In yesterday's Tech Tip (see TTOTD 10/31/03) we talked about the mic modeling built in to Roland V-Studio machines, which raised some questions that will surely come if we don't answer them. They are: Does mic modeling really work? Can one really turn a C3000 in to something like a U87?

There are some related questions, such as, "Isn't modeling nothing more than fancy EQ?" and "If mic modeling can make an SM58 sound like a large diaphragm condenser mic then why do people still spend thousands on such mics?"

These are all valid and important questions, and make no mistake: there are some very strong opinions out there on these matters. Some professional engineers will tell you that mic modeling is a joke; that it doesn't work. Others do claim it is just EQ, or some users just assume it is, especially when they first hear it in action. Then, on the other hand, there are some very reputable manufacturers who are known for excellent products with high quality engineering making some big claims about modeling. What are we to believe?

It's not uncommon for some people to draw cynical conclusions about manufacturers. "Of course Roland would say mic modeling is the greatest thing since sliced bread. They want to sell product." The other side of this same coin is that there are an awful lot of engineers and producers out there who've sunk tens of thousands of dollars into exotic mic collections, and you can be sure they aren't going to outwardly buy in to anything that may threaten their way of thinking, not to mention the dollar value they place on their collections. Or, perhaps, they just believe it doesn't work because it sounds too good to be true.

The fact is there is some validity to all of these points of view. Mic modeling technology is certainly much more than EQ, even in it's simpler forms. Although, even if it was "just" EQ there is still some value in having very precise EQ's set up to do very specific things like make one mic sound like a different mic. That said, the degree to which these models sound authentic is going to depend a lot on how they are applied, and to some extent how discerning the listener is.

For clarity, let's pick on some obvious ways in which modeling is going to come up short. Let's say you set up an SM57 to record an acoustic guitar. You want it to sound like an omni directional small diaphragm condenser. Can a mic modeler do this? In a word, no. The SM57 has a cardioid pickup pattern. The modeler can't synthesize the sounds not picked up from off axis, so the result is going to certainly sound different with regards to off axis response (presumably room ambience in this case). The SM57 is also a dynamic mic (moving coil specifically). Due to the relatively high mass compared to a small diaphragm condenser some transient information is bound to be lost, as in not captured. Again, the technology of modeling hasn't reached a point where it can synthesize this information and put it back into the recording.

Those are some easy and dramatic examples of how modeling can fall short. There are many more subtle issues. But before we jump to the conclusion that it's "no good" one should take a listen sometime. The things that mic modeling can do are pretty incredible. In the example above, the tonality of the guitar in the recording could, in a general way, take on some of the characteristics of the other mic. Will it truly make a C3000 sound like a U87? Not entirely. But it is capable of fundamentally changing the sound of a microphone in ways that are difficult or impossible to recreate using conventional technology such as EQ and compression. And in many cases it can make one microphone come very close to the sound of another — close enough that big, golden ear experts have been fooled. It does mean that an engineer can take a given microphone and fundamentally change the sound so that the "essence" of some other model's character is reproduced. The end result depends on how far you're trying to push it, and on the exact application's circumstances. As shown above, it might be hard to get an SM57 to sound convincingly like an Earthworks mic. However, it may not be as much of a stretch to get something like a C414 to sound similar to a U87, or vice versa. Success will depend, to some extent, on the source, microphone distance and positioning, and so forth.

Alas, all of this somewhat misses the point, in my humble opinion. It seems somewhat pointless to try to exactly model another mic, unless you just happen to know that's exactly the sound you need. It would be much more useful to model whatever happens to sound good. And that's where this technology can really shine. Mic modeling gives you access to many more sounds than you have, regardless of how many mics you own. Through experimentation you can find settings that will accentuate the good sounding aspects of a track while deemphasizing things that aren't as flattering. Ultimately that's what it's all about.

Whether mic modeling is really for you is probably debatable and can be discussed with your Sweetwater Sales Engineer. If you have the budget to buy 10 or 15 awesome microphones you probably wont use modeling as much, and at the end of the day you'll probably be able to get better results overall…assuming you know what you're doing and/or are willing to experiment with different mics and techniques. For users who have just enough budget for one or two decent microphones, or who either don't have the time or skills for productive experimentation it is nice to know that you can get a pretty wide range of sounds by using mic modeling. There aren't a lot of parameters to tweak. It's pretty much a process of dialing up presets until you find the right fit. Also keep in mind that modeling is a technology that's still very young — it's getting better all the time.

For a little more background on this topic you can read the TTOTD from 4/24/2001.