Recording Techniques (Part 5)
Mixing Console (1)
Part 5: The mixing console is the largest, most complicated, and probably the most expensive item in the recording studio. In this first installment, David Mellor begins his journey through its component parts.
One of my favourite mixing consoles is the Audio Developments Picomixer. It has six channels, two outputs, and is about the size of a briefcase, if somewhat on the chubby side. It is small, but it is not cheap - it costs around £2000, depending on the options you have ordered. Does this sound like a lot of money for a 6-channel mixer? Well, £2000 pounds is by most standards a lot of money, but what you get is a very professional, robust little unit which is precisely suited to its intended function - location recording. If you don't need more than six microphones, but you do need good quality sound and the feeling of confidence that you will never be let down, then the Picomixer is just the thing to have.
The point that I am working up to is that to many people the words 'mixing console' are pretty well synonymous with 'large' and 'complicated'. A mixer does not, in fact, have to be either of these. It has to be able to do a job of work. This job may demand a lot of facilities, or it may demand a smaller number which are exactly appropriate to the task in hand. What I intend to look at here is a typical recording console and see what facilities it has, and to show at each stage how they fit in with other items of studio equipment and with the recording process as a whole.
"It's very impressive, but what does it actually do?" the A&R man asked about the mixing console in an unguarded moment...
The recording process basically consists of taking sounds from a variety of sources, modifying the character of those sounds if necessary, mixing them together in a subjectively pleasing way, and storing them on a permanent medium - usually tape. The mixing console is the control centre for all of these processes. In fact, virtually every signal flowing from one piece of equipment to another in the studio will flow through the console. Figure 1 shows the typical selection of sources and destinations.
Inside the console, the signals do not simply pass from input to output. The essential component of the mixing console is the mix bus. The phrase 'mix bus' derives (just like 'London bus') from the Latin word 'omnibus', meaning 'for all'. In fact, if you think of how a motor bus operates, you will not be far from the mix bus concept: a bus travels along a fixed route from suburban point A to city centre B. In the suburban areas, it picks up passengers from various points along the route. When it reaches the city centre, they all get off and go to work.
The mix bus plies a route as well, from the left-hand end of the console to the right. It starts by picking up various signals from the inputs, and takes them to a destination, which will be an output from the console. In the console, there is only one suburban area (the input section) and all buses pass through it. There will be several outputs, and each bus goes to one output only. Figure 2 shows the signals from four channels flowing into one bus.
The basic mixing console has a number of inputs, known as channels, and outputs, known as groups. A console with 12 channels mixing into two groups would be fine as a PA console; 12 mics could connect to the channel inputs and the two group outputs could feed the left and right loudspeakers via a stereo power amp.
However, a recording console needs two extra sections: the monitor section, which enables you to hear the output of the multitrack tape recorder, and a master section which - among other things - has a switch which lets you hear the main stereo output of the console or alternatively routes the output of the stereo tape recorder directly to the monitor amp and speakers. Figure 3 shows where to look for these sections on a typical console. This orientation of facilities is known as the split monitoring design. The alternative is called in-line, in which the monitor section is combined into the channels. I am going to stick to the Split system for the moment, because it is easier to explain and perhaps a little bit more logical.
The channel is where the input signal is conditioned, so that it is suitable for further processing in the console; it provides equalisation (EQ) to change the tonal character of the signal; finally it routes the signal to one or more buses and controls its level. Let's take a stage-by-stage look...
The first part of the channel is the input stage. This is where the signal first enters the console environment and is brought up to a suitable level for the circuitry to work on. Figure 4 shows the controls involved. Each channel of the console has two input sockets, mic and line. 'Mic' basically means what it says, you plug a mic in here. 'Line' refers to virtually anything else that isn't a microphone. This is, of course, a generalisation that needs further clarification.
Microphones, conventionally, have low output levels. A typical figure would be around one millivolt (1mV) - ie. one thousandth of a volt. The console likes to work on a signal level of around one volt to keep well above the inevitable noise voltages that will be present in the circuitry. This means that the signal from the microphone has to be boosted by 1000 times, or 60dB (a gain of 60dB means exactly the same as 'multiply the voltage by 1000'). Of course, when I say that 1mV comes out of the microphone, that depends on the level of the sound source, its distance from the mic, and also on how the sound level varies. In practice, a mic input needs to have a range of gain from 20dB (10 times) up to 60dB (1000 times). A desirable range is from 0dB (no gain at all - the signal stays the same) to 80dB (10,000 times). This would cover all situations, from the mic being placed inside a bass drum to a watch ticking at 20 paces. There would be no advantage in providing more than 80dB gain for even quieter sounds, because the noise produced by the mic would be amplified above the console's noise level. (More on noise later.)
The mic input on our hypothetical console has a switchable 48 volt phantom power supply.
Pretty well any equipment that is connected to the mains produces a line level output, or near enough - so this is connected to the line input. Ideally, the gain control should still operate when the channel is switched to line input. On some consoles, variable line gain is dispensed with as an inconvenience-causing, cost-cutting measure. A good range of line gain would be - if you could get it - from -20dB (which would actually reduce the signal level by a factor of 10) to +20dB. (One reason why you might like to reduce the signal level is when you want to use only a small amount of the signal on that channel in the mix. If you can reduce the input gain, the fader can be operated at a higher, more convenient, level).
