5 Serban Ghenea and Max Martin Mixing Techniques from Don’t Call Me Angel

“Don't Call Me Angel” by Ariana Grande, Miley Cyrus, and Lana Del Rey was mixed by Serban Ghenea — one of the most credited mix engineers in pop and hip-hop — and produced by Max Martin's team. Opening the studio multitracks reveals that the surface-level simplicity of the record hides a small number of deliberate decisions that separate it from the thousands of pop mixes that sound almost-but-not-quite-commercial. None of the moves are expensive. None require rare plugins. What they require is intention.

This article walks through five of the most instructive production and mix details uncovered in the session: hi-hat pitch and panning, kick-to-808 waveform interaction, 808 drive automation and widened sub-bass in the Lana Del Rey verse, creative low-pass filter sweeps on the recurring main riff, and the kick-tuning question — the kick lands on A# while the song sits in C minor, which is technically the “wrong” note, and the record went platinum anyway.

1. Hi-Hat Pitch Automation (and the Mono-Center / Off-Center Trick)

The hi-hats on Don't Call Me Angel do something that is almost invisible on casual listening and obvious the moment you open an analyzer on them. The pitch of the closed hi-hat shifts subtly from bar to bar: one bar sits slightly higher, the next sits slightly lower, and the pattern alternates across the song. It is a very small pitch difference — probably a semitone or two — but it is enough to create an alternating ebb and flow in the rhythm that keeps the groove from feeling stamped and mechanical.

This is a production-stage decision, not a mix decision, and it is almost certainly a Max Martin-team move. The Martin camp is famous for treating repeated elements as opportunities for micro-variation. Anything that plays the same thing for eight or sixteen bars is either getting an automation move, a filter sweep, a pitch shift, or a layered variation. The closed hi-hat sample in Don't Call Me Angel happens to get pitch automation, which is the cheapest of those tools and the hardest to detect by ear.

The panning decision is just as deliberate. All closed hi-hats are mono and placed dead-center in the stereo field, which anchors the rhythm and keeps them from competing with the vocal and snare for edge space. But one open hi-hat is panned slightly to the left. That single off-center element introduces width and movement without collapsing the mono foundation. It is the opposite of the common amateur move of spreading every percussion element across the stereo field and ending up with a blurry, mid-scooped groove. The lesson is not “keep everything mono” — it is “keep most things mono and let one small element break the symmetry.”

2. Kick and 808: Waveform-Level Sculpting

Aligning the kick and 808 tracks in the session and zooming into the waveforms reveals the kind of interaction that only happens when a beat has been mixed as it was produced. Every time the kick hits, the 808 waveform dips close to 0 dB for the duration of the kick transient, then returns to full level. The kick punches through. The 808 holds the low end. Neither element overlaps with the other at the moment of the kick attack.

This could be sidechain compression, intentional level automation written in the DAW, or a committed bounce of the 808 with ducking already baked into the audio file. Any of the three produces the same result visually. Many top producers commit the ducking at the production stage rather than relying on a sidechain plugin at the mix, because printing the behavior removes one more thing the mix engineer has to decide later. For a full breakdown of the different kick-808 ducking techniques, including how engineers like Bainz combine volume automation with Waves Factory Trackspacer for the same result, see our guide on how to sidechain kick and 808: 4 advanced tricks, and our breakdown of 5 Bainz mastering techniques. The takeaway from the Don't Call Me Angel session is the same either way: whatever technique you use, commit to a result where the two elements do not step on each other at the waveform level.

3. 808 Drive Automation and Widened Sub-Bass

The 808 on Don't Call Me Angel stays mono and relatively clean for most of the song. It holds a consistent, solid low-end presence through every verse and the earlier choruses. But the final chorus introduces a subtle drive or distortion on the 808 — enough to add harmonics and some higher-frequency content, not enough to change the overall tone. The effect is that the 808 suddenly sits slightly more forward in the mix, cuts through better on small speakers, and adds a textured edge that complements the emotional climax of the song.

This is a sound design choice disguised as a mix decision. The same 808 sample carries the same role all the way through, but its tonal character evolves once across the arrangement. Listeners do not consciously notice the change. They notice that the final chorus feels bigger, and they do not know why. That is exactly the intended outcome.

The Lana Del Rey verse does something even more unusual. The sub-bass in that section appears on the analyzer as visibly wide — stereo content in a frequency range that is conventionally kept mono for compatibility across playback systems. Widened sub-bass can cause phase cancellation when the mix is collapsed to mono (in clubs, on TV, on Bluetooth speakers), and it will always sacrifice some punch compared to a mono sub. Top producers occasionally accept that trade-off in specific sections of a song for the sake of atmosphere. The Lana verse on Don't Call Me Angel is deliberately dreamier, more spacious, and less rhythmic than the Ariana and Miley sections, and the widened sub-bass reinforces that mood. It is a rule-breaking move executed with intention. If you try it on your own tracks, mono-check it obsessively.

4. Creative Low-Pass Filter Sweeps on the Main Riff

The main riff of Don't Call Me Angel plays consistently across the entire track. Left alone, any repeating element eventually fatigues the listener, no matter how good the initial hook. The production team handles this with one of the cheapest and most effective sound-design tools in pop music: an automated low-pass filter.

