Can 'modeling' microphones deliver on their copycat promise?

The humble microphone. For many, it’s little more than a tool for a Zoom call. For streamers and podcasters it’s a utilitarian bit of kit. But for some — singers, musicians and other recording artists — it’s an instrument as characteristic and expressive as any guitar or piano. The microphone is often the biggest investment these performers ever make. If you want a different “timbre” to your recordings, then, you could just buy a lot of microphones. Alternatively, you might consider a “modeling” mic - one specifically designed to imitate the character of famous (and usually expensive) models.

For those not wanting to spend “small condo” levels of money on a well-stocked mic locker, something like the $1,000 Sphere LX from Universal Audio offers a tantalizing prospect. It’s one of the aforementioned modeling mics, and it claims to mimic 20 of the most sought after microphones of all time.

The Sphere LX itself is a condenser microphone, but thanks to its dual-capsule setup (most mics only have one) it’s able to do some pretty funky stuff. Not only can it pretend to be a dynamic or ribbon mic - both of which use different technology to a condenser - it can imitate environmental factors such as distance from the mic and/or different recording spaces. You can even blend two different virtual microphones at the same time for added versatility. It all sounds very clever on paper, but does it actually sound, well… good?

Before we get to that, it’s worth understanding the setup required. As the Sphere LX has two capsules (the “sensor” part) it outputs in stereo unlike most vocal microphones that record in mono. This means it ships with a specific XLR splitter cable that requires two ports on your audio interface. So, if your setup currently only has one, you’d need to find a new one with multiple inputs (Shameless plug: I wrote an audio interface buyer's guide that you can read right here).

There are some important steps to take before you can start experimenting with your new virtual microphones. First, if possible, you’ll want to “bind” the two input channels on your interface so that they operate as one. This locks their gain levels so you don’t need to worry about the physical volume controls being different (which could impact the effectiveness of the modeling).

It’s also important to remove any processing your interface might apply. For example, Universal Audio’s own Volt 276 interface has hardware compression and EQ which you don’t want applied here. Other interfaces also can apply light processing or compression by default, so you will definitely want to check your interface’s settings. Sometimes such processing requires turning off via software - so watch out for that, too.

With the hardware set, you’ll want to open the Sphere’s companion software. It’s a VST plugin and therefore runs within other software — Ableton Live, Logic Pro or even Garageband will do. Drop the plugin onto the same audio channel assigned to the Sphere LX’s output and you’re set to record. Actually, you can even record first and then add the software later, as the emulation can be applied to any recorded audio (but only recordings from the Sphere will “match” the emulated mic).

In the most simple use case, you would record your vocal with the Sphere LX and then choose the microphone you want it to sound like in the software. Then all you need to do is export the audio and you’re done. You can even go back, change the emulation to another microphone and export it a second time, but why stop there?

One of the key benefits of the dual-capsule system in the Sphere LX over, say, something like Slate’s M1 VMS system, is that it allows for things like changing the direction you were addressing the “mic” after the fact. If you recorded into the Sphere head on, but wanted a slightly off-axis sound (useful for taming high frequencies from a guitar for example), you can do that in the software after the fact. You can even change the polar pattern - the shape/area around the capsule in which sound is received.

While this feels a bit like witchcraft — changing the physical qualities of a microphone after something has been recorded — it serves both practical and aesthetic purposes. On the practical side, you can theoretically record once and “try out” different microphones, mic placements and polar patterns. This not only saves time with the vocalist or musician, you won’t need to keep moving gear around, perfect if you have a smaller space.

Engadget · Universal Audio Sphere LX mic modeling examples

In practice, unless you have all of these classic microphones for comparison, it’s obviously hard to know how close the Sphere LX comes to the originals. I happen to have three of the microphones that are modeled by the system - Sennheiser’s MD421, Neumann’s TLM103 and Shure’s SM7B - although I use them primarily for spoken word, podcasting and streaming.

On simple tests with voice recordings, the Sphere LX comes really close for all three. There are definitely differences, but given that two units of the same mic can develop variations from each other over time, the LX likely falls within those differential boundaries. This is further reinforced by listening back to the raw audio captured by the Sphere which is very, very far from what it sounds like with emulation applied.

I was particularly interested in how well the Sphere LX would emulate the SM7B and the Sennheiser MD421, as these are both dynamic microphones. Condenser microphones work very differently, so the idea that one could imitate the other was interesting. Condenser mics are generally favored in vocal studios as they capture more detail, but dynamic mics are better for those with less than ideal recording conditions. Being able to flip between the two with one mic would be both convenient and impressive.

The MD421 in particular has a unique character for a dynamic microphone with a surprisingly detailed, lively sound. With a straight voice test, the Sphere LX doesn’t quite capture those trademark “sparkle” frequencies, but it definitely does a good job on generally sounding like the Sennheiser — and it’s distinct from its imitation of the SM7B (also a dynamic mic), but with more emphasis on the lower frequencies.

Engadget · Modeling mic versus modeling software

While the comparison shows a decent facsimile of the Sennheiser’s sound, it’s worth noting that both the MD421 and the SM7B have a physical highpass filters built in that can be adjusted through different settings of bass roll off. This isn’t something that’s replicated in the Sphere’s companion app, so if you wanted any sound other than their default, you would need to apply some post processing.

There are some other challenges, too. In almost every other recording situation, once the track is laid down, the take is set in stone. You can manipulate it after the fact, of course, but the take itself is immovable. With the Sphere LX, and similar systems, you can reimagine the recording the moment the singer’s lips stop moving. Flicking through each virtual mic and the related setting until you find the one you like the most. This could obviously be a good thing, but option paralysis is the death of many a good audio project.

The appeal of these systems could also depend on your budget and whether you feel confident that they will remain supported by the manufacturer. This is particularly prudent for professionals. “I find with anything like that you're into ‘what's it going to be worth in the next, you know, five years?’ Or when they decide to make the mic redundant, which I’ve found on products that involve anything software.” Chris Denman, CEO of Skyrocket Audio and professional sound engineer told Engadget. The redundancy isn’t a huge risk, unlike many app-dependent gadgets, as VSTs can live outside of an app store, but something to consider.

If you’re curious about microphone modeling but don’t want to jump right into a hardware system, there are software products that claim to do a similar thing without the $1,000 outlay. Something like IK Multimedia’s “Mic Room” plugin, for example, offers a similar roster of mics for well under $100. To get the best results, your current mic will need to be one the app has a reference for, so there’s possibly an additional spend, and the results aren’t as accurate or as configurable as the Sphere LX, but it’s an easy way to experiment before committing to something like Universal Audio’s solution.

All in all, whether mic modeling is right for you will come down to personal preference, economics and the projects you need it for. For casual users looking for a do-it-all mic, it’s perhaps a little complex and costly. However, for singers looking for a streamlined, versatile setup, it almost feels like a no brainer. There will be plenty of use cases that fall between and either way, it’s a really interesting concept that will likely only continue to become more accessible.