If there's one dream most of us, musicians and engineers alike, take a pleasure in mulling over and escorting through the corridors of fantasy, it's surely the concept of setting up your own studio. Next to worldwide fame and fortune, that is. It's hardly surprising that images of tranquil, timeless sessions appeal to the musician. Likewise the opportunity to install, experiment with, and operate selected equipment in a tailored environment is a salivating thought to the engineer. But can such a dream ever become a reality? And how difficult is it to set up?

Well, it's here that you have to be clear about the complexity of your proposed setup. Approaching a studio project on an organic growth basis, where each expansion of facilities is achieved by tacking on extra gear and wiring, makes for a messy, unreliable and ultimately expensive installation. Such a studio is a sadly common phenomenon with working bands where capital has been dribbled into the project without foresight or co-ordination. The more professional approach is to design a system that could feasibly be afforded in the future, and install the wiring and connections which would ultimately handle such a system. An important factor that frequently inhibits such an approach is ignorance — users might, for example, discover the value of talkback after they've done all the hard wiring and so have to tack it on awkwardly later.

In a vain attempt to reduce this sort of mess I've written this article, throwing in the odd diagram of basic interconnections necessary for small studio set-ups. Fig 1 shows a studio termination box for 8-track balanced mic recording giving two mono foldback options over six sets of phones and a mono talkback PA. Total interconnection with control room requirement would be 8-pair individually screened multicore cable, stereo (figure-8) screened foldback cable and one talkback screened cable plus an odd wire for the recording light. By making the talkback link a figure-8, this low voltage dc power could be sent up the spare arm, thus neatening interconnection. Balanced mic lines have earth returns summed at the mixer — summing earths this late and at one point obviates hum looping. Of course, refinements could be made, such as: individual can level controls; stereo foldback to cans; a selection of switchable foldback lines; provision for foldback on monitors; and reverse talkback from studio to control room. But I've just shown a basic arrangement. Having said that, some would argue that the inclusion of a recording light is superfluous. I would agree it's not essential, but it can be very useful, particularly where visual contact between studio and control room is poor or non-existent.

And so the mother multicore and its infant cables wend their way through partitions, up walls or between rafters (more about the structural aspect later), until they reach the vibrant, magic world of the control room. It's here that standards of interconnection are given every opportunity of reaching an all-time low, mainly because the manic forces of enthusiasm tend to force you round every last short cut, and compromise your natural standards of neatness. Also some people aren't very good at wiring. The main thing is to stay cool and reject the 'lash-up' approach. Combining the patience of a saint with the standards of a perfectionist has never been easy, and terminating the balanced mic multicore provides a good example.

In some 'installations' I've seen component cables soldered straight on to Cannon XLR or jack connectors, no termination box having been used. The die-cast boxes so abundant in electronic component shops are ideal for terminating cables, entering by grommet or cable clamp and being wired on to tag strips or screw connector blocks. The anchored box provides a solid, screened base to which patch-bay and console connections can be easily made.

See fig 2 for a schematic control room configuration following our example of an 8-track installation. You will notice that signal lines have been extensively brought up on patch panels, with the exception of foldback and monitor options, which are almost invariably selectable on the desk. In practice some patch provision in the diagram may be duplicated by console switches, reducing the patch-panel requirement.

I shall now proceed to bore some readers (who may well be bored already) by briefly discussing the rudiments of patching. After all, the word appears a lot and I know for sure that some people don't fully understand what it means or, more important, how to incorporate such facilities into a system.

Eggheads may sneer, but in fact many nasty fundamental problems like hum, radio interference and noise can ultimately, after hours of channel card swapping, repositioning of power supplies and other checks, be routed back to patch panel faults. To start the ball rolling I'll describe the two methods of patch panel connection. One is the straightforward 'normally open' method — this allows balanced line working throughout and is totally 'graphic' in so far as inter-connection of equipment can be checked at a glance by simply tracing patchcord routes. The second is the 'normally closed' method where a pre-determined routing configuration is in operation until patchcords are inserted. Balanced working is not possible, but cord links are kept to a minimum and in that respect make for a tidier system. The principle is that each jackfield socket represents not just an input or an output, but a circuit. By inserting a jack plug, connection between input and output arms of the circuit is interrupted by switch contacts and replaced by contact with the inserted plug's tip and ring. This is a 'send and return' arrangement, in fact, signal being sent down the tip and returned on the ring. Sockets used in this way are called 'break jacks.'

