Variable rooms
Chapter 2 was rather long and convoluted, but it had to be in order to set out the groundwork for the expansion of our design concepts. I described in considerable detail rooms which, due to their limitations, cannot be the ‘all things to all people’ that they were perhaps originally intended to be. Such rooms are excellent for recording jingles, one-off radio broadcasts, and many other recordings which must be cost-effective and of good quality. They are real ‘work horses’. Do not think that I am relegating these rooms to a second division, as I am not. They are, however, not the ideal rooms for many of the ‘big production’ recordings for CD release, which will be listened to countless times, frequently by people with expensive equipment and critical ears, and from which ‘special’ sounds have come to be expected. One of the main failures of large neutral rooms is not that they do not produce good recordings, but that they frequently do not inspire the musicians or help in the production of ‘magic’ sounds. As mentioned in Chapter 1, they were originally conceived at a time when it was the ‘accepted’ view that the engineers recorded just what the musicians played. The studio rooms, themselves, had not yet been fully integrated into the creative process of music recording.
In the late 1960s, such neutral rooms as those described would have been well received by the majority of recording personnel, but around the same time, groups like The Who, Led Zeppelin and the Rolling Stones were beginning to use the old Olympic Studio in Barnes, London, which had a large recording room of distinctly characteristic acoustics. Perhaps for the first time, artistes and producers were beginning to get more general freedom relating to how, where and when they recorded. They were becoming less tied by the instructions of their record companies. The ‘supergroups’ began to emerge, having so much money that they could even start their own record companies, and, not too much later, have their own front-line studios built. It is true that in the late 1950s and early 1960s, visionary producers like Phil Spector in the USA and Joe Meek in the UK were innovative and controversial, but these people were very rare exceptions, and faced much opposition from the ‘establishment’.
3.1 Time for change – musician power
In times gone by, the technical people often seemed to look down on the musicians, considering them pawns in the ‘scientific’ recording process. Given their new found independence, the musicians began to make their point ‘with their feet’, walking out of the studios where they felt uncomfortable, and into the studios where they felt they could play their music to the best of their potential. Time after time, the newly ‘in-vogue’ studios were those which had been frowned upon by the more mainstream acousticians and recording staff. Yet, ironically, it was very often the technical weaknesses of these studios which the musicians found that they could use to their advantage.
Sounds in general were becoming a more important part of the recording process. ‘Records’ were becoming more of a complete creative process. Rather than a studio just recording what the musicians played, they were becoming more of an instrument in themselves. This was somewhat different to what had begun in the late 1950s, when at that stage it was mainly the electronic processes that were beginning to be usurped by the adventurous producers and musicians. The studios were still largely sacrosanct. Studio staff in the more ‘comfortable’ studios were also often of a more flexible nature, with maintenance staff who were a little artistically inclined themselves, or at least who were interested in new ideas and experimenting with novel techniques. It was a long time however before the old attitudes were swept away, partly because there was a certain ‘superiority’ seen in being ‘technically correct’. I am viewing this at the moment from a European perspective, as in the USA, greater emphasis always seems to have existed on the subject of musical acoustics. In the 1960s, many British groups began to go to the USA to get the sounds that they sought, with few, if any, American groups coming to Europe specifically to record, and when they did, it was usually for the European musicians and not the studios.
It was clearly becoming apparent that acoustic conditions, other than musically neutral ones, were being more widely appreciated, but all too frequently the conditions were idiosyncratic. They were great for a certain type of recording, but imposed their character too much on some recordings which did not need that ‘help’. It was thus also becoming clear that a greater degree of acoustic variability was needed. What was more, that variability needed to be something more than the provision of acoustic screens, which at the time were usually on wheels, and perhaps, if one was really lucky, had one side hard and the other side soft. It was also becoming no longer acceptable to enclose the drummer in a box, crudely arranged from screens, solely for the purpose of achieving better separation from the other instruments (and vice versa).
Unfortunately, just as this necessity for acoustic variability was beginning to be more widely appreciated, 16-track recording was beginning to become widespread, followed in 1972 by the first practical 24-track machines from Ampex and MCI. (A Unitrack 24-track machine went into Morgan Studios in London in 1970, but reportedly was never fully operational.) The new vogue word was ‘separation’, and for the first time all the instruments of many groups, and even the individual drums in a drum kit, became recorded on separate tracks. This supposedly offered much more scope for processing in the mixing stage, but inter-instrument leakage from track to track could limit some of the possibilities. In prior times, musical errors would usually require a re-take of the whole backing track, but with multi-track recording, single instrument performances could be ‘repaired’. However, if, for example, a guitar needed to be replaced, then even if the old track was erased, if the original had spilled over into the drum microphones, the unwanted guitar, which may have played out of time in the original, or played wrong notes, would still be audible via the drum mix. The repair of the original guitar would thus only be undetectable if the leakage between microphones was minimal, and that meant either recording in dead spaces or in multiple, isolated rooms.
