Editor's Note: Audiophile Style community member George Graves has kindly allowed us to publish his five part series on high quality audio recording. This series is a primer that many audiophiles will find interesting and educational. It isn't a treatise, textbook, or master class designed to cover every detail in depth. As a music lover and audiophile I want to understand a bit more about recording, but I don't want to become a recording engineer. This series is right in my wheelhouse, and I hope it's in yours as well. - CC
Part One: Commercial Recording Quality (link)
Part Two: Recording Media (link)
Part Three: Microphone Mixers and Pre-amps
This time we are going to discuss how to get the microphone signal to the recording device. For this discussion, we are going to assume a two-channel stereo recording, but, keep in mind that these things apply to multi-channel recordings as well (whether they are mixed to two-channels for production or to 5.1 channels or any other surround-sound format.)
Professional quality microphones, whether condenser, or dynamic have one parameter in common: they all need to be amplified before they can produce a recordable signal. Usually, microphones need between 40dB (100X amplification) and 60dB (1000X amplification) to get their signals up to what is regarded as "line level" which is required to give maximum record volume. Some microphones, such as ribbon mikes have such low output that sometimes as much as 70 or even 80dB of gain is required. This means that pre-amps required for microphones must have the following characteristics: They must be quiet, have low distortion, lots of headroom and wide, flat frequency response. They also must have differential inputs to take professional "balanced" microphone cables.
Noise, is a major factor here. At these levels of amplification, the "self noise" generated by active components such as bipolar transistors, FETs, Integrated Circuit (IC) operational-amplifiers (op-amps) and vacuum tubes becomes a major factor in the quality of the finished recording. Much of this self-noise is thermal. Active electronic components use the physics of electron attraction and repulsion to move signals around and to amplify them. Moving electrons through the devices creates a certain amount of random, or non-correlated noise (along with heat). In fact, vacuum tubes (or valves as they are sometimes called) operate by heating an element inside the tube hot enough to actually "boil" electrons off of its surface. The rest of the tube's elements (the grid(s)) control the flow of those electrons and actually determine how many of those boiled electrons get passed it and on to the plate where they constitute the amount of current that the tube conducts; I.E., less current flow represents small signals, more current flow represents larger signals. Since all of the tube's amplifying ability comes from the fact that it is a heat-operated device, the amount of random electron flow, and thus noise, is characteristically quite high.
Solid state devices work differently, and while they still create heat it is much less than a tube, and rather than the heat being the method of operation for these devices, it's more of a by-product with them. While noise is still an issue, generally speaking, solid-state electronics are much quieter than tubes. Does this mean that tubed electronics cannot be used to make modern digital recordings? Not at all. The reason is because there is another way to get voltage amplification of a microphone signal; transformers. Voltage gain, in a transformer, is largely a matter of the turns ratio of the transformer's coils between the "input" (the primary coil) side of the transformer and the "output" side (the secondary coil). As an example, if you put 10 turns on the primary windings (coil) of a transformer, and 100 turns on that transformer's secondary windings and apply a one-volt AC signal to the primary, you will get about 10 volts out of the secondary. This is an oversimplification, but it does show how the voltage gain of a transformer is determined by its turns ratio. Since the transformer is not an active component, it adds no noise, but it will amplify any noise in the signal applied to it right along with that signal. There is no free lunch, after all. So if we take the output of a good mike with decent noise characteristics, feed it into a transformer before applying it to an active microphone preamp, it is possible to get by with far less gain in the preamp itself. Less gain equals less thermal noise being added to the signal making it very feasible to use tubes in modern microphone preamps and still get signal-to-noise ratios that are compatible with even high-resolution digital recordings. Of course, there is a downside with transformers. Good ones, which have flat frequency response across the entire audio spectrum are expensive. Transformers also have problems with maintaining phase integrity at all frequencies and especially have problems coupling low frequencies through them. Most modern solid-state microphone pre-amps and mixers don't use transformers, but use a type of input circuitry called a differential amplifier. These are very good at rejecting noise that is common to both legs of the balanced interconnects from the microphones. These include, hum, air conditioning spikes (when the compressor cycles), noise created by the proximity of light dimmers in auditoriums and other public venues, etc. This ability is called common-mode rejection.