Greatest flexibility of interconnection is achieved by using patchcords with connections reversed at one end, so that any output circuit arm can be patched to any input, the other solution being to use ordinary leads throughout, but reverse connections on the equipment socket instead of the leads. With this system, you can keep track of connections by remembering that a signal is sent from the desk to the equipment down the tip of the jack, and is returned from the equipment to the desk down the ring. This method obviates any confusion with leads: you just remember to connect outboard gear in reverse. But there's no accepted standard practice here. The only convention is to colour code phase-reversal cords with yellow (a British PO standard) to avoid accidental phase reversal when patching balanced sources. Fig 3 shows a standard ¼in PO B-type jackfield socket and plug, fig 4 a normally open connection and fig 5 a normally closed connection. Next to each connection diagram are sketches indicating the way in which each system is used. Note that on 'equipment' sockets in fig 5 the wired link between switch contacts used on 'circuit' sockets is replaced by a link shorting equipment inputs to earth when not in use. This can also be incorporated in 'equipment' sockets on the normally open connection, shown in fig 4.

There are a couple of practical drawbacks to the 'normally closed' method which you may or may not consider serious. Apart from having to remember the configuration of the jackfield when empty (which isn't so difficult once you get used to it) and being denied balanced working, such a system involves double the wiring and fiddly things like shared earth tags. But the worst disadvantage is the dependence on switch contacts for normal (empty panel) use. A duff contact here can mean big trouble, sometimes cured by frantic spraying with cleaner/lubricant aerosols and equally frantic jack plug actuations. If such methods fail, you may have to remove the socket by undoing a single nut from behind the panel with a socket screwdriver — rather a tedious operation.

Mic lines should always be balanced where possible, but in a small studio set-up, balancing line sources can be expensive and futile in relation to the improvement in signal transmission, provided you aren't next door to a taxi service, electricity sub-station or policy transmitter (or any other strong electrical field). But it is important to standardise levels and impedances if you are to retain complete patching freedom.

So we'll start with impedance matching. Signal transmission theory tells us that the impedances at either end of a transmission line should be equal for maximum efficiency of power transfer. An impedance mismatch results in a signal being 'reflected back' to the source impedance; power is lost. It's important to observe this when terminating microphones whose operating level is low, but the rule doesn't work out too well for line sources where patching is involved. This is because more than one input is frequently connected to a single output; when input and output impedances are the same, attenuation occurs.

Consider the standard line impedance of 600 ohms in this example. Maximum power transfer would occur initially between a 600 ohm input and output. But if another 600 ohm load was patched into the circuit, the level would drop by 3dB. Further patching would incrementally load the output towards a short condition which, apart from reducing level, could damage the source equipment. So, more typically, a 'bridging' arrangement is used for line levels. Here the 600 ohm output looks at a much higher input impedance, say 10K ohms. Power is lost; but a number of inputs can be patched on to the same output, and the lost power is fairly insignificant at line level.

Next, the question of level matching. Again the objective is to make interconnection between equipment as easy and instinctive as possible. So you need to line up all output levels and input sensitivities to the same figure, typically 0dBm or 1mW into 600 ohms (0.775 volts into 600 ohms also amounts to the same thing). Inputs are aligned with a 0dBm 1KHz test tone commonly available on mixing consoles; outputs are aligned with the same tone having passed through the equipment concerned. Having said that, some equipment is not designed for 0dBm use, (for example, Teac/Tascam gear) in which case it is quite permissible to line up to a lower level, provided that all equipment is lined up to the same level. So much for patching and interconnection.

Two of the requirements of recording most overlooked in the small studio sector are acoustic treatment and sound insulation. A lot of people confuse the two; they make a stab at acoustic treatment and consider they've somehow dealt with sound insulation at the same time. In fact, a clear distinction exists between them. Acoustic treatment is concerned with tailoring room response to shape and, ultimately, subjective preference. Common objectives are the reduction of standing waves (the frequencies of which are a function of room volume), and the controlling of room reflections across the audible range to produce a natural, even reverb characteristic. That's a delicate, subtle process. Sound insulation is a heavy, more systematic affair concerned with controlling sound entry and exit to and from the studio.

It's an unfortunate but inescapable fact of science that the sound insulating efficiency of a medium is inseparably proportional to its density. So the layers upon sandwiched layers of polystyrene and glass fibre so popular with budding studio managers, on which their future career and reputation precariously hang, are all in vain. Quite evidently sound insulation is a much misunderstood business. It can come as something of a shock when the realisation of density hits home. After all, the heavy materials required can be expensive and indeed, in some situations, impossible to install. By now you may well be wondering what materials are used. Well, a lot depends on practical, economic and (just to complete the cliche) social considerations. The first two are obvious enough, while the latter is simply a metaphor for neighbours. A good start is achieved by using cavity-walled premises. If you've got the run of the place, you can then consider applying a plasterboard, such as Gyproc, and then face this with more plaster. Composition materials such as Stellite can also be used.

