How To Create 'Depth Of Field' In Your Mixes

In our simplified scheme, the room is divided into three zones.

Any reverb plug‑in can be used to create the perception of depth — and your mixes will be more engaging and three‑dimensional as a result.

Mixing engineers frequently talk about 'depth of field', and placing instruments in a space. That sounds grand and complicated, but often it simply means using two or three different reverbs. In this article, I want to share a way of creating a really realistic depth of field that can be used to glue your entire mix, with three instances of just one reverb.

To be clear, there’s absolutely nothing wrong with using different reverbs. In fact, every engineer I know does that, myself included. But doing so does not usually create a realistic perception of depth of field. Rather, it’s about the sound and feel of the specific reverbs used, which is why I call them ‘special sauce’ reverbs.

So what is a realistic depth of field?

Points Of View

Imagine you’re standing on a hill overlooking a small town, and beyond that, you can see all the way to the open sea. In effect, you’ll see three planes of distance. The houses nearest to you will appear quite sharp, and you’ll be able to spot some details. Beyond that, things start to get a little blurry: you can still see the individual houses, but details soon get lost with distance. And further still, everything kind of blends together, with the waterfront and horizon being the only clear things you can distinguish.

If we transfer this basic concept to audio, we end up with three basic planes: near, mid, and far. It is not by coincidence that you can spot these terms in many engineers’ reverb setups. Often they are used more or less interchangeably with the terms short, medium and long. What this usually means is that, whatever three reverbs are used and whatever they are called, one is set for a small ambience, one is set for something bigger such as a room, chamber, plate or hall, and one is set for something really large, which could be anything from a hall to a plate or cathedral. This is all fine and valid, and I’m not saying you shouldn’t do this. This is how hit records have been mixed for decades. What I’m saying is that using different verbs with different lengths for the different planes does not deliver realistic depth of field perception. So let’s jump right in and see how you can achieve this.

I said at the beginning that you only need one reverb. That’s true, and it can be pretty much any reverb as long as you can set up the following basic parameters: pre‑delay, high cut, and the balance of early reflections and reverb tail. Of course, in the rare case that your favourite reverb doesn’t feature pre‑delay, you could just insert a delay plug‑in before the reverb. I find algorithmic reverbs are preferable for the task, but feel free to use a convolution reverb if you prefer. I use Softube’s classic TSAR Reverb for this myself because it offers all the controls needed, sounds good and works well.

Before setting up any parameters, let’s think about the theory behind it. Imagine you’re not standing on a hill overlooking the town, but at one end of the live room in a recording studio. What actually defines the three planes within a rather small, confined space? It’s the proximity to a wall: not only does a close‑by wall give the most obvious early reflections, it’s the boundaries that define the size and sound of a closed space. So the near, mid and far planes in the live room correspond to close to the wall, further into the room, and finally deep in the centre of the room. You could also think of them as progressing from ‘sharp’ to ‘blurry’. What we want to do is set up these positions in the reverb, but all based on the same room algorithm.

(I’m aware, of course, that this is a simplification of reality. Every room has a lot of boundaries, and they are usually made of different materials, especially in a recording studio. The goal of the whole approach is to provide a very convincing and effective illusion of the real world that’s very easy to use.)

Inner Visions

First, we have to decide on the actual patch. This is completely up to you, but since the goal is not to swamp your tracks in verb, the actual reverb time probably shouldn’t be too long. And because we are aiming for depth and ‘glue’ rather than obvious reverb, the artificial room should sound fairly controlled, just like a good studio live room. This is where a damping control on the reverb plug‑in may come in handy.

After you have decided on a nice patch, set up three stereo sends, all with the same patch loaded on the same reverb plug‑in, and label them something like AMB A, AMB B and AMB C. Route their fader returns to a bus or group called something like AMB Master, which will let us hear and control our virtual room as one.

AMB A will be our position closest to the wall, so let’s start with that. Leave your pre‑delay at 0ms, set the blend of early reflections and reverb to around 75 percent in favor of the early reflections, and your high cut at 10kHz. This tames the sharpest highs but keeps the effect very up‑front, which is what we want. AMB B will be our position further into the room, so we will introduce some pre‑delay. Start with about 25ms, set the ER/tail blend to 50 percent and the high cut to 5kHz. Moving away from the wall is mimicked by introducing the pre‑delay and rolling off more highs, in line with real‑life physics of sound: the farther away you get, the more the high frequencies are attenuated (and lows for that matter, but wait). AMB C, you guessed it, will be the position far away in the room, so we double up the pre‑delay to about 50ms, set the blend to 25 percent early reflections only, and further cut the highs, setting the high cut at 3kHz.

