I always had my own order of operations for how to see a mix through from start to finish, but I never crystalized it until this year. Teaching has a way of doing that for you. In the pandemic, a couple of longtime clients started showing an interest in learning the craft of mixing, and they came to me for weekly zoom lessons. This expanded to zoom lessons with other engineers who had read my columns or heard my work.
Quickly I was asked for a roadmap for how to proceed through a mix, and so I put one together. it became the foundation for my students, and I’m sharing it with you now.
Whether you’re mixing your own productions at home or have a few hours to put up a mix at a hired-out studio, this order of operations should help.
When starting your work, first check that everything in your studio sounds exactly as it ought to sound. There’s so much to keep track of in the studio, even if the setup is simple (one audio interface, a computer, and a pair of speakers). Cables wear out, software can glitch. So it’s good to check.
By no means do you need to run a comprehensive set of diagnostics here—unless something sounds immediately wrong. Listen to your favorite references at the proper level in your studio and make a quick determination. If you have a hybrid setup, run a test-tone through your customary, starting-point chain and make sure everything sounds as it should. It won’t take long, and it will save you a ton of time later.
Check out some iZotopians’ favorite reference mixes in the video below:
The client will send you a rough mix, or you will have a production session that you’re working from. Listen to what you have. Make initial notes of what you like, don’t like, and potential issues (I find paper lists help me stay more focused, though I’ll often use text docs if my hands are cramping).
Use the rough mix to gather the appropriate references for the song, though these often come from the client.
You should use their references, but also compile your own to foster an emotional connection to the material. Gather the references in one place; either drag them into the session, use Ozone’s Reference pane, or any number of referencing products. Do try to find lossless files whenever possible. Bill the client for these if they expect you to do so; eat the cost if you’d rather not look so cheap; I don’t mind spending five bucks or so to get my mind situated, and possibly hear some cool music as a result.
Import the assets into your session from the client. Or, if you’re mixing your own production, click Save As and call it “.mix” or something clear to indicate this is the mix.
I prefer to use a basic mix template that speeds up my workflow while allowing for some flexibility. By all means, use your own template, or ask us if you’d like help in crafting a basic mixing template.
Organize the session in a way that makes sense to you. I tend to organize my sessions thus:
Submix aux returns first (This is part of my template)
SFX, if applicable
Harmonic instruments (guitars, chordal synths, pianos, etc)
Melodic instruments (anything playing a melody rather than a harmony)
I also have a color coordination scheme that I use for every mix. You should have one too, provided you are biologically disposed to find color-coordination useful.
Don’t route anything just yet—that comes later. Just keep stuff organized so you know where it all is, which will help you in the next task:
You need to make sure you actually have all the tracks that the client sent—or that you actually have everything from your original production session.
There’s a scientific way to do this, and a not-so-scientific way. First, we’ll cover the scientific way:
If the files look like they line up with the rough mix—if playback between the rough and session files seems to start at the same moment—try a null test.
Flip the polarity on the rough mix and see if you get silence on playback. This will tell you if you’re missing any parts, or if you have additional parts that aren’t quite right.
90% of the time, a null test will not work. Files almost never line up with the rough in the real world. Many “bounce in place” operations leave everything a few samples off. You can try aligning to correct for this, but bounce in place can be volatile for a variety of reasons.
Sometimes, you get an “almost null” where the volume is consistently quiet, and nothing sticks out; it’s just not perfect. This can tell you there was stereo bus processing that you don’t have. So long as the playback doesn’t have obvious leap-outs, you can largely use this result as a satisfactory, relatively conclusive asset check.
Now you’ll see why we advised you don’t route your tracks at this stage. You’re far more likely to disturb any eventual null if you change the routing order, accidentally double-routing a piece of material in the process, or sending some submix to a part of your chain that’s always engaged.
A quick tip: you can save time in the null test by exporting the phase-inverted rough mix against your raw tracks and examining the resulting file visually. Read the spectrogram in RX, or the examine waveform in your DAW. So long as playback is either silence or consistently quiet, you’re probably in good shape—though be aware that the word “probably” does a lot of the heavy lifting here.
Here’s the truth: most of the time, you’ll have to use your ears to determine if you’re missing anything. That’s fine, it only takes five to ten minutes depending on the song. You may have to set up a cursory static mix to balance against the rough. This is also okay.
You’ll see that the next steps pertain to editing; sometimes, the producer has done such a fantastic job that you don’t need to edit a thing. Sometimes this is not the case—only the initial listening experience will tell you.
Hopefully, while you were taking notes on the rough mix, you had a chance to ask yourself if anything sounded unintentionally off; you’ll have noticed if a fill felt sloppy, or if a bass part felt wrong. If there were any such issues, now is the time to fix them: use your ears to get the proper timing for instruments.
No way around it: you have to use your innate musical sensibilities to help you here. The grid will not save you. If you’re not sure if something is an intentional choice, find a way to ask the producer nicely. I find doing it over the phone helps, as it eliminates potential “text tone” issues.
Would that all audio was pristine, free of clicks, pops, and horrid distortions! Sadly, this is not the case. Often we get audio that craves de-noising, de-clicking, de-clipping, and more. It is our job to provide these fixes—to scrub the audio clean. Otherwise, clients might not be satisfied, and mastering engineers will post screenshots of your audio under the RX microscope to Twitter, where the masses will laugh at you.
I have personally never had to comp a track for a project I wasn’t also producing. I find this task is best left to the producer and the production process. I have, however, heard of mixing engineers being asked to compile one usable take out of many. I wouldn’t do it unless there was a separate hourly rate involved; strictly speaking, it’s not part of the job of mixing—at least, not in my opinion.
Sometimes I do have to choose between two takes, and this is a different issue.
“Hey,” the producer might say, “on the second chorus, I felt the singer delivered two completely different but usable vibes. I like them both, so let me know which works in the mix.”
This is fine for me, and falls outside the purview of a dedicated hourly rate; I take care of these issues around this point in the process.
Often you will be responsible for getting the singer in tune. Here, a question arises: do you do this during this round of editing? Or later? Many people prefer to correct pitch in the beginning or have their assistants do it for them if they have the luxury.
I’m not one of those people, but maybe that’s because I mostly work in genres where pitch choices are always intentional. You don’t get a lot of unintentionally bad singing in country, Americana, rock, and mood pop. It does happen sometimes, however.
In my workflow, I prefer to leave pitch correction to the actual creative part of the mix, because there are so many avenues for it. Melodyne? Sure. But it doesn’t end there: other algorithms work better for specific purposes. Maybe the answer isn’t pitch correction, but detuned stereo harmonizing; this is a valid option.
Also, mixing requires giving your ears breaks, and deadlines are inimical to breaks. Pitch correction work can be a break of sorts because I’m doing something completely different from balancing instruments with each other.
As with instruments, sometimes you need to have a conversation about pitch with the producer or artist. I find this is best done on the phone, and I tend to use the word “honest” a lot.
“Hey dude,” I might say, “The way you sing the second pre-chorus—are you going for something shiny and polished, or do you want it to feel more honest and earnest?” This is a nice euphemism.
In the next article, I’m going to cover the creative steps in getting a mix over the finish line. These steps will be somewhat adaptable to suit your workflow and your studio.
But the ones we’ve just covered are absolutely mandatory for any professional job. I also find they work best in the order listed above—that flow gets me from A to Z in a relatively quick manner. Now that you’ve digested these tips, please move on to the next article, in which we cover all our bases throughout the creative process of mixing.