I recently read a question in a Facebook group. It was from someone who didn’t have a great deal of experience recording, and as is so often the case, it is usually the people who know the least who ask the best questions. He asked, “Why do people who have spent upwards of $2,000 on studio condenser microphones have to EQ them to get them sounding right?”
It’s a great question. The obvious answer is that recordings made with individual microphones rarely exist in isolation and to manage conflicts between the various overlapping elements, which make up a busy mix it’s usually desirable to tweak the spectral content of individual mics. However, we all know that most microphones aren’t flat and exactly how their coloured response presents itself is complex and varies from application to application. After all, if the flattest mic was the best mic, we’d all use that one wouldn’t we?
It has to be said that when someone is EQing a mic they aren’t really EQing the mic, they are EQing the sound the mic is capturing, a subtle but important difference. The sound arriving at the mic has already been coloured by its environment in all sorts of ways, many of them good, some of them, less good.
The flattest mics are rarely the most popular or coveted mics. There is a reason measurement microphones are invariably omnidirectional condensers with small diaphragms but if you consider the most coveted studio microphones they tend to be distinctly coloured large-diaphragm condensers, often featuring valves and when a choice of polar pattern is available they usually spend the majority of their time in cardioid, which is an inherently coloured polar pattern, then it starts to look like flatness isn’t a priority.
Why does this colouration occur? Is it the reason these mics are favoured or is it a price worth paying to access other, more desirable aspects of studio microphones?
Angle And Distance From The Microphone
I’ve become more and more interested in the contribution angle and distance make to the response of microphones. It’s frequently overlooked by users but unless you are recording in isolation, from exactly on axis, in an exceptionally dry room, the contribution of off axis sound will be present. Spill from other performers, reflected sound and reverberance or even the off axis sound from other parts of a large instrument like a piano all contribute. Not all many sounds are a point source. But ‘not flat” doesn’t necessarily equate with “bad”.
In his blog post Why Directional Mics are Inherently Colored Chris Townsend of Townsend Labs says that:
“Most recording engineers are not used to hearing flat microphones, and flat might sound boring or maybe even slightly odd. Even though it might theoretically be more accurate, in the end, it’s a subjective artistic choice depending on many factors such as the musical genre, the creative vision, and the sound sources themselves.”
Taking the example of a good cardioid condenser, the transformers, valves and deliberate voicing of the mic are obvious sources of colour. But after these have been discounted there are still major sources of colour inherent to the design, principally the proximity effect and off-axis colour due to the cardioid operating principle. Taking these in turn:
Proximity Effect And Off-Axis Response
The familiar bass lift we hear when directional mics are used from up close is one of these sources of colouration, the flattest response a cardioid mic occurs from a specified distance but we usually don’t know what that distance is. Luckily the bass lift caused by moderate proximity is reasonably easy to adjust for. However, the second issue is rather more complex and more difficult to fix.
The Townsend Labs Sphere is best known for its modelling of exactly the kind of desirable coloured mics mentioned earlier but a consequence of the modelling system is that the Sphere can model “hypothetical” microphones. Chris Townsend again:
“While the Sphere microphone is designed to model other microphones, it can also create new kinds of microphones that don’t exist in reality. One of these new types is the “Sphere Linear” microphone model, located under the “Custom” mic type category.
Just like any large diaphragm condenser, the frequency response of the Sphere microphone is not perfectly flat, but it is possible to apply DSP processing with a compensation filter so you end up with a nearly flat response. The Sphere Linear model by default is flat at a distance of about 50cm, but by enabling Off-Axis Correction it's possible to have a flat response any distance. Off-Axis Correction has an On-Axis Distance control and an Off-Axis Distance control. When using the Sphere Linear model the On-Axis Distance control adjusts at what distance the on-axis frequency response will be flat. The Off-Axis Distance control adjusts at what distance the polar pattern will have the most off-axis rejection.
Of course, it’s not unusual to record multiple sources with one mic, such as an orchestra. This might result in only some instruments having a flat response, although once the sources are more than about four meters the distance doesn’t matter that much. Like a zoom camera lens, you can set the focal point to infinity as long as the sources are far enough away.”
Off-Axis Colouration
We all use cardioid mics and most of the time don’t put much thought into how they do what they do. The fact that the intricate arrangement of carefully designed openings of different lengths which shift the phase of sounds arriving from the rear work at all is kind of amazing so the fact that they have undesirable side effects has to be forgiven.
All mics sound different depending on the direction from which the sound arrives. In the best omni mics this difference is negligible but cardioid mics suffer quite badly. To test this out just connect a cardioid mic, monitor it through headphones and just make a hissing sound through your teeth. Rotate the mic in front of your mouth and hear how dramatically the sound changes. Record the results with a few different mics (tip - speak the mic model at the beginning of each recording, saves potential confusion later!). You’ll hear that some mics perform very differently to others off axis.
The Sphere Linear Diffuse model uses the off-axis response, which is so important to the authenticity of the mic modelling that the Towsend Labs system is so well known for, to push the off-axis response of the mic as close as possible to flat, which is a lot closer than can be achieved in a conventional mic:
“Sphere also goes a significant step further by flattening the off-axis response. Although it’s impossible to perfectly flatten the response simultaneously in all directions, it is possible to produce a considerably more ideal polar response.
Another Custom microphone model is "Sphere Linear Diffuse" which provides a flatter response for sound coming from all directions, such as reverberation. This can be particularly useful for room miking or other techniques where the majority of sound is off-axis.”
While using a very flat off-axis response would be useful if recording in a particularly live space or in the far-field to capture a drum room mic or similar, these applications could arguably be fulfilled using a conventional Omni mic, with corresponding changes to placement. Something which absolutely couldn’t be achieved using Omni mics is to flatten the off-axis response in a particular direction to address one of my long-standing issues with most coincident stereo mic techniques. Chris Townsend again:
“If desired, it's even possible to adjust the Sphere Linear model so that it's flat in a particular off-axis direction… useful when using a single Sphere microphone to record in stereo because the dominant sound sources are likely to be off-axis.
For example, if the microphone is 45 degrees off-axis to the dominant source then setting the Axis control to 45 degrees will make the microphone flat at 45 degrees off-axis. Keep in mind this does mean that the microphone will now not be as flat on-axis. Depending on the pattern this axis correction works up to about 90 degrees, and best results are obtained in the +/-45 degree range. Beyond 90 degrees the Axis control provides a moderate high-frequency roll off.”
Using conventional microphones I’ve always favoured mid/side arrays over XY pairs of cardioids precisely because in an XY pair the centre of the stereo pair, usually where the most important information is coming from, is off-axis to both mics. In an M/S pair, this is not the case. The linear models, while lacking the superstar appeal of the vintage models, uses the technology used to facilitate these models to address some of the issues and limitations of mics which are designed to be clean and neutral, allowing them to be even cleaner and more neutral - Clever stuff! To find out more about the off-axis modelling of the Townsend Labs Sphere and how it is used in the Linear models read Why Directional Mics Are Inherently Coloured at Townsend Labs.