We open on wide shot of a forest. A river runs in the distance. Not far from the river, emerging from the trees is a bloodied man in a torn business suit, limping and desperate for water. Cut to an over the shoulder shot of him staring at the river. Cut again and the camera is right on the water as he leans in for a drink. The focus (for our purposes) isn’t the man or his torn and blood soaked suit (I just added that for some flair). From a sound editorial standpoint, the complicated element here is the river. It’s far off in the distance, now it’s close to us, now it’s full frame in an extreme close up. As a viewer, the camera is our proxy here. Wherever the camera sits, so do we. And so, as the perspective of the camera changes so does our perception. Distant in that first shot, the river needs to very low in volume. As soon as the camera brings us closer, we need to feel more presence from the river. When we jump to the extreme close up, that river should engulf the soundscape as it engulfs the frame. A sound editor makes this possible by properly cutting for perspective.

Why do we cut for perspective?

It’s all about making our editorial mixable. Picture that scene I described above, wherein the editor provides just one long river sound, across all cuts. The mixer can’t simply play it this way in the scene - eyeballing the cuts and making rapid fader movements to match. Even the most seasoned of mixers won’t necessarily nail this. Between panning and volume changes, it’s pretty much an impossibility. In reality, it's much more likely the mixer will have to take the time to recut your work themselves for perspective just to make it mixable. Not good. When the editor cuts the perspectives for the mixer, the shot cuts are clearly delineated, with a ton of lead time to prep fader moves on the fly. Even better, the mixer can simply highlight the sound over each cut and adjust automation offline. 

When do we cut for perspective?

Perspective cutting can be used on dialogue, foley, even source music cues, but the majority of perspective cutting you’ll be doing will be on what we call steadys. These are sounds that persist in a scene and are often ambiences (for more details on what constitutes an ambience vs a background, check out this previous post). It’s very common that you’ll need to introduce these sounds and then hide them away so they are not distracting.

So how often should we make the cuts? The instinct here is going to be to cut every time the camera shot changes. In reality, you only need to cut when the position of the camera demands a significantly different treatment of your sound. Think of most situations as two options - either very close to your sound or somewhat near your sound. In rare cases (as in my river example above), you may want to add far away from the sound as a third option but to be honest, you’d need to have a pretty quiet soundscape to make something so distant play in a final mix. Often times the choice then would be to simply cut the sound out all together.

Let's have another example for illustration purposes. I walk into the Boom Box kitchen, turn on the microwave and make some popcorn.

Close up perspective (very close to your sound)

Close up perspective (very close to your sound)

Medium perspective (somewhat near your sound)

Medium perspective (somewhat near your sound)

Here you have a close up of the microwave and a shot of it simply running in the room. These should be your only two perspectives. This is the part that bears repeating - if the camera stays at an even somewhat relative distance to the microwave but moves around the room, you do not want to cut those similarly distant shots for perspective.

How to perspective cut

Step 1: Slicing up Your Audio

Find the frame where the shot cuts and split your region on the cut. You now have two regions, one on the A Side of the cut and one on the B Side. Repeat this process, cutting your regions to match the visual cuts for perspective only where necessary.

Step 2: Organization

Now that you have your cuts, the next step is to checkerboard your regions, pulling every other region down to the next track below. Even if you are cutting between more than 2 perspectives, I suggest you simply use two tracks to accomplish this. It's cleaner editorial and takes up much less real estate, especially if your steady has multiple elements.

Checkerboarded steady
Single layer perspective cutting with vertically aligned fades

Step 3: Create One Frame Overlapping Cross Fades

Extend the B Side region back one frame and create a one frame fade IN. This fade should start on the A Side of the cut and end on the B Side. Create a one frame fade OUT on the A Side of the cut. The result is two identical fades that are lined up vertically one over the other.

Step 4: Extra Credit for Volume Automation

Lastly, if you want things to play well for your clients, you’re going to need to set some volume levels. I suggest you do this entirely with volume automation (rather than clip gain) as it will be much easier for your mixer to revert back, should they want to mix these level from scratch. A warning here - don’t go too crazy with the difference in volume or it will feel unnatural.

Perspective Cutting Volume Automation.png


Proper perspective cutting is invaluable. That said, an improper understanding of these rules could lead to more trouble than simply avoiding it altogether. Here are a few pitfalls worth mentioning.

Multi-Layered Builds

Perspective Cutting Stacked Regions.png

A steady build that has more than one element, such as the microwave motor and popcorn popping within from my earlier example, should be layered in groups. The vertical order of regions should remain the same in each group and the regions should stay clustered together. Again, this makes each build easy for the mixer to adjust all together.

All or Nothing

There are no situations that call for perspective cutting only part of a build. All elements of a build need to be cut together for perspective. If for example you have a stampede scene with a bunch of animal vocals and a rumble element, it would be absolutely necessary to perspective cut both the rumble and all the animal vocals as the camera changes perspective since all of these things remain the same relative distance to the camera.

Location, Location, Location

Perspective cutting can take up a lot of track real estate. I suggest you try and keep as much of it as possible at the bottom of your general sound effects editorial tracks to stay out of the way of the rest of your editorial.

This is a complicated topic but it's absolutely essential. Of course no single method for perspective cutting is going to satisfy everyone but these rules should give you a great baseline.

What essential skills do you recommend learning for today's sound editor? Share your ideas in the comments.