Not everything that produces a signal can be connected to the mic or line inputs. To reduce noise levels, these inputs are normally designed to be fairly low impedance - jargon meaning that the inputs need to see a lot of electrons coming down the cable. Some signal sources just cannot produce a lot of electrons, or a lot of current which amounts to the same thing. The most typical example is the electric guitar. Although a guitar may put out a healthy voltage, it hasn't got the 'oomph' to drive a mixer input directly. In this case, a device called a 'DI (Direct Injection) box' is connected between the guitar and console.
The other switches in the diagram are the pad, filter and phase switches. The pad, in a properly designed console, will cut the level of an input signal by 20dB before it reaches any active device in the circuitry. This is so that if you have a really high level input signal, it can be cut down to size before it has chance to produce any distortion. If a console has a switched gain control, as opposed to one with a continuous travel, the pad may be invisibly incorporated on one of the wafers (sections) of the switch.
In theory, the filter ought to be an adjunct to the pad. As you know, directional microphones produce high levels of low frequency signal when they are used close to the sound source. This can also happen when mics are suspended in a large auditorium. Convection currents in the air create high levels at subsonic frequencies. The mixing console filter, if it comes before any active circuit component such as a transistor or integrated circuit, can deal with these low frequencies. Sometimes the pad and filter come after some of the active circuit elements, in which case they don't work as protection against distortion, they are just accessory level and EQ controls. Figure 5 shows the preferred internal arrangement.
Figure 6 shows the effect of the phase switch. It simply turns the signal upside down. This is useful when one of your microphones, or one of your cables, is wired in reverse - pins 2 and 3 of an XLR connector being swapped, for example. This happens more often than you would imagine and produces horrible frequency cancellation effects. Also, where several mics are being used close together, there may be random phase effects even when they are all wired correctly. It is sometimes helpful to experiment with the phase buttons on your mixing console to see whether the sound improves subjectively.
Going back to Figure 5, you will notice the insert point. The insert point is typically used for connecting a compressor or noise gate. A unit connected here will operate only on this one channel. The insert point on many consoles is in the form of a stereo jack socket, the 'tip' connection being the insert send and the 'ring' connection being the insert return. You will not be surprised to learn that some consoles have it the other way round; there is no fixed standard. When there is no jack plug inserted in the socket, a switch contact passes the signal straight on to the rest of the channel circuitry.
There is another use for the insert point, beside sending the signal through an external effects unit: the insert send can be used by itself as an extra output from the console, but only for the signal on this channel. You would normally only do this if you were already using every other available output.
To use the insert point in Send/Return and Send Only modes, as I shall call them, you need two different types of cable. The first is shown in Figure 7. Notice how two cables must be forced into one jack plug for Send/Return operation. For Send Only, you need to link the tip and ring of the jack with a piece of wire so that the signal passes through the insert as normal (Figure 8). Remember that when you plug a jack into the socket, the internal connection is broken away.
Sometimes the insert point is positioned directly after the input circuitry. This is great for noise gating, because once you have set the gain, the gating threshold doesn't vary. Other consoles place the insert point after the EQ. This is subjectively better for compression, but it makes gating difficult because you have to adjust the gate threshold each time you change the EQ. Some consoles let you have it both ways (at a price!).
After passing through the input circuitry, the signal is big enough and strong enough to be sent through the rest of the circuitry. Its first test of manhood will be the EQ section.
EQ stands for 'equalisation'. It is very strange terminology, because we are not making anything equal to anything else. The word originates from telephone systems and we seem to be stuck with it, at least until someone comes up with something better.
The EQ section of the mixing console's channel is a glorified tone control. It changes the frequency balance of the signal, boosting the highs, cutting the lows, sucking out the middle, adding a presence peak... but let's cut the jargon. Let's look at a typical console EQ section, as shown in Figure 9.
Figure 9 depicts a three band EQ section. The HF control cuts or boosts the level of the signal above a set high frequency. The LF control similarly cuts or boosts the level below a set low frequency. The Mid section has two controls: one which sets the amount of cut or boost, and one which selects the frequency at which this occurs. One possible point of confusion for the newcomer to mixing console operation is the fact that if the Mid Level control is set to its centre position, the Mid Frequency control will have no effect on the sound. This, of course, is entirely logical when you think about it.
An essential feature of any EQ section is the in/out switch. This has two functions. The minor function is so that if you are not using any EQ on that channel, the extra circuitry (which adds noise) can be switched out of the signal path. The major function of the EQ In/Out switch is so that the mixing engineer can assess the difference between the EQ'd and unEQ'd ('flat') signal. It really is very important to know that you are actually improving the sound as you twiddle the knobs. Without this switch, comparisons are difficult and rarely made.
The EQ described above is really pretty rudimentary, although surprisingly few consoles go further than this. Higher up the console price range, you get an EQ that can really do things to the sound. It will look rather like Figure 10. Some console EQs get more complex even than this.
The real advantage of this EQ is in the switched frequency HF and LF controls. Fixed frequency EQ controls may or may not operate on the required part of the audio spectrum for the instrument you are recording. With switched EQ frequencies, you can home in on the part of the sound you want to modify.
The switches shown close to the HF and LF sections are the peak/shelf switches. Figure 11 shows the difference between the two. A peaking response boosts (or cuts) a particular band of frequencies around a centre frequency. The shelving response boosts all frequencies above (in the case of HF) or below (in the case of LF) a set frequency. These two EQ responses offer distinctly different sounds. The shelf response is particularly effective when the slope of the curve where the response changes is fairly steep.
The mixing console is too large and full of possibilities to cover in one installment. So far I'm only half way down one channel! More next month.
Feature by David Mellor
Previous article in this issue:
Next article in this issue: