How To Make A Clap In Serum

In this post, I’m going to explain the process used to create 100% original claps that fill out the frequency spectrum, and sound great.

If you want to know how to make a clap in Serum, you’re in luck! In this tutorial, we take it a step further to really create a nice thick clap, ready to spice up your next deep house tune.

The way you should approach sound design, especially for things like drums, is to break the sound up into layers. It’s far easier to create each layer separately as they are usually many simple sounds, combined to create the final, more complex sound we are after.

When building claps I like to divide the process into 4 main parts.

  1. Initial transient(s) – Synthesized
  2. Body and tail – Synthesized
  3. Real hand claps – Recorded
  4. Assembly, Layering and Finalizing

Initial Transients

For this stage I’d like us to look at a 909 clap waveform and analyze the waveform.

What do you see?

If you follow the waveform you will see 4 transients, each with slightly varied amplitude and decay lengths, unevenly spaced apart. What we want to do is to re-create this shape. To achieve this we are going to use an LFO inside Xfer Serum. The main sound source we will be using is White Noise.

At this point, I would suggest hopping into your DAW, opening Serum and following along.

  1. Set LFO one to Envelope mode, and toggle off the BPM switch so we can control the exact frequency of the LFO. This gives us a lot more control over the length of the clap.
  2. Visually try to re-create the 909 shape using the LFO editor (double click to add and remove points). Try to create slight variations in curves and don’t place points perfectly on the grid. Remember, we want to “humanize” the sound as much as possible.
  3. Turn off OSC 1 and set the Noise OSC to “BrightWhite”. Assign LFO 1 to NOISE LEVEL.

I then resample the output to a new channel and compare the waveforms.

Adjust the LFO rate until the transients match up with the 909 clap.

The result looks like this:

And sounds like this:

Ok, we now have the transients in the right place but this still just sounds like white noise, not very useful!

We need to get the transients sounding like actual hands, well, as close to real hands as possible. For this, we are going to use an EQ and some Tube distortion.

The frequency range that gives claps their character, in my experience, tends to be around 800-1000 Hz. When these frequencies get boosted they tend to give the “SSHH” character we associate with a clap.

What I like to do is have a bell curve boost the frequencies up around 1000 Hz at the top of the transient and slide down to 800ish Hz at the base. 

  1. Go to the FX tab in Serum, and turn on the EQ. 
  2. Set the EQ on the right to a bell curve and turn down the Q so it isn’t so narrow.
  3. Assign LFO 1 to the cutoff, set the cutoff to around 800Hz and the modulation amount to around 1000 Hz.
  4. Assign LFO 1 to the gain and have it boost the gain slightly so that the top of the transients are a bit louder.
  5. Finalize it with some Tube distortion.

Great! Now our white noise is starting to sound more like a Clap.

I would then tweak Noise pitch, try adding chorus and other effects, change the shape of the LFO, and just resample 4-5 different sounding transients. We want as much choice as possible in preparation for the final step.

Let’s move on to the stage 2…

Body and Tail

These are combined into one step because I don’t tend to make the body and tail separately.
The way I think about a body/tail is really just a swell of white noise that we shape to give it a nice thick, full character. Think about creating a clap, but with no transients.
Open another instance of Serum and turn off OSC 1. Enable Noise OSC and set it to “BrightWhite”.
For this part, we are trying to make what I described earlier as the “SSHH” sound of a clap.

  1. Set LFO1 to ENV mode and turn off BPM rate, just as we did in when we made the transients. Assign LFO 1 to NOISE LEVEL.
  2. Draw the shape of the “SSHH” sound (see image below).
  3. Adjust the Rate to taste, 8.0Hz worked for me.
  4. Go to the FX tab and enable EQ, boost around 800Hz. This should give you a nice, beefy sounding body.
  5. Add a low cut to the Filter, assign LFO 1 to the frequency and have the modulation pushing the frequency down so that the filter opens up at the top of the amplitude and thins out the tail at the base.
  6. I like to add some stereo width with a ping-pong delay with very short delay times, you could also use a chorus or dimension expander. (optional)
  7. Finalize with some Tube distortion.
“SSHH” sound shape (See point 2 above)
FX panel including Distortion, Delay and EQ

As per stage one, try creating a few different variations of the tail, use different noise types, add reverb, try different types of distortion. The idea is to have a nice range of building blocks for us to mix and match when we create our final clap.

This is how my body/tail sounds:

Real Hand Claps

This stage is fairly simple. Use whatever mic you have, if you don’t have a mic, use your phone. Record a bunch of claps in different spaces and rooms, under a blanket for a really dry sound or in an empty room for a reverbed sound.

Drop the recording into your daw and find 3-5 different sounding claps that you like.

Here are mine:

I added some short HAAS/Pingpong delay for width. Again, process them differently so we have some variation to play with. 

Now, on to the final stage where we get to put everything together and finally make some damn claps!

Assembly, Layering and Finalizing

In this stage, we play Legos and mix and match all our variations together.

Here I’ve lined up a Transient from Stage 1, Tail from Stage 2 and a couple of the Hand claps from Stage 3 in a way that sounds good to me, and placed them in a Group/Bus so that they can be processed together.

Sounds alright… Let’s process the bus a bit.

I added an EQ to cut out the lows as we don’t need them. I then boosted around 1kHz to give it more character. Added a glue compressor to compress everything together, followed by a reverb with the dry/wet automated to add a little reverb tail, and finally added some OTT to bring the mids out a bit.

You can see my processing chain and the reverb automaton lane in the screenshot below.

The result:

And that’s it! 

From here I might try layer in some foley sounds, or add in a different tail. Really, just get creative and experiment!

We have more articles on drum design in serum, make sure to check them out.