Serum 2 Spectral Mode: A Producer's Deep Dive, the Dankest New Feature

Serum 2 is here and it's amazing. So many new features, so much to talk about. Based off all the questions y'all had, I wanted to talk about one of the most exciting new features for me, which is spectral synthesis, and the unique way we can now use Serum to get access to it. It feels like a really fresh type of synthesis. Even though it's been around for a while, now that it's in Serum it's easily accessible and totally worth talking about.

One of the things we like to stress at Futureproof is why things work, not just how to do them. If you can understand a concept at its core, then you'll be able to get better on your own and become the amazing music producer you were always meant to be. So let's get right into it.

Watch the Video

Prefer reading? Here's the full breakdown.

What you're actually looking at

The first step is to change from wavetable to spectral mode. There's nothing in here out of the gate, and you can select samples from the drop-down menu if you want: bass, brass, flute, and more. These are all worth exploring, but I want to bring in my own sample, because that's the way most people are going to use it and it's also a really easy way to explain it. I'm going to bring in a vocal run.

Spectral synthesis is the concept of breaking a sound down into individual frequencies, or partials. Normally when you're looking at an audio file, you're looking at what's called the time domain, which is how a sound progresses over time. We have that in spectral mode, but now we've moved into what's called a frequency domain, and that's just how this is visually represented. We can generate a bunch of unique tones by manipulating these frequencies.

Here's how to read the display:

• Vertically (Y-axis) it's like an EQ. Low end on the bottom, high end on top. You can use the fader knob to select what part of the frequencies you're using spectral synthesis on.
• Horizontally (X-axis) shows how the sample is played, left to right.
• The darker the graph, the less there is of that frequency. As you remove the densely populated white color, you start to lose the fundamental of the sound, which could be cool.

You get access to unison, which gives you stereo width and detuning. You also get one shot, forward loop, reverse loop, forward-reverse loop, and tailed, all the things you get in sampler mode. The blue lines represent the part of the loop that's playing, and the little X next to the loop mode is the fades on the loop itself. Drag it up and you can get fades in any direction you want.

Get it in key first

One of the things I like to do whenever I drag a tonal sample into an oscillator is make sure it's set to C, because if I add in oscillator B or C, I want these to play in key, or at least play a harmony. So if this is an F note, to bring it to C we go to minus 5 semitones. Now if I turn on oscillator B, we should be in key, and we can layer them and make a harmony if we want, or just keep it as a root note. This is super important.

The scan knob

The next important thing is the scan knob, located right below one shot. Scan determines the speed and direction of playback. At 12 o'clock it's stagnant, not progressing through the sample at all, which can be cool. Move it to the right and the further you go, the faster it plays the sample. Go to the left and it plays backwards. Having scan set to the middle almost has a granular effect, which I really like. At the very least, you can play through samples and scan through them in a really unique way.

The filter

If you click on filter, you can edit an actual filter that's scanning through the spectral wavetable. To create a low-pass filter, just click and drag, kind of like what you do in the LFO view. And you could definitely use an LFO to modulate this, because this is a cutoff knob. This is an interesting way to create movement, and we all know movement is really good for sound design. Sounds without movement are stagnant. Stagnant can be good for something like a pad, but I love to have movement in my sounds, even if it's very subtle.

There's also a mix knob, which is just the dry/wet for the filter. A really cool thing to do in sound design is to have something very aggressive, like distortion or in this case a filter, and pull the dry/wet back. That way you're able to mix it in almost like it's in parallel. A really cool effect that might normally be way too aggressive for your sound, but pulling the mix knob back lets you slide it in there and make things smooth and nice.

Warping is where the dank stuff happens

We're in Serum, which means we get access to warping, and it's just like the next level of spectral synthesis. This is where the dank stuff actually starts to happen, because it's going to become something totally different. The spectral warp modes are unique in Serum 2 but also work really well with spectral mode:

• Detune +/-: visually it's literally changing it, and we immediately lose our tonality, which is okay if you're doing a different type of sound design.
• Bend +/-: stretches it from the middle, pulling the middle up into high end. It's almost like a pitching effect.
• Smear: as you turn it up, the fundamentals start to disappear. It's almost like a tape stop effect.
• Spread: really warps it around. A lot of these sound like we're frequency shifting around.
• Harmonics and subharmonics, and comb.

Here's the fun part. Take a metallic snare drum, drop it in, and scan through it. This works so well in electronic music for getting these metallic, strange sounds. It might not logically make too much sense, but I smell a new meta coming on. Add my holy trinity of effects and we're starting to get a bass sound out of a snare drum. Yeah, sounds like dubstep to me. If you're a tear-out producer, putting a snare drum in here and combining it with the convolution reverb, you will be cooking so quickly and so easily. With distortion and a comb filter, we're turning a vocal chop almost into a flute.

The technique: swap the source

Experimenting with different source samples is definitely the tech here. What I would do is create a really cool sound, a sound you like a lot where you're doing some weird stuff, then save the patch, duplicate it, and just swap out the source sample and see what happens. You're going to get some really unique results.

A key element to spectral synthesis is keeping an open ear for possible new additions and new results. You might go in thinking "I'm going to make this type of sound," and yeah, you could totally do that. It's like the difference between sound design on purpose and sound design for fun, where you're just hoping for random results. Spectral mode is a really great way to do both. If you brought a bass sound into spectral mode you could probably make a pretty rad bass, but I love that we can bring a snare drum in and still make a pretty rad bass. You can do the same with a vocal or whatever, because now you can play it at any octave you want, scan through it any way you want, and loop it any way you want. Very creative, lots of flexibility.

Spectral mode is dope.