How To Mix Vocals Like Dua Lipa: Josh Gudwin EQ, Compression, Plate Reverb, Delay, and Parallel Compression

A Dua Lipa-style vocal usually feels tight, polished, modern, and confidently forward. The source transcript points to a workflow inspired by Josh Gudwin that relies on carefully staged cleanup, smooth compression, dynamic resonance control, and a set of aux effects that add width and depth without clouding the lead.

One of the smartest ideas in this workflow is that vocals should not always be mixed as one single track. Different sections and takes often need different treatment, especially when the singer moves between registers, dynamics, or tonal attitudes. That mindset matters as much as the plugin chain itself.

1. Start with Corrective EQ Before Compression

The first move in this chain is Pro-Q 3 used as a cleanup stage. The transcript describes rolling off below 100 Hz, dipping around 250 Hz by roughly 1.5 dB, and cutting around 4 kHz by roughly 3 dB. The logic is not just tonal. It is also dynamic. Compressors often react unpredictably when too much low-frequency energy or harsh upper-mid buildup is left untreated.

By removing those problem areas first, the compressor sees a cleaner and more relevant signal. This is one of the clearest answers to the question why EQ before compression on vocals?Sometimes the best reason is not tone alone, but compressor behavior.

2. Use LA-2A Compression for Smooth Control

After the first EQ comes an LA-2A-style compressor. This is a classic move when you want a smoother, more musical leveling stage rather than an aggressive peak clamp. The transcript notes that the sweet spot is moderate gain reduction, enough to stabilize the vocal without making it feel overly grabbed.

This stage works especially well after the first EQ because the compressor is no longer being pushed unnecessarily by rumble or overly sharp energy. It can focus on musical movement in the lead vocal instead of random distractions.

3. Add a Touch of SSL Channel Compression and Tone

The chain then introduces an SSL E-Channel style processor for a slight touch of additional shaping. The important idea here is restraint. This is not described as a heavy-handed stage. It is more of a finishing touch that tightens the vocal a little more and adds character in a familiar console-style way.

Using small amounts of compression across multiple stages is often cleaner than relying on a single aggressive processor. That theme keeps appearing across these vocal workflows.

4. Follow Up with Secondary EQ and High-End Finish

Another Pro-Q 3 stage is then used, dipping around 250 Hz and 450 Hz by about 1 dB and boosting the top end around 20 kHz with a high shelf of roughly 3 dB. That tells you something important about this chain: the low-mid cleanup is being done in small, layered moves rather than one dramatic cut.

The high shelf at the end adds polish and air, but because the chain already dealt with buildup earlier, the brightness has a cleaner foundation underneath it.

5. Use Pro-MB to Control Resonances Dynamically

According to the source, Dua Lipa’s vocal also needs resonance management, and that is handled with FabFilter Pro-MB. This is a smart choice because resonances are often inconsistent. A static EQ cut might be too much in one phrase and not enough in another. Dynamic multiband control allows the vocal to stay natural until the problem region actually builds up.

This is especially useful on pop vocals where the arrangement may be sparse in one section and dense in another. A dynamic stage adapts better to those changes than a fully static tone-shaping move.

6. Use Three Separate Auxes for Depth and Space

The transcript notes three aux channels. The first is a plate reverb with 2.5 seconds of decay and no pre-delay or low-cut. That provides a clear, smooth vocal ambience with a familiar pop character.

The second aux is more complex: a quarter-note ping-pong delay followed by Valhalla VintageVerb, with the reverb mix around 43% and about a 5 dB cut around 2 kHz to clear room for the lead vocal. This is a great example of building an effects return that sounds beautiful but is still carved away from the vocal intelligibility zone.

The third aux uses a Bricasti M7-style reverb for additional depth and richness. Separating these spaces lets you blend more deliberately instead of forcing one reverb to do everything.

7. Filter the Parallel Compression Bus Aggressively

Parallel compression is also part of the chain, but its routing is disciplined. Before hitting the compressor, the signal is filtered below 100 Hz and above 14 kHz. Then it hits a CLA76-style compressor with ratio 8, super-fast attack, and fast release, aiming for up to around 10 dB of gain reduction.

After compression, the bus is filtered again below 200 Hz and above 10 kHz. This is one of the best details in the whole workflow. It keeps the parallel path focused on useful midrange thickness and excitement rather than adding low-end mud or scratchy top-end hiss.

8. Blend, Don’t Smash

Once the filtered parallel bus is set up, it is blended into the lead vocal rather than replacing the main tone. This reinforces a core principle of modern vocal mixing: the parallel path should support the emotional density and apparent size of the lead, but the dry lead should still carry the identity and detail of the performance.

Practical Workflow Summary

1. Clean up the vocal before compression so the compressor reacts more musically.
2. Use an LA-2A stage for smooth leveling rather than aggressive clamping.
3. Add only a slight touch of SSL-style compression and channel tone.
4. Layer secondary EQ moves instead of relying on one extreme curve.
5. Use Pro-MB for dynamic resonance management rather than static over-cutting.
6. Separate plate, delay-plus-reverb, and additional depth reverb into different auxes.
7. Filter the parallel bus before and after compression so it adds density without junk.

Frequently Asked Questions

Why not mix all the vocals on one single track in this style?

Because different vocal parts often need different settings. A chorus may need more air and density than a verse, and a lower register may need different low-mid control than a higher register. Treating all sections identically can make at least one section wrong.

Why cut low frequencies before the compressor?

Low-frequency energy can trigger gain reduction more than intended, especially on smooth optical compressors. Removing unnecessary low-end before compression helps the compressor follow the vocal more musically instead of reacting to rumble or proximity effect.

What does Pro-MB solve here that EQ alone cannot?

It reacts only when resonances become a problem. A static EQ cut is always active, even when the frequency area is not misbehaving. Pro-MB gives more control over inconsistent harshness or buildup without flattening the vocal the whole time.

Why filter a parallel vocal bus so much?

Because the job of the parallel bus is density and presence, not full-spectrum replacement. If you leave too much sub or extreme top end in the parallel path, you usually get mud, hiss, or harshness. Narrowing the band focus keeps the blend cleaner.

Why cut around 2 kHz on the delay-plus-reverb return?

That range is crucial for vocal intelligibility and forwardness. Reducing it on the effects return helps the ambience support the lead without masking the part of the vocal that carries the words and presence.