Growing up, I was always told by amateurs and pros alike—you can’t master a song you’ve mixed. It was interesting then to see engineers like Luca Pretolesi and Don Gehman break that rule. Could I do both too?
With technology like Ozone 9 and the improved Tonal Balance Control, I find it’s more possible than before—provided I follow concrete rules of thumb. The following are the rules I live by when I’m asked to master my own mixes, which is more and more often these days.
As a mixer, I’m always trying to suss out the right personality for the song and to translate that personality into technical processes. If I get a blues-rock mix in the vein of the Black Keys, I know what that means in terms of EQ choices, harmonic distortion implementation, and compression.
When I’m mastering, I do the same thing to a certain degree: yes, I try to achieve a balance across the entire album, one that translates to all delivery media and speaker systems. But I also note the intended vibe of the mix, and, if it falls short, I use my tools to achieve that vibe.
If I am doing both mixing and mastering, I have to think very carefully about where to impart color—where I employ big, tonal-shaping gestures to add flavor. Much as I must favor either the kick or the bass in the low end, I have to let one process win the battle of color. Either it’s going to be mixing or mastering. If I try to take on adding color in both processes, it winds up being a mess.
So, when I’m wearing both hats, I save adding color for the mix process, and aim for transparency on the stereo bus itself—on the master. This helps me to keep perspective in both arenas.
What does this mean, in terms of individual processes and tools? I’ve drawn up a list of how I operate:
Tonal issues of color and vibe are relegated to the mixing process; issues of frequency balance, clarity, translation across speakers, and competition with other releases are handled on the stereo bus. Learn about EQ in mastering here.
Issues of glue, groove enhancement, and color compression (box tone) are relegated to the mixing process; issues of global dynamic reduction—done as transparently as possible and in the service of raising the overall level—are for mastering. Take a deep-dive into compression in mastering here.
If I’m using multiband compression to fix a harsh issue in one instrument, I need to go back into my mix. If I’m using multiband compression to achieve an extra bit of competitive level, I may do this in mastering.
Issues of width are best handled in the mix itself; however, mid/side EQ for instrumental clarity can be saved for mastering.
When mastering engineers advise you to shoot for a target level or a target loudness, they’re not trying to shame you. They want to help you preserve headroom, so they can achieve their goals: balancing your entire record, making it sound good on all playback systems, and competing with similar products.
With headroom preserved, a mastering engineer has room to maneuver. Without headroom, they find it much harder. So, when mastering your own mix, leave headroom for yourself to work with.
Take the following experiment—a little loop made I made with stereo bass information, prominent low mids, and harsh mids/highs.
Here it is presented with headroom:
Here it is compressed and limited for maximum loudness; it’s quite high on the meters, but we’ve level-matched it for comparison.
If I wanted to take this loop to the final stages, I’d have to correct these tonal issues—the stereo bass mud, the prominent low mid tubbiness, and the screaming highs. We add EQ, and we get this on the original mix:
But on the mix with no headroom, it sounds like this:
Note the general lifeless quality of this example; that’s what happens when you EQ something with no headroom. But why does it happen?
When you’ve used compressors and limiters to get as loud as possible, you’ve processed the transients a whole heck of a lot. You’ve downright reshaped them in some cases, depending on how hard you’ve limited.
If you emphasize or deemphasize those transients with further EQ, you’re manipulating transients already divorced from their original intent. It’s like adding salt to a dish already seasoned—it’s going to be too much.
It’s more than the transients, however—the entire dynamic range has been constricted, and this also affects our EQ choices.
Consider a mix with lots of dynamic range, boosted at 300 Hz to impart warmth during the chorus. In loud sections, you really hear the effect of that boost, but at quieter levels, there’s less of the overall mix feeding the boosted EQ. So you can make an informed decision to keep that boost for both sections; it doesn’t affect the verse so much, simply because the verse isn’t as loud as the chorus.
Now, if we constrict that dynamic range too much, the quiet verse is much, much closer to the loudest chorus. We hear the 300 Hz boost as undue warmth on the quiet part, and it may not work, given the material.
If you are going to be the mastering engineer, you must keep this in mind. You have to set a level that gives you enough room to allow for global changes in EQ—changes that serve translation on multiple speakers, and help the master compete.
There is, however, a flip-side to this coin.
The previous section imparted the value of preserving dynamic range for the mastering process. But in mastering your own mix, you should be careful not to have too much dynamic range either. It’s counter-intuitive, so let’s explain:
Let’s say you have a mix with quiet verses and loud choruses. The difference between the last chorus and the first verse is drastic, say 7 LU short term.
In calibrating your mastering chain, you start where the energy is the highest, with the biggest, loudest chorus. You get it sounding right, with just the right amount of EQ, compression, and limiting to preserve the mix’s integrity and assure its translation and competitive level.
But what happens to your initial verse?