The next general phase in variability, not surprisingly, was to provide a main recording area, with one or a series of ‘isolation rooms’, usually in good visual contact with the main area. Vocal booths had existed for a long time, but these new isolation rooms were usually larger. In response to the need for more variable acoustics, they were often built with different characters, having varying degrees of life (both acoustically, and before (in many cases) rejection by the musicians).
Musicians rarely like being isolated: it frequently makes it difficult for them to play as one unit. Perhaps the most emphatic of all of the musicians who rejected these ideas were the drummers. Technological advances and the perceived benefits of greater isolation led to the introduction of the disastrous drum booths of the mid-1970s. Engineers and producers saw much possibility in the prospect of more isolated drums and a dead environment, but drummers usually hated playing in such dead surroundings. This will be discussed in more detail in Chapter 5, but suffice it to say here that there was a period of several years of experimenting with multi-track recording and high degrees of separation. Once again, largely due to a revolt by the musicians, the pressure was back on to produce rooms with a good ambience for playing (both visually and acoustically) and where the musicians felt happy, yet where the recording staff could also achieve their aims. Larger rooms began to arrive on the scene with either zones of different acoustics, or movable walls and/or ceiling surfaces which could change the ratio of absorbent, diffusive or reflective surfaces facing the room. From time to time, movable ceiling panels could be raised and lowered, thus being able to change their reflexion patterns from the late to the early type.
Figure 3.1 Recording room with fixed live and dead areas
Figure 3.2 (a) Recording room with graded acoustics. (b) Detail of side wall
Figures 3.1 to 3.3 show three possible ways of achieving acoustic variability in larger spaces. Although the type shown in Fig.3.1 is relatively cheap to construct, and highly effective, it has the drawback of being awkward to use if a large ensemble is in the room, as a uniform acoustic could not be shared by the whole group. Figure 3.4 shows how the room could be used effectively for a typical rock group, allowing the musicians to play as one unit in close contact with each other, yet with each instrument in its own, desired, acoustic space. The results can be excellent, but rarely are large rooms dedicated solely to the recording of a small number of musicians. Unfortunately, if a larger ensemble is placed in such a room, filling more than one half of the space, either the front/back or left/right balance of direct to reverberant sound will not be uniform. It would simply not be possible to achieve a balanced, overall sound. However, in small rooms where the recording of acoustic ensembles was never likely to occur, this technique can be used to good effect. What is more, if a room is large enough, two distinct areas of differing acoustic nature can be very useful. In the area between the two halves, an interesting area can exist where very different recordings can be achieved by the varied positioning of directional microphones, either in the vicinity of the instrument, or more distantly as ambient microphones. In the live end of the studio, walls can be made quite reflective, with angles chosen which will encourage lateral reflexions, and resonances of lowish Q. This area would contain far more reflective surfaces than the neutral room of the previous chapter. The live end of the variable room under discussion here would not be musically neutral, but it would not be intended to be so.
Figure 3.3 (a) Recording room with adjustable acoustic panels. (b) Detail of hinged panel system
Figure 3.4 Typical usage of room shown in Fig. 3.1 by a rock group. In this example, a five-piece group could record ‘live’ using the characteristics of the room to good effect. The drums could be set up in the live area, to produce a full sound and a good feel for the drummer. Bass and guitar amplifiers face the absorbent walls, thus reducing the overspill back to the other instruments. The amplifiers help to shield the microphones from the drums and percussion. The percussionist faces the other musicians, and the percussion microphones do not point in a direction where they are likely to collect excessive overspill, even though the percussion is in a relatively live part of the room. The vocalist, in a dead corner, faces the other musicians, but the directional vocal microphone faces the absorbent surface
We must always give due consideration to how musicians feel in the recording spaces, because if they are not comfortable, they tend not to play so well, and if they are not playing well then it is barely worth recording their performances. Much music is based around instrumentation, and instrumentation is often based around the sounds that they produce in certain spaces. The evolution of music, instruments, performing spaces and human perceptions has been inseparable. Instruments often rely on the fact that they will be surrounded on at least three sides by walls. Side walls produce lateral reflexions, and those reflexions in effect become part of the sound of the instruments. For this reason we cannot, as was found in the high-separation era, take musicians into alien environments, expect them to perform optimally, then hope to add whatever is necessary later. An inspired performance must occur at the instant of the recording.