The riff sounds nearly identical in most sections, but at the start of Verse 1 a low-pass filter cuts everything above 7 kHz. Over the next bar, that filter gradually opens back up to its original ceiling of 11 kHz. The listener's brain registers the reintroduction of the high frequencies as a fresh event, even though the underlying riff has not changed at all. That single automated sweep is enough to reset attention on a repeating element and make the arrangement feel more dynamic than it actually is.

The same technique works on rhythm guitars, pads, arpeggios, and any looping element. The effective version of the move is subtle — 3 to 5 dB of apparent brightness change, not a dramatic dubstep-style filter sweep. Listeners should feel the variation without necessarily hearing it as an effect. For a deeper look at how top engineers use automation to keep long-form arrangements from fatiguing, see our breakdown of 5 Jaycen Joshua mixing techniques, which covers the same principle applied to transient shaping rather than filtering.

5. Should the Kick Be Tuned to the Song? The Don't Call Me Angel Answer: No.

Whether to tune the kick drum to the key of the song is one of the most persistent debates in modern production. Don't Call Me Angel provides a surprisingly clean answer: the song is in C minor, and the kick lands on A#. That is not the tonic of the key. That is not the dominant. That is not a harmonically obvious choice. And the record went multi-platinum.

The analysis itself is easy to replicate on your own kick. Load FabFilter Pro-Q 3 on the kick track, set the analyzer resolution to Maximum and the speed to Very Fast in the settings menu, and enable the piano-roll overlay. The first strong peak in the low end is the kick's fundamental, and the piano roll translates it directly into a pitch name. For a full guide to Pro-Q 3's analyzer settings and dynamic EQ workflow, see our walkthrough of FabFilter Pro-Q 4 features that transform your mix.

The practical takeaway is that tuning a kick is a creative decision, not a rule. If you are mixing, your priority is blending elements cohesively, not retroactively pitching the producer's kick to the key. If you are producing, the question that actually matters is whether the kick and bass complement each other in the overall low-end structure — whether their fundamentals collide, whether their attack envelopes overlap, whether the two elements are fighting for the same space in time and frequency. A kick on A# in a C minor song can sound perfect if the 808 and the harmonic content of the mix swallow the dissonance into the groove. A kick tuned to C in the same track can sound wrong if the kick and bass are fighting at the fundamental frequency. Pick kicks for fit, not for theory.

Why These Five Techniques Work Together

The five techniques above look unrelated in isolation, but they share a single underlying principle: repeated elements in a pop arrangement need controlled variation over time, and every decision about the stereo field should be deliberate. The hi-hats vary by pitch. The 808 gets drive in the final chorus. The riff gets a filter sweep on the first verse. The sub-bass widens in the Lana section. Every one of these is a small, specific move on a single element. Stack enough of them across an arrangement and the song sounds alive even though each individual layer is doing something extremely minimal.

This is also the difference between a mix that sounds like a Max Martin and Serban Ghenea record and one that sounds close but not quite commercial. It is not the plugin chain. It is not some secret compressor. It is the sum of a dozen invisible decisions that most producers and engineers do not make because they do not know they need to. If you want another example of the same principle at work, read our breakdown of how to mix vocals like Ariana Grande and our full Lana Del Rey vocal chain breakdown — both artists perform on this same track, and both vocal chains follow the same “minimal moves, all intentional” philosophy.

Frequently Asked Questions

Who mixed Don't Call Me Angel by Ariana Grande, Miley Cyrus, and Lana Del Rey?

Serban Ghenea, one of the most credited pop and hip-hop mix engineers in modern music. His credits include Taylor Swift, Justin Bieber, Bruno Mars, Beyoncé, The Weeknd, and hundreds of other chart records. The track was produced by Max Martin's team, which is why it carries the hallmarks of the Martin production sound — ruthless arrangement clarity, deliberate stereo field decisions, and repeated elements kept interesting through subtle automation.

Should I tune my kick to the key of my song?

Usually no. The kick on Don't Call Me Angel lands around A# while the song is in C minor, and it still works on a chart-topping record. Tuning is a creative decision, not a rule. The question that actually matters is whether the kick and 808 complement each other in the overall low-end structure, not whether they share a fundamental with the musical key.

Why are closed hi-hats kept mono and centered in pop mixes?

Mono, centered closed hi-hats anchor the rhythm and keep them from fighting the vocal, snare, and other central elements for edge space. Closed hats are repetitive by nature, and spreading them across the stereo field would blur the groove. Keeping them mono while placing a single open hi-hat slightly off-center introduces stereo interest without disturbing the mono foundation — a common Max Martin production move.

Can sub-bass be widened in stereo without causing phase problems?

Yes, but only when done intentionally and with rigorous mono monitoring. The default rule is mono for compatibility on clubs and PA systems, but top producers occasionally widen sub-bass in specific sections for atmosphere, as heard in the Lana Del Rey verse of Don't Call Me Angel. Limit the widening to specific sections, mono-check at every stage, and accept that some punch will be traded for immersive character.

How do you keep a repeating riff interesting over the course of a three-minute song?

Automate a low-pass filter. On Don't Call Me Angel, the main riff is nearly identical across most of the song, but at the start of Verse 1 a low-pass filter cuts everything above 7 kHz and opens back up to 11 kHz over the next bar. That single automated sweep resets listener attention without changing the underlying riff. The technique works on any looping element — guitars, pads, arpeggios — and is especially effective in genres built on loops.