Ideally, it should all fall into place—the compression decisions, which you made in accordance with the track’s time and feel, should carry over. The limiting should sound markedly better, as nothing is so loud against the digital ceiling.
But in practice, something different happens:
The limiter’s algorithm subtly changes the tone. Often it sculpts the lows, and your EQ decisions account for that. Your compressor reacts to EQ as well; the loudest parts of the frequency spectrum are telling the compressor when to clamp.
Because these processes are inter-relational, things sound markedly different when a section drops drastically in level. Suddenly, the EQ signature that pushed the compressor so sweetly isn’t there; suddenly the whole low-end of the tune changes, thanks to how it hits the limiter. Now you find the track actually lacks the color it had on the chorus.
Mastering engineers usually work with a stereo file, so they have quick ways of handling this. They can gain or automate the quieter sections up by the appropriate number of decibels, preserving both density and dynamic range in the process.
In the analog world, some mastering engineers print the master in separate passes—they set the chain one way for the choruses, and then use a modified version of that chain (perhaps with slightly lower thresholds on the compressors) for the chorus passes. The two passes are then cross-stitched together.
You, however, are working from your own mix—you don’t have the simplicities of a stereo file. Sure, you can automate, but that can muddy up the waters; if you’re automating for both the mix and the master, that’s a lot to keep track of, and you could accidentally draw in the wrong move.
Forget about the logistical problems, the CPU hit could be astronomical. It could bring your session to a standstill.
So what’s the solution?
For me, it’s working harder in the mixing realm to do less in the mastering realm—to incur fewer trade-offs at the end of the chain. This involves attending to issues of dynamic range on the track and submix level, so my limiter and stereo bus compressor don’t have to work as hard in the mastering portion of my endeavors.
It goes back to our first tip: all tools impart some color, and I don’t want too much color in self-mastering. If I’m trying to have a transparent hand in compression across the stereo bus, I want that compressor to work less hard. If it’s working less hard, it’s adding less color.
The same goes for the limiter: if I’m achieving 1 dB of gain reduction to reach my desired final level, rather than 5 or 6 dB, it’s adding less color.
When you’re the sole line of defense between the artist and the real-world, you must take special care to ensure your balances are correct, both in your room, and on all speaker systems. You must make doubly sure your mix is balanced, in other words.
Take great care to make sure your low end isn’t overwhelming the mix; this is especially hard to do in home studios, where getting the bass sounding right in your room is a herculean task.
You can achieve better bass balances through improved Tonal Balance Control in Ozone, which not only shows you if your bass matches the appropriate curve for your genre, but also lets you make EQ changes from the plug-in itself.
Likewise, make sure there aren’t any unwanted, resonant hotspots in your mix. The improved Tonal Balance Control also helps for this endeavor, particularly if you’re viewing the curve in fine mode.
Watch for instrumental imbalances as well: a sweet, brilliant vocal can be obscured by overly crispy high hats. Again, Tonal Balance Control comes to the rescue, particularly when you solo a specific range.
Check out the highs in solo: that’s where you may clearly hear the rub between competing instruments, and make a judgment call as to whether one needs to be turned down.
Make sure the stereo field is balanced too. In the improved Tonal Balance Control, solo your bass frequencies to hear if too much is going on in the sides in mono. You can also solo the sides in the EQ section of Ozone 9. Do both, and you’ll have a great picture as to whether or not you meet filtered out lows on the sides.
Use Ozone to throw the mix into mono every once in a while. This will help you see if you’re losing important instruments; if you are, either your panning is off, or your phase relationships are out of whack.
Say you’ve gotten a static mix up, with level relationships, pan positions, but no EQ. Say you know the tune will end up on CD, and will need to reach a loud LUFS target commensurate with that goal.
You can throw up the Master Assistant, set it to CD, and see what it suggests for an EQ curve—its broad suggestions serve as an indication of how to proceed.
This screenshot above tells us to try and boost high end information—I can do this in the mix by adding more treble to the vocal or shimmering up the drum overheads a bit. Perhaps this helps.
Mix a little more, and do it again—only this time, throw your reference track into the Master Assistant. See what it suggestions for EQ and compression.
If it’s giving you a bump in the high-mids, try focusing your vocals there a bit more; if it’s cutting your low-mids, see if your instruments are a bit tubby, and try notching them out a little. Always make sure, of course, that you like the changes.
In mixing, you’re always racing against aural exhaustion (ear fatigue is a real thing!), and you’re always looking for ways to stay objective. You can use the Master Assistant as a check in preserving the objectivity of your ears.
As you jump from the mixing level to the Master Assistant level, make sure you maintain a constant volume on your monitor controller. If your mix jumps up 4 dB when you’re monitoring a possible mastering chain, bring the controller down by 4 dB. A consistent, healthy monitoring level will help you make the same objective choices over and over again.