You can then attach acoustic tiles either at random or all over, to absorb mostly mid to high frequencies in the room (the acoustic treatment bit). These tiles are made of compounded wood fibre, their front-facing or 'active' surface being punctured by a random selection of holes. A more restricted form of absorber is regularly perforated hardwood backed by mineral or glass-fibre. The widely available 'pegboard' has a hole density of about 17% and is effective at high frequencies only, but a more useful range of absorption is achieved with 25% perforated hardboard. Another method entirely is to build, next to plastered walls, a wood batten matrix at a depth of about 4-6 ins. Into the resulting vertical boxes is packed a mineral fibre such as Rockwool — glassfibre can be used, but tends to be less effective, mainly because of its uniformity and ultimately glossy fibre surface. The entire batten structure is then covered with chicken wire or similar, to stop the fibre falling out, and finished with hessian. This type of absorber really sucks sound out of a studio and it may be prudent to counter this by randomly scattering a number of (say) polythene panels across the surface to retain a certain proportion of reflected sound. In fact, if you don't, problems can arise with musicians unable to hear themselves. This inevitably results in amplified gear being turned up and undermining the sound insulation, as well as feeling uncomfortable. Polythene is a good material because it doesn't throw everything back — just the important frequencies.

Provision for a naturally 'bright' sound is useful and probably the cheapest way to incorporate this is by tiling a corner. Large studios often use mirror glass (the mirror property is incidental — the glass surface does the work) or marble to achieve the same effect. If you've got a few square feet of mirror kicking around, or a few slabs of marble you no longer use, fine; if not, stick to glazed kitchen tiles.

Conventionally, the ceilings and floors in a small studio are treated so as to reflect as little as possible, the wall finishes being adjusted primarily to tailor reverb requirements. You may, however, prefer the particularly live sound that goes with vertical sound reflection. Things to watch, though, are messy hi-hats and an uncanny sensitivity to foot tapping and shuffling. If you want to sample the phenomenon, just visit the local church hall where the right (or should I say wrong) conditions invariably exist. Floor reflections can be minimised with a good thick pile carpet, but the prerequisite to floorbound noise suppression is a reasonably solid floor. Freely springing floors tend to boom at the bass end and transmit foot taps direct to mic stands — paving stones or concrete floors as found in garages are ideal. The more domestic wood or vinyl tile finishes are also acceptable, provided they sit on a firm base.

Treating ceilings can be an expensive business. One well known studio designer incorporates a 6ft false ceiling filled with suspended Rockwool as part of his standard design. A little more applicable to the small studio world is the by now familiar notion of acoustic tiling.

Acoustic screens can be useful in a small environment to increase instrument separation. An idea for such a screen is shown in fig 6. As you can see, the materials quoted are quite commonly available, the board coming in 8 x 4ft sheets, the glass fibre in rolls from a roof insulation dealer. Thick fabric could be used to cover the hardboard, improving appearance and further reducing sound reflection. Also plasterboard, though upping the cost, could be incorporated to increase the screen's sound-stopping power. The main things to remember in selecting materials are the essential components of an absorber, shown alongside the diagram.

Finally, drapes are useful — particularly in preventing undue reflection from windows (or any other reflective surface). Their great advantage is that they can be drawn open or closed to suit. The heavy fabric required is expensive new, but old cinema curtains (seriously — they are replaced from time to time) or other stage drops are ideal. A certain amount of heavy old-fashioned curtaining regularly finds its way into jumble/garage sales and the like.

Designing a control room and its acoustics is a rather subjective area. Generally though, it's logical to follow studio acoustics on the premise that they represent your subjectively ideal acoustic environment (or close to it). Most engineers would prefer less reverb in control rooms than studios to get 'up front with the sound' — that's only the tip of the jargonistic iceberg now alive and living in control rooms everywhere. A thick drape or even a blanket on the wall behind monitor speakers absorbs a lot of stray sound from the back of monitor cones and so is a first step towards a more intimate sound.

But what worries most people setting up a small studio is where to put the control room in the first place. The main consideration here, apart from structural and disturbance factors, is whether or not visual contact with the studio is essential. If you feel it is, then prepare for the decorated tranquillity of your intermediate wall to be shattered to let you have your way with the proverbial double glass. Having visited many double glassed big-time studios, however, I've been struck by the casual and occasional use of the feature. In fact, some engineers tell me they prefer to record with their eyes blankly focused on middle distance, without any distraction from musicians. Once you've accustomed yourself to the idea of 'blind' recording, it can work quite well and gives a lot more scope in the siting of the control room. One last thing about control rooms, minor though it may seem, is to make sure everybody can sit down when it comes to a replay. Most musicians are on their feet when playing; let them listen in comfort and they're likely to be more relaxed and objective about the mix.

So, broadly speaking, you now have the makings of a small studio. You've all heard the one about the musician who spends so much on building a studio that it has to be hired out — well, that may be a trap to avoid. Alternatively, it might appeal to you as a deliberate strategy, and if this is the case watch out for a future article on the subject of making studios pay. In the meantime, good luck.