Three instances of Softube’s TSAR reverb configured for close, mid and far distance settings. You can use almost any algorithmic reverb for similar purposes.

Take a look at the illustration to visualise the concept again. As I said, it’s a simplification, because we are not taking the ceiling or the rear wall into account. Doing so would make for a far more complex setup, without sounding significantly more realistic. By contrast, what does make this simplified approach more convincing is narrowing the stereo width of the returns progressively: keep AMB A wide, pan AMB B in at about two thirds on each side and AMB C about one third. That way, you are really “looking into the three‑dimensional room” as on the picture. Feel free to experiment with this.

Level Pegging

So far, so simple, right? Now comes the fun part. In my experience it’s super important to establish a trusted ‘unity’ reference post‑fader send level. This is completely dependent on the actual reverb plug‑in you use. Sending too little dry signal into the verbs will not produce enough ambience, while sending too much will sound nasty and unnatural. With TSAR I found that ‑9dB works well on the sends, with the actual reverb returns or the AMB Master at ‑6dB. These are just arbitrary numbers I ended up with after listening.

Let’s say we have a rock/pop band mix, because that’s where this approach works best. The drums are usually positioned rather close to the wall, so they get sent to AMB A. And with drums, I mean the whole drum kit. Of course, overheads and rooms will already have significant room information baked in, but we want the whole kit to be placed in our virtual room for the approach to work properly. Of course, feel free to experiment. I tend to send all drum mics at reference level to unify the kit, with only the kick being drier. Same with strummed acoustic guitars: we typically want those rather dry, so I usually start with sending a little less. Close‑miked electric guitar amps definitely benefit from a little more room, so start by sending them to AMB B at a little less than your reference level. For your pianos, keyboards, strings or horns, start with sending them to AMB B at reference level and adjust from there. Finally send your lead vocals and lead instruments to AMB C. These get the most depth from the room position, but since that has the longest pre‑delay, the dry portion stays up front.

The beauty of this simple approach is that you can realise not only those three positions, but intermediate positions as well. This works pretty well for background vocals, for example. By sending them to AMB A, they get some tight room information surrounding them, and by also sending them to AMB B or AMB C you can introduce some of the size of the room, too.

Diving In

Now it’s time to listen to the AMB Master in solo. You should hear a nice, three‑dimensional ambience that gives you the illusion of standing in the live room, surrounded by the instruments. The drums should sound rather loud and tight, the acoustic guitars barely noticeable, the electric guitars blasting away into the room, and everything else somewhere in between, with the vocals sitting nicely in the music. Unsolo the AMB Master and turn it all the way down. Now slowly raise the fader until you can really hear your mix opening up and losing its two‑dimensional dryness. Note that position and raise further until you can really hear the effect and the mix starts to sound ‘wet’. Note that position, too. Return to the first position and listen for a while. Mute the AMB Master. The mix itself shouldn’t sound different now, but flat. Unmute the AMB Master and lower the fader until you feel you can’t hear the effect. Note that position again. Now you have a window from ‘almost dry’ to ‘rather too much’, which you can work within for different parts of the song.

It takes some practice, but when done right, it’s astonishing how loud you can actually make the AMB Master without your mix becoming obscured or overly wet.

It takes some practice, but when done right, it’s astonishing how loud you can actually make the AMB Master without your mix becoming obscured or overly wet. Instead the result feels not unlike using a parallel bus compressor that fills the space between instruments and glues your mix together. I’ve prepared a couple of audio files, so have a listen yourself. After you’ve got the hang of it, there’s still room for further tweaking. Try cutting some lows on the returns (either with a high‑pass or with a shelf): start with 100Hz on AMB A, 160Hz on AMB B and 220Hz on AMB C. Try EQ’ing the AMB Master: sometimes some midrange frequencies can get harsh, and in a busy mix, it can help to further roll off lows and highs here. Something like Oeksound’s Soothe 2 also works wonders here to keep rogue frequencies under control. Also, try panning the sends for your widespread guitars to the opposite side, for a nice sense of width and better mono compatibility.

As an extension of this idea, you can also try using mono reverbs instead of stereo instances. For AMB A, set up three instances hard left, right and centre. For AMB B set up two instances and pan them half left and right. And for AMB C, try just one dead centre, or keep the stereo instance, but narrow the stereo field in your DAW to spread it just slightly. The mono setup is a little more cumbersome in use, but can give you a really nice and articulate separation and a super‑realistic depth perception.

As I said at the start, this technique is not meant to replace your favourite reverbs. By all means, keep using your favourite vocal or snare reverb as well. But I find that with this technique you’ll need less of those ‘special sauce’ verbs — and the ones you do use can really shine. I call that a win‑win.