How can you tell you’re monitoring level is consistent? Buy an SPL meter, or an SPL app for your phone, and make sure the reading is in the same range every time. Depending on your room, 79 to 83 dB SPL is a good monitoring level.
If the mix sounds good compared to a reference mix at the same level, and it does so consistently over a few trusty playback systems, your mix may be done. If it holds up the next day, it’s more than likely done.
However, you must also have a personal set of goals for this mix. These are subjective, emotional achievements that go beyond the technical. Does it groove? Does it make your spine tingle? Tabulate a personalized list of these adjectives after you get your static mix up. When you think you’re near the end, see if you can check these adjectives off on your list.
If the mix hits all the right notes, you’re probably ready to take a break, come back, throw on Ozone 9, and get started finalizing.
There’s a few ways to set up the limiter—the Maximizer in Ozone 9—which is a vital part of your chain, and should be handled sooner rather than later as you transition to mastering.
You can gain the actual mix into your limiter, which may look like this:
Note the input slider on the right of the GUI, gained up 3.5 dB to push audio into the limiter. That’s a valid approach. You can also lower the threshold of the limiter, which looks like this:
There are differences between the two methods. The former juices the input on any modules between the mix and the limiter, which may or may not be what you want. The latter may cause the limiter to work too hard—to knock off all the gain reduction at once, creating unpleasant squashing or artifacts in the process.
There is a third option: to gain stage within a plug-in like Ozone.
Say you run a multiband compressor in parallel (so it’s not too prominent in sound), with 0.8 dB of makeup gain across the board.
You then follow this with a dynamic EQ damping down obvious harshness in the high-midrange.
Then, you opt for a little more compression with the Vintage Compressor, auditioning the tone of the various modes, knocking off no more than a dB, and making it up a dB or so of gain.
You could then push a little into the Vintage Limiter, for 1 dB of boost with a little character. Choose among the three modes to see which character complements the tune right.
This is then followed by the Maximizer to give you one more dB.
Instead of the Maximizer being set with a threshold of -4 dB—or something else that’s absolutely insane—you now only have it set to, say, -0.8 dB.
You’ll note each processor is doing only a bit of the work. Some are imparting color, to be sure—but you’re always A/Bing against the original mix to make sure the color is complementary, and not too heavy-handed.
It’s a bit of a balancing act. Indeed, even though a lot of “in the box” engineers prefer this multi-stage approach, it’s tricky to get right and requires practice.
One way to practice is to switch between two chains. Set up Master Assistant, and then, in a different instance of Ozone, try the multi-stage approach. Toggle back and forth to hear which sounds better. Try to beat the Master Assistant as much as possible.
If you’re new to mastering your own mix, it’s wise to seek all help you can find. The improved Tonal Balance Control in Ozone 9, with its expanded preset curves and inter-plugin communication, is a great help.
The whole objective of the improved Tonal Balance Control is to give you an average idea of what the frequency response of your master should look like. This is done in comparison to references either culled by iZotope or selected by you.
The preset menu gives you many genres, from EDM to Classical, with Metal and Jazz in between. For specific targets (say, all the songs on your album, as you’re mastering them), it can be quite helpful to load your own songs into the improved Tonal Balance Control.
During the mastering portion, it’s helpful to compare your master to references from applicable genres—and this is easily done in the reference section of Ozone. At minimum, load up a genre reference. If applicable, load up the tracks immediately preceding/proceeding your present master.
Use your ear to get the EQ working as best you can, and save Tonal Balance Control comparisons until you feel you’re nearly done. This is where a couple of strokes from the Tonal Balance Control window may come in handy.
For instance, I recently had a mix for which I used Ozone 9. But at the end, I checked the Tonal Balance Control graph and noted something interesting: I was below where I should be, according to Tonal Balance Control, in the midrange.
I liked the sound of my master, but I was curious to see what the improved Tonal Balance Control would get me—so I loaded up the EQ in Ozone directly from the Tonal Balance Control panel, and the necessary adjustments.
The master, I thought, sounded much better. Furthermore, it was easy to make the comparison right there, in the meter window; this sped up my workflow quite a bit.
Using the improved Tonal Balance Control, you can see if the crest factor of the bass has dropped too low for the verses, then make changes accordingly by increasing the compression, harmonic distortion, and EQ of low-end instruments, including the bass and kick drum.
Watch your tonal graph in Tonal Balance Control, and try another iZotope-specific approach: open up Neutron’s Visual Mixer, and see if gaining or moving any particular instrument in the Visual Mixer gets you the frequency response you’d like to see.
As for how to use references in Ozone, peruse this article for more tips.
While it is now more possible than ever before to master your mix, you still must be ruthlessly careful. Always check your work against the competition in a variety of listening spaces, and have an impartial friend listen to your work to hear their thoughts. Because of the nature of perspective, there’s a lot of room for error. Thanks to tools like the improved Tonal Balance Control and Ozone, there is a lot of room for joy as well.