Certainly for most acoustic instruments, lateral reflexions are expected by the musicians; they are critical to the development of the expected tonal character of the instrument. Furthermore, those reflexions must be lateral; they cannot be adequately substituted by reflexions from other directions. This is one reason why acoustically neutral rooms so often fail to inspire, and why they failed to gain wide appreciation for serious music recording. In a more lively room, we need to provide some reflective wall surfaces to add richness to many instruments, but, on the other hand, for recording of electric bass guitar and amplifier, such surfaces can steal so much of the punch from the sound. What we need to do, to add greater flexibility, is either make the reflective surfaces in some way removable, or confine the bass to a non-reflective area. Lateral reflexions are also a problem as far as separation is concerned, as if we need to collect them in the recording microphones, we may also collect the reflexions from other instruments in the same room. This begins to highlight the need for variability, as what is excellent in one set of circumstances can be detrimental in another.
3.2 Introducing practical variability
Figures 3.1, 3.2 and 3.3 show some steps along the road to variability. In the first case, Fig. 3.1, position A would enjoy a rather live acoustic, whilst position B would be much more dead. A is surrounded on three sides by reflective walls. B is surrounded on three sides by very absorbent walls, and the only live surface directly facing it is the far wall of the live area, which is quite some distance away. Figure 3.2 again has positions A and B, live and dead respectively, though this time the method of achieving the effect is rather different. The sawtooth arrangement of the reflector/absorber surfaces creates a situation whereby as one moves from the absorbent wall to the reflective one, there is a progressive reduction in the reflective surfaces facing the signal source. If an instrument is at position B, with a cardioid microphone at position C, facing B, then an almost dead room response would be recorded, as all the reflexions would be passing away from the front of the microphone diaphragm. Conversely, by positioning the microphone at D, many reflexions would be captured by the microphone, both from the hard wall and from at least the first three ‘sawtooth’ reflectors of each of the side walls.
Figures 3.1 and 3.2 are acoustically variable in terms of both the positions of the sources of sound and the microphones, but Fig. 3.3 shows a move toward a more variable room. In this instance, hinged panels can be moved through 90°, exposing either hard or soft surfaces to the room. By this means, either the whole room, or different sections of it, can be radically altered between live and dead. I have built many small rooms of this type, along with a variant where the hinged panels are in the centre of the long walls (see Chapter 8: Fig. 8.12 and Section 8.6), allowing the rooms to be sub-divided. They have generally been very well received, but in the case of larger rooms, this type of system is somewhat impracticable.
Figure 3.5 shows a proposal for a large room with quite comprehensively variable acoustics. From this it can be seen that a large room with seriously variable acoustics does not come cheap. Space is consumed by the variable elements, which means that considerably more floor area is needed in the building shell than will be realised in the final recording room. Highly variable rooms tend to be both structurally complicated and expensive, but they can also be very effective recording tools when a truly multi-functional room is needed. Essentially, though, it will be seen from Fig. 3.5 that almost every surface of the room needs to be capable of being changed from hard to soft if a very high degree of variability is to be achieved. An added possibility would be the provision of carpets, which could be laid down over the floor, or removed, as necessary.
Figure 3.5 Room with high degree of acoustic variability (but at a cost of much floor area). (a) Plan; (b) side elevation
Figure 3.6(a) shows the details of the rotating panels of Fig. 3.5, with four different designs, though many other variations on this theme are possible. The diffusive sides show the options of either curved surfaces of different radii, or the quadratic residue types of both the pit and relief forms. This type of room variability technique is now widely used in concert halls, though it was largely the Japanese designer, Sam Toyashima, who brought it to high profile in recording studios around the world. The wall panels of Fig. 3.6(a) have three surfaces, which can be rotated into any position desired. Having three surfaces gives the option of mixing any of the diffusive, reflective or absorbent characteristics by intermediate positioning. The room can be divided into live or dead areas, and, by the rotation of the panels, the reflexions are to a large degree steerable, which can create interesting effects for ambience microphones. The ceiling panels can be raised, lowered or rotated, and therefore can be absorbent, diffusive or reflective with varying degrees of timing for first reflexions.
Figure 3.6 Figure 3.6 (a) Detail (end views) of rotating panels – four variants. Rotating contrivances such as these can provide reflective, diffusive or absorbent surfaces, either wholly, or in combinations, by being rotated to intermediate positions. They can be applied to walls or ceilings, and can be motor driven and controlled from the control room, where their effect on the recording acoustics can be judged whilst listening via the microphones. (b) Mounting arrangement for rotating panels shown in (a). Ceiling mounted devices would use a similar arrangement, but would be mounted horizontally
With