These days, a decent microphone preamp will give signal to noise ratios of somewhere in the region of about -125 to -130dB, the greater the number, the quieter. Any microphone preamp with figures in this region will give recordings that are, for all intents and purposes, essentially silent. Even with the playback gain set quite high, one should hear nothing but blackness when no instruments are playing.
Another characteristic of microphone preamps to consider is headroom. Live music can get quite loud. Cheap preamps can clip (distort) when fed a microphone level that's too high. Modern condenser mikes can handle sound pressure levels of as much as 150dB before clipping. It would seem like it would be nice to have a microphone preamp that had similar characteristics. Thankfully this isn't necessary. All microphone inputs have controls on them to vary the amplifier's gain and most have a light on them to indicate when that mike channel is clipping. A rule of thumb here is that the louder the source is playing. the less gain is required from the pre-amp. It is always advisable to ask the ensemble's leader to have them play the loudest part of their program before the recording starts so that you can advance the gain to clipping, and then back-off until the clipping indicator light goes out for all microphones. Then, back-off a bit more - just in case somebody wasn't really playing their loudest. This insures the highest possible signal-to-noise ratio without worry of overdriving the microphone preamps. Some recording engineers use a calibration box on each input to feed a signal of known amplitude into each microphone preamp. This might work in a studio situation where things like room-loading and individual microphone characteristics are well known, but for location recordings, I'd rely on the actual musicians to tell me how loud they're going to play rather than count on some unrelated "standard". After all, conditions will vary from venue to venue and musical group to musical group. No two situations will ever be the same. Of course, if you are in a position to record the same group in the same venue time after time, experience will guide you in setting your microphone preamp gain. Ultimately, the amount of overload protection built-into one's microphone preamps is down to their design. The higher the power supply voltage feeding the amplifiers, the more head-room they will have. On modern mixers, even fairly inexpensive ones, this shouldn't be a problem.
Many mixers contain, for each input, a group of controls called "EQ" or equalization controls. These are essentially, "tone controls". Usually there are at least three and sometimes more. They are usually marked "high", "mid" and "low". Sometimes the frequency at which they come into effect is also marked on the mixer and sometimes that frequency has its own control and can be varied somewhat. There is usually a set of these for each input on the mixer. If one is doing an 8 or a 16-track recording where the mix will be finalized at a later date and every instrument or instrument group has been assigned it's own microphone channel, then I can see where such controls would be very useful. On the other hand, most of the types of mixers that are used for location recording are "X" number of channels in but only two channels out and are designed for mixing on-the-fly while the performance being captured is actually occurring. So, no matter how many mikes you are using, the end result is two channels recorded to media, and that is cast in stone. There is no going back and "tweaking" this mike feed or that one. Since there is no way to "undo" an injudiciously applied amount of EQ in these cases, I tend not to use it. The exception would be if I had a certain microphone that was deficient in some way (like a ribbon mike that had little response above about 10 KHz) I might use a bit of EQ on that channel to accentuate the area that was a little lacking. It is also possible to judiciously add a little presence to a vocalist by lifting the midrange a bit, or to reduce the chestiness in a male vocalist's voice by reducing the bass on his mike's input. Other than that, I feel that it's best to leave these controls out of the picture. They're great to have when you need them as long as you keep in mind that a little goes a long way, and the results are not reversible.
Each microphone mixer or pre-amp has its own features such as built-in reverb effects, or busses for external effects and it is beyond the scope of this article to discuss them. But what I do think is necessary is to talk about the size of mixer needed for location recording.
When I got back into recording after a long hiatus (see the first installment of this blog entitled "Commercial Recording Quality"), I figured that since all of my earlier recordings had been made with mostly two microphones, and when confronted with a chorus as well as an orchestra, a maximum of four, that a four microphone input mixer would more than suffice.
