DIY horror sound design: make your own signature sounds, hits and drones

Right, this is my favourite bit of making horror music. Honestly, it should just be called the sound playground, because that’s what it is.

I don’t go into sound design with a blueprint. It’s playing, exploring and asking one question over and over: “What happens if I do this?” Here’s how I make signature sounds, horror hits and drones from a microphone and the effects you already own.

Your toolkit is smaller than you think

I call these sound design toys rather than tools, because that’s how you should treat them.

Something to record with. This is the most important one, and I don’t want you to panic about not having a “good” microphone. I use a Rode NT1-A, a little Rode shotgun mic (it’s smaller and easier to point at things) and, a lot of the time, my phone. The phone is genuinely enough. We’re not building sample libraries here. We’re capturing sound to add interest and texture.

The effects in your DAW. Five do nearly all the work for me:

• reverb
• compression
• saturation (distortion, overdrive, whatever grit you’ve got)
• reversing
• pitch shifting

Sometimes reversing a sound is enough. Other times I’m layering a load of things. There’s no go-to recipe. The stock plugins in Logic or Cubase are more than enough, so don’t go shopping before you’ve played. If you do want an extra toy later, a dedicated pitch shifter such as Pitch Monster or a free extreme time-stretcher such as PaulXStretch fits this way of working nicely.

What makes a sound a “signature” sound

The thing that makes a signature sound signature is that there’s something almost human about it. The note wobbles. It bends. It’s inconsistent. Sometimes it’s literally a voice. There’s something alive in it, and that’s what we’re trying to catch.

That’s why I always start with a microphone. A preset can’t be alive in the same way.

If you’ve got an instrument, pick one where you can bend or slide the note. Strings are perfect. Woodwind and brass work too, because you can force a bend. Piano is harder to get a signature sound out of, which is why I so often fall back on my voice.

And don’t try to sing it beautifully. Mine are out of time and out of tune, and that’s good.

Step by step: a vocal signature sound

Here’s the exact process I use. It really is this simple.

1. Lay down a quiet bed if you want something to react to: a soft pad or drone, nice and low in the mix. You don’t need it, but it helps.
2. Record a few takes in mono. Do one fairly normal and one slightly bonkers. Wails, groans, breaths, whatever comes out. They don’t all have to be weird, but it’s more fun when they are.
3. Reverse it. My weird take sounded a bit comedy played forwards. Reversed, it suddenly became unsettling. Reversing shifts your perspective on a sound, and that’s often the whole trick.
4. Cut it midway so it swells and then stops dead.
5. Transpose it down an octave. Then try two octaves. Keep whichever one gives you the creeps.
6. Add grit. A saturation or distortion plugin changes the texture rather than the pitch, so it turns gravelly. (I reached for Decapitator. Use whatever you have.)
7. Give it space. A big reverb is what makes it huge. Then try a low pass to take the edge off the top.

Then keep asking “what if?” What if I reverse it again? What if it’s lower still? Can I change the texture, the pitch, the space? Keep going until it has its own wonderful, horrible feeling.

Sometimes I end up with something else entirely that becomes the foundation of the track. That’s the point of playing.

Chop up a sound you already have

You can do the same with any organic recording: a cello note, a door, or (one of my students did this) a dog. Here’s what I’ll typically try:

• Don’t use the whole sample. Chop out the most interesting part, often just the tail, and fade it in and out.
• Play with pitch, carefully. Pitch moves feel very human, so use that on purpose. If the sample clashes with your track, retune it to fit. One of mine was sitting on an E against a track in A sharp, so I dropped it six semitones.
• Drench it. I chuck it in a big hall, take the dry signal out and go fully wet. I’m not picky about reverbs at this stage. I just want the sound I’m hearing in my head.
• Add a tape delay on a dotted quarter note. Turn the flutter down if it gets too wobbly, and high-cut the repeats.
• Put a reversed copy in front of it. Reverse the sound (reverb tail and all) and butt it up against the original so it sucks in and then speaks. Try the reversed copy pitched down an octave or two.

DIY horror hits: three layers

When I say hits, I don’t just mean big trailer booms. I mean any percussive sound that punctuates, the way a cymbal does on a drum kit. A good horror hit comes down to three layers.

1. Impact. Usually a drum. It doesn’t have to be: a slamming door works. It’s the big hit.

2. Sparkle, or found sound. Something small: paper, sticks, glass, beads, a shaker. In the course I literally used a handful of lollipop sticks. The key is that it feels close. Mixing up the spaces, with something right in your ear while the impact sounds far away, is what makes it eerie.

3. Pitch. Usually a cluster, or at least something with no obvious tonal centre. I scratched and played an open string on my violin for this. It shouldn’t relate neatly to the note your track is sitting on, because then it jars. You hear it and think, we’re back, it’s fine. No, it’s not!

Here’s how I build one:

1. Record your found-sound layer first, close-miked, and normalise it.
2. Record a pitch layer: a cluster or scraped string, and normalise that too.
3. Line them up, then add the impact underneath. Suddenly the whole thing makes sense.
4. Balance the three. Some hits are mostly impact, some mostly sparkle. Use them in whatever proportions the moment needs.

Picture being bashed round the head with a drum that has beads and strings hanging off it. That’s what you’re after.

Capture a bank of sounds before you start a track

People always ask me what the starting point is. My answer is sounds. Find sounds you like, capture a load of them, and only then start writing.

When I sample an instrument for this, I’m not making a detailed multisampled instrument. I’m after four things:

• Signature sounds: bends, slides, the instrument played out of its normal context
• Long pads and drones: sustained, textural notes you can stretch and stack
• Pulses: gritty bowed pulses, pizzicato, muted repeats
• Percussive sounds: taps, knocks and scrapes on the body of the instrument

Record to a click. Your timing just needs to be good enough to chop (mine isn’t brilliant). Then normalise, chop, delete, chop, delete. One afternoon gives you a palette nobody else has.

Even a tap on the body of my violin became the pulse a track was built on. Want a drone? Take one of those long notes, drop it an octave or two, reverse it, drench it in reverb and layer it under a second one. Stretching it out to extreme lengths in PaulXStretch is another fun “what if”.

Build the weird world first

This is the bit I really want you to take away. I start every horror track with the sound. I create my own atmosphere first, then lay everything else on top.

If you start with pre-existing samples, your creativity gets shoved to the back. You’re saying, “let me do the important stuff first and add my cherry on top at the end.” No. Build a decent, interesting, weird world first and then layer the blocks on top. That’s your foundation.

So go and record yourself making weird noises, turn them into something alive, and let the track grow out of that.

Happy composing! Rich