Illustration: Michael Haddad
You’ve probably heard countless reasons why analog is “better” than digital.
Countless musicians and sound engineers claim that analog gear just sounds “fatter,” “warmer,” or “thicker” compared to digital’s tinny, cold thinness. While that’s not always true, there are reasons for these claims and misconceptions. A lot of the digital synths that became famous in the 1980s, like the infamous Yamaha DX-7, sound a bit thin and maybe even cheesy in retrospect. But at the time, studios were selling and throwing away analog synths to make room for the futuristic sounds of digital synthesis. Some analog gear certainly might have a warmth and fuzz that most digital gear doesn’t, but it’s far from impossible to find warmth in digital sounds. One of the best examples of this gray area can be found in the world of analog emulation.
Almost every piece of classic gear has an emulation that’s a fraction of the price of the original hardware. It can be hard to even tell what hardware is worth owning these days, and it can be even harder to tell how these digital devices differ from their analog counterparts. Let’s talk about the primary ways analog emulation plugins are created, how they differ from each other, and how they can differ from the gear they’re trying to recreate.
How analog emulation plugins are made
There are two main ways to go about recreating hardware in the digital realm. First, you could reconstruct the circuit from scratch digitally. By taking apart the device and making notes of how each piece of circuitry is configured, it’s possible to build a digital device that functions the same way as the analog gear you’re trying to recreate. This is the more straightforward method; if you know exactly how each piece of the hardware works, it’s easy to make a corresponding piece of software.
The other method is developing a Digital Signal Processor, or DSP. By sending a variety of audio test signals through the device and recording the inputs and outputs, you can write code to emulate the effect of the hardware. If the hardware you’re emulating is a synthesizer, rather than an FX unit like a compressor or reverb, you’d be creating sounds with it instead of processing test signals, of course. This method gives a lot of creative freedom to the designers, and often, engineers can create a shockingly similar device that arrives at the same types of sounds in a completely different way than the original does.
Both of these work quite well – before we dive into their pros and cons, it’s important to note that each method can create incredibly accurate results and give the original hardware a run for its money. That being said, there are some key differences in the results you might get from each.
The challenges of analog emulation
Certain pieces of hardware are much more difficult to emulate with DSPs than others, largely depending on how many possible control settings there are. Often, that key difference lies in the separation between stepped controls and continuous controls. Stepped controls only allow a setting to rest at a finite number of positions, whereas continuous controls allow for a nearly infinite amount of possibilities. For example, it’s much easier to emulate a compressor with a ratio that can only be set to 2:1, 4:1, or 10:1 than a compressor that can have a ratio at any value between 1:1 and 10:1.
This doesn’t mean that continuous controls can’t be emulated by DSPs, but it can take much more time and dedication to achieve a convincing emulation. Hence, the wider the variety of control settings you test, the more accurate your emulation is going to sound. Because of how time consuming this can become, devices with lots of continuous controls can be much easier to emulate by taking apart and recreating circuitry rather than coding DSPs.
Alternatively, recreated circuitry will often sound a bit “too perfect” to make a convincing emulation clone. Modeling individual pieces of circuitry isn’t always as difficult as creating a DSP, but this method can sometimes miss the unpredictable interaction between circuitry that makes analog gear lovable. Some DSPs manage to capture a few of these quirks in their own way, but, at the end of the day, no software emulation is going to have quite the same type of unpredictability that analog gear does. Software plugins all function completely separately from each other, whereas hardware can interact with nearby hardware, creating subtle distortion and modulations that can be quite palatable. Such audio artifacts, like those from drifting, shifting, and deterioration, are very difficult to model in the digital realm.
Another major downside of recreating circuitry is the limits of computer processing; especially with more complex devices, it can take most (if not all) of your computer’s CPU to perform the way that analog devices perform with ease. By creating a DSP, your computer might be able to use more efficient methods to achieve similar effects.
On the flip side, when creating DSPs, unwanted audio artifacts and issues can arise from within the test signals themselves. This can sometimes make it difficult to tell which sounds are coming from the gear itself, and which are just a result of a faulty test signal. Because of this, engineers need to use many test signals to create a convincing DSP, requiring even more time for the completion of a worthwhile clone. But, this shouldn’t completely deter you from DSPs; after all, more time spent is going to make a more authentic emulation no matter what method you choose.
The last thing we’re going to consider is creativity. Software is far more fluid than hardware could ever be; the possibilities are nearly endless in what you can achieve with code. Some engineers opt to create DSPs solely because recreating circuitry gives you so little creative control. When building a DSP, you have a massive palette of options and all sorts of choices to make; in comparison, recreated circuitry can feel a bit like a color-by-number.
To summarize, DSPs are often preferred because of the creative control you have when designing them, and because they often allow you to find CPU-saving alternatives to the original circuitry that will run smoother on computers. But, the more possible control settings and the wider array of sounds you’ll be processing, the more difficult a convincing DSP will be to create, making recreated circuitry a much easier path to take for certain gear. And this duality isn’t even cut and dry – some engineers even use a combination of both methods.
At the end of the day, digital emulations of analog gear sound better than they ever have before, and they’re improving every year. Without an awful lot of experience with analog gear, you probably won’t be able to tell most modern emulations apart from their counterparts. That isn’t to say that analog gear isn’t inspiring and exciting, but if you’re tight on cash, don’t be deterred from software. And if you want to play around with some analog synth emulations of your own, check out KORG’s Collection 2 or Arturia’s V Collection 8 on Rent-to-Own.
Do you have any questions about analog emulation? Let us know in the comments below.
January 26, 2021