Exploring the Sonic Frontier: The Power and Potential of Physical Modeling Synthesis

Despite the myriad capabilities of synthesizers, recreating the subtle variations found in acoustic sounds remains surprisingly challenging. This issue has been prevalent since the dawn of synthesizers. Even Wendy Carlos, who popularized the Moog synth, stopped using it because she couldn't achieve the variation she desired.

Consider this: if you open your DAW, start with a basic patch, and play the same note multiple times, you'll get an identical output each time. Real-life sounds don't behave this way. Whether you're strumming an open guitar string or snapping your fingers, you'll notice a wealth of variation in the timbre, intensity, and the duration between each note.

Various methods exist to manage this issue, like modulating synth parameters over time using LFOs or automation. However, matching the level and subtlety of variation found in acoustic sounds remains arduous. This is where physical modeling shines - it can introduce a ton of subtle variation from note to note without having to modulate any parameters at all.

Physical modeling synthesis achieves this by simulating the physics of sound production in real instruments or objects. For example, it models the impact of a mallet on a marimba bar, the vibrations in the wood, the resulting sound in the air, and even the subtle variations in these processes. This allows physical modeling synths to create a plethora of different sounds with a high degree of realism and expressiveness, especially in emulating acoustic instruments.

Furthermore, physical modeling synths like Collision or Chromophone allow you to map MIDI velocity and pitch to various parameters. This means that not only will each note have variation depending on its velocity, but also the timbre can change significantly across the different registers of the keyboard. This feature is invaluable when trying to achieve more organic and expressive sounds.

With physical modeling synths, you can explore the "in-between" areas that traditional synths may struggle with. You can adjust parameters to create hybrid sounds that, for example, start as a struck string sound and morph into something more like a plucked membrane.

This is not to say that physical modeling synthesis is a panacea. While it excels in some areas, it may fall short in others. For instance, some physical models can sound less convincing at extreme settings, and others may have a steep learning curve, especially for beginners. Yet, the depth, expressiveness, and variation it offers are well worth the effort.

So, why should you care? Because physical modeling synthesis expands the boundaries of your sound palette. It can breathe new life into your productions, allow you to create more realistic and nuanced sounds, and push your creativity to new heights. Embracing this technology could just be your next step in advancing your sound design skills and broadening your musical horizons.

How Does It Work?

Physical modeling synthesis distinguishes itself from other types of synthesis by using a different approach to sound production. Most synthesizers adopt a signal-based approach where the primary goal is to reproduce the output signal of an instrument. For example, one can recreate the sound of a marimba using an FM synthesizer by modulating one frequency with several others, adding some noise, and adjusting the envelopes accordingly. While this might yield a similar sound to a marimba, the process of frequency modulation bears no resemblance to the physical process that generates sound in a real marimba. It is merely an imitation of the final result, ignoring the nuances of the physical interaction that makes every real instrument unique.

Physical modeling synthesis, conversely, operates on a completely different principle. It uses mathematical models to simulate the physical behavior of real-world objects when they produce sound. Instead of trying to mimic the output signal of an instrument, it tries to mimic the actual physical process that gives rise to the sound. This could be a string being plucked, a piece of wood being struck, or air being blown into a tube.

The modeling process takes into account a range of variables. These can include the manner in which the object is interacted with, such as the strength and location of a string pluck or drum strike, the characteristics of the object itself, such as its material, size, or shape, and the environment in which the object is placed, like the characteristics of the surrounding air or resonating bodies.

This means that the synthesizer doesn't just produce a static sound. Instead, it generates a dynamic, evolving sound that changes over time, much like how a real instrument behaves. For example, when you pluck a virtual string in a physical modeling synth, you're not just getting a static sound sample of a string being plucked. Instead, you're getting the sound of a string vibrating back and forth, with all the subtle variation and complexity that this entails.

What's even more exciting is that because this process is entirely virtual, you aren't bound by the laws of physics. You can create imaginary instruments that could never exist in the real world. You can have a string that changes its thickness over time, or a drum that shrinks when struck harder. The possibilities are nearly endless.

The reason physical modeling synthesis isn't more widespread despite these advantages is largely because it functions differently from other, more familiar types of synthesis. It requires a different mindset and approach to sound design. However, once you understand its principles, it's an incredibly intuitive and powerful way to create a wide range of interesting and natural-sounding sounds.

To fully appreciate and understand the power and flexibility of physical modeling synthesis, we will use several different synthesizers in our tutorial. Each one will help demonstrate different facets of this fascinating technique.

Once you understand the core principles, the specific synth you use becomes less critical. With a firm grasp of the fundamentals, you can explore, experiment, and create using any synth that supports physical modeling. Whether you're working with Collision in Ableton Live, Sculpture in Logic Pro, or Chromaphone by Applied Acoustic Systems, you can apply the same principles and techniques.

Towards the end of this tutorial, we will explore a variety of these synthesizers to give you a sense of the range and diversity available. Each has its unique quirks and capabilities, and by familiarizing yourself with a range of options, you can find the tools that best fit your creative vision and workflow.

Now, let's delve into the hands-on part of the tutorial and start exploring the magic of physical modeling synthesis!

Tutorial

For the demonstration, we'll use Collision for Ableton Live, a product of collaboration with Applied Acoustic Systems. This synth is a favorite of mine due to its user-friendly interface and exceptional velocity and key tracking capabilities. It enables effortless mapping of MIDI pitch and velocity to the parameters of the synth, which is vital for achieving a diversified sound.

Collision includes a resonator section where you choose the physical model, and an exciter section that represents the imaginary mallet or the filtered noise used to stimulate the model. By manipulating each of these sections, you can create unique sounds that defy the constraints of traditional synthesis.

There are other remarkable synths out there that incorporate physical modeling. One such example is Kyvo from Madrona Labs, which introduces a granular oscillator as an exciter. Applied Acoustic Systems' Chromophone is another example, serving as a lightweight alternative to Collision. Then there's Substantia, which stands out by using a sample as the exciter.

In conclusion, physical modeling synthesis opens up a wealth of opportunities for diversifying your sound palette and achieving more natural and expressive sounds. With this technique, you can explore new sound design territories and break free from the limitations of conventional synthesis. While it may not be the be-all and end-all solution, it's certainly a valuable addition to any sound designer's toolkit.