Serum-like sounds in SunVox

Post Reply
User avatar
AutumnCheney
Posts: 34
Joined: Sun Dec 29, 2019 8:16 am
Location: usa
Contact:

Serum-like sounds in SunVox

Post by AutumnCheney » Tue May 26, 2020 9:41 pm

Abstract

If you're a producer working in electronic music (especially in the genres of dubstep and drum and bass), then you may have heard of a program called Serum. Developed by Xfer Inc., it is a wavetable synthesiser with many built in effects and the ability to make many different modulations and envelopes between parameter values. It is extremely powerful and is used many producers working in the harder genres of EDM to create screeching, overpowering sounds and timbres. However, many aspiring producers believe that you must use Serum and only Serum to create these types of sounds, and this view is incorrect; there are many synths capable of making these sounds using the same or similar techniques as Serum, and I will show you how to implement these ideas in SunVox.

Wavetable-style synthesis in SunVox

Serum uses a method of digital sound synthesis called wavetable synthesis. It works by taking an array of various, usually somehow related, waveforms or samples and switching between these sounds during playback. It can create a variety of morphing, dynamic sounds. In Serum, you can create new wavetables by taking a static sound and changing it over a period of time; Serum can sample this and turn it into a wavetable. Alternatively, Serum can also interpolate between two or more sounds to create a new wavetable.

Although SunVox does not yet have the capability to perform true wavetable synthesis or interpolate between waves, it is possible to approximate this by fading between two or more modules playing some sound using volume adjustment and a custom MultiCtl curve (this can be seen in the "crossfader" project file included in the "simple_examples" folder in most SunVox versions). It should be possible to crossfade three or more modules, but I find that two modules is enough for almost all sounds.

To emulate Serum's recording capability, you can record any sound into the Sampler module and then play back an arbitrary short snippet using a granular synthesis technique devised by forum user Saltbearer, which they've described in this post: https://warmplace.ru/forum/viewtopic.php?f=20&t=5413.

Serum's crazy waveforms

By default, Serum comes with a variety of wavetables to use, and you can import others into Serum. These waveforms tend to be very chaotic, with a variety of irregular shapes and curves. These give them a variety of harmonics and timbres, which make them very useful when creating harsh sounds.

SunVox can easily create these same waveforms using several methods. One powerful method is the "drawn" or "drawn spline" waveforms in the Generator and Analog Generator modules, which allow you to draw your own waveforms. You can also use the WaveShaper module to heavily alter the wave and give it unique harmonics and shapes.

Serum's "unison" feature can also be recreated in SunVox using multiple MultiSynth modules with differing "Finetune" values; my "Supersaw" module does this using seven regularly spaced voices.

Modulation

Serum can implement several types of modulation between its two oscillators, including common ones such as AM (amplitude modulation) and FM (frequency modulation); FM can create very harsh sounds, especially when used with complex waveforms.

SunVox also has modulation capabilities in the form of the Modulator module, which can impliment AM, PM (phase modulation), and absolute PM (phase modulation with the absolute value of the modulator wave). Phase modulation is very similar to frequency modulation, except with some minor limitations: waves with only or mostly sharp edges and corners, such as the saw or pulse waves, do not have a significant effect when using PM. Instead, PM uses the derivative of these waves, and the derivative waveforms of these types of waveforms are basically silent DC signals. I won't explain too much about this, since it involves a lot of off-topic mathematics (you can do research if you're curious), but just know that smoother waves work best for PM; the best wave to use, in my opinion, is the "drawn spline" mode on the Analog Generator, as it can produce complex waves with smooth edges that work best with PM. You can also apply a light lowpass filter to any other wave to slightly smooth it out; I'll explain more about filtering later.

Envelopes

Serum allows you to define envelopes for all of the internal parameters in the synth, and you can adjust how much the envelopes affect the values.

Making envelopes in SunVox is fairly straightforward; you can simply use a flat DC signal (which can be produced by the "drawn" modes on the Generator and Analog Generator) and the Sound2Ctrl module, which translates current into a controller value. You can create simple ASR (Attack, Sustain, Release) envelopes using these modules, but you can also create much more complex envelopes using the Sampler module, which allows you to draw an envelope. You can change the envelope amount using the "OUT min" and "OUT max" controllers on the Sound2Ctrl; making the "OUT min" value larger than the "OUT max" value will invert the envelope.

The envelopes are most smooth and accurate when the "sample rate" and "smooth" controllers on the Sound2Ctrl are highest and lowest, respectively. However, these configurations cause the module to require slightly more CPU power, so make these changes lightly, depending on your system.

Filters

Serum also comes with a variety of filters, including simple ones such as the lowpass and highpass filters, as well as more complex ones. A simple technique commonly used by professional sound designers using Serum is modulating the cutoff values, as well as using a high resonance (often labelled "Q").

SunVox comes with two filter modules: the simple Filter module, and the more powerful, 64-bit Filter Pro module. The Filter Pro module also comes with the ability to visualise the filter cutoff, which can be very helpful when customising the module. Complex filter types can be created by using multiple filters in a chain and using the MultiCtl module to change any needed values shared between the modules with one controller. Some filter types included with Serum, like the "comb filter", can be created using a Flanger module with a static LFO (i.e., the "LFO amp" value is set to 0).

Effects

Usually, the final step in making a sound with Serum is applying effects. Serum comes with a variety of effects to change your sound, and they can easily make your sound much more full and complete.

SunVox can also replicate these same effects. Some, like Serum's Flanger and Delay, are relatively straightforward to replicate. Others, like the Phaser and Chorus, are not easily replicated with SunVox's native modules. However, it is still possible to build MetaModules with these capabilities, and you should be able to find the effect you're looking for in the "SunVox Modules" page of this forum. as well as the custom effects included with most SunVox versions. If you have the know-how, you can also build them yourself.

Conclusion

These are some general recommendations for making Serum-like sounds in SunVox! I hope that they help you! If you have any questions or recommendations, feel free to reply to this topic.

I have also included a few projects showcasing examples of these methods; feel free to study them and use them in your own songs.

NOTE: Example 2 uses my "Triple Oscillator" module. which uses three oscillators and two built-in PM modulators.
Attachments
AutumnC Serum-like Sound Example 2.sunvox
(32.67 KiB) Downloaded 35 times
AutumnC Serum-like Sound Example 1.sunvox
(23.68 KiB) Downloaded 38 times
please check out my newgrounds account, i post music and talk to people here: https://autumncheney.newgrounds.com

Post Reply