole balance of the track. Out of context the track may sound well rounded, but if you put it back into context of genre and compare it to the greats, it may sound flatter or warmer.

    I find that with a lot of Dubstep and EDM (My genre of choice) there is a huge dip around the mids. A lot of sub bass and a lot of treble are present, to accentuate the whole 'clarity of electronic music thing'

    Whereas with more rock-y kind of stuff, or alternative. There seems to be more mids and that adds to the whole warmth of the sounds.

  5. well, I am learning these mix….for my own albam, i play guitar, bass, kbd, drums(seq), i don't have 4 hands or….is this possible? by the way I have done the mastering (!), but now ill start from the beginning, wish me luck.lmao.

  6. anymore I use my analyzer and set it several different ways. You can do that to check some 10 or 15 songs you like and just remember or LOG the info of each and what you liked about them. Then you KNOW what the db level of each and every Freq is and is not. I use that along with a very high end parametric that can limit and or compress/expand individual freqs. This is NOT for the beginner for sure but after all these years on doing it, it really is spot on to the enth degree. The analyzer I use can expand a view as small as one bar or the whole offering at a glance. You gave very good advise and I recommend viewers add you to their arsenal of info they can count on. ROCK ON!

  7. Interesting approach John! Personally, I'm not a fan of "spectral matching" because the spectra will vary according to the arrangement. As an extreme example, matching the spectrum of a guitar/vocal acoustic track to a rock track sounds TERRIBLE. I'd argue that even matching two rock songs by the same band in the same key is problematic for the same reasons, but of course to a lesser degree. In my 15 years of mastering full-time, I haven't found any technical solutions that beat out just using my ears.

  8. I'm not into spectral matching. My comments are around the DISPLAY RESULTS. Say you have something a bit TUBBY but you've looked at all the tracks. When combined and run thru an analyzer, one finds that a cumulative freq has been compounded via more than two tracks together created a boomyness that can't be found individually. A synergistic result. You use the analyzer to DISPLAY that one narrow freq that is summing freqs together. You THEN NOTCH OUT that freq with a NARROW band covering ONLY that tiny area of the boomyness. THe arrangement has NOTHING to do with what is sounding dominate. Say it is red lining FREQ wise at 300hz at =4db. You just use the parametric to NOTCH OUT a negative 300hz @ _4db. DONE DEAL. Boomy B gone. Not spectral matching at all. In spectral matching, you need something to MATCH TO. THIS is not that. This is just "Hey look what is too loud freq wise" and you just notch that out. Only comparing the track synergy to itself. Not some spectrum from some other song. Once you get the hang of it your tracks become so clear and clean you wonder "I'm just doing this now?".

  9. Use your ears AND an analyzer and you will fine tune your ears to become magic magic. I've only been mastering for some 35 years so I DO have a way to go yet i"m sure. Once you start using a parametric and an analyzer to fine tune YOUR EARS…your pride goes out the window and your confidence goes stellar.

  10. I believe each song is it's own. Each track MUST be a solid song standing alone. When combined one needs to pull back on frequencies that double or triple in certain areas. AN Analyzer, no matter HOW GOOD YOUR EARS ARE can answer for things that you can't hear when your ears get tired.Ears get tired after 2 hours. I don't care who you are. Your ears will compress sound after 2 hours. Fact of life. Only mix with cold ears. You are on the clock…add an analyzer to your day to save the day. It doesn't lie whilst your ears NATURALLY will. God bless…I got off on a tangent there. It was a topic of a class I was just teaching for live mixes of a 9 piece horn band. Sorry…time to go to bed y'all. g'nite. JV

  11. I do not understand what we're doing here, are you mixing several songs for an album, are you mastering a whole album or mixing one song by comparing various mixdowns of the same song and choosing the best one?

  12. quite useful. even though i am just starting out, it still makes sense and seems like common sense now.

Leave a Reply to Pete Alberti Cancel reply

Your email address will not be published. Required fields are marked *