Behringer 1202 Mixer sports four excellent microphone inputs and 8 line-level inputs. These can be had for less than $120
When I bought the above pictured Behringer 1202, I was astonished by the street price of just a hair over $100 (US). My previous mixer, a TAPCO, had cost about $1200 and wasn't anywhere near as good. Behringer calls their microphone preamps in this line of mixers "Xenyx" pre-amps and they tout them as being very quiet. They are. I have made some astonishing recordings with this mixer. The circuitry sounds so good and is so quiet that instruments just "appear" out of a velvet black background. I realize that manufacturing this mixer in China (from a German design) is part of the reason for the low-cost, and the advancements in solid-state technology is responsible for the rest, but still, I was blown away by the quality. Soon I realized, however, that the types of recordings that I was doing required more than just four microphone inputs. Not the least reason was because I was using a stereo microphone in the M-S pattern (which will be discussed in another installment), and that required three microphone preamps to yield two channels. That left one. Realizing that I needed more microphone inputs, I went back to the Behringer catalogue and found the 1832FX. This mixer sported 6 microphone inputs as well as built-in reverb effects. I hesitated due to the size of this mixer (one more thing to carry). It is much larger than my 1202. But I figured that where the 4-input 1202 was sufficient, I could cary that, and where I needed more, I could carry the 1832FX.
The Behringer 1832FX Mixer has 6 of Behringer's excellent "Xenyx" series microphone pre-amps and 8 line level inputs. It also sports built-in digital special effects and a graphic equalizer on the outputs. The street price on this mixer is around $250.
While I have chosen Behringer mixers, that doesn't mean that there aren't others just as good, and while I find the Behringers excellent performers and suburb values, Mixers from Peavy, Mackie, Allen & Heath, Edirol and Yamaha are probably just as good. Choose according your projected needs and keep in mind that you will have to tote around whatever mixer that you eventually choose. If you do find, at some point down the road, that you need more microphone channels than your current mixer can provide, that there is another alternative to buying a whole new mixer.
Add-On Microphone Preamps
Most mixers on the market today come equipped with a certain number of microphone preamp stages. In the case of the Behringer 1202, that number is four, and with the 1832FX it's six. The Peavy PV20, for instance, is close in price to the Behringer 1832FX and offers sixteen microphone stages but lacks the Behringer's comprehensive features list. It is also immense. Many of these same mixers also have a number of line-level inputs. While these, lacking the gain, are not designed for microphones, but rather for other sources such as recorders (for mixing-in pre-recorded material and adding to it), and even other mixers. My Behringer 1202 and 1832FX both have 8 such inputs and other mixers may have similar. This is a perfect application for outboard microphone pre-amps. These devices can be had for as little as about $35 for a single-channel tubed unit from Behringer up to several thousand dollars. SM Pro Audio sells an excellent 4-channel solid-state microphone pre-amp called the Q-Pre4 which is available for a street price of less than $80. For my purposes, the Behringer, again, proved to have the most bang for the buck. Behringer's MIC100 is a solid-state, stand-alone, single-channel microphone preamp with a tube output buffer to impart "the tube sound" to the microphone being fed it. The 12AX7 vacuum tube used is not for gain and therefore adds no appreciable noise. You will, of course, need one for each extra microphone you connect. For either Behringer mixer, that means eight in total. I carry two in my recording kit and have even used them in place of an entire mixer when only two channels are required! They sound excellent and would, in fact, constitute a fine starter system. A pair of MC100s, a pair of decent big-capsule cardioid condenser mikes such as the Samson C01 at less than $80 each (street price) or a pair of Behringer Pro 2Cs (multi-pattern mikes) at slightly more along with a Zoom H2 solid-state recorder and you will be recording 24-bit, 96KHz stereo recordings for a a maximum investment of only about $500! This kind of price/performance combo would have been unheard of just a few short years ago.
The Behringer MIC100 Tube buffered Microphone preamp. At $35 street, it's excellent and hard to beat for flexibility and control
The SM Pro Q-Pro4 4-channel Microphone preamp. at around $80, it would be hard to go wrong having one of these in one's kit
Next time we'll discuss Microphone types and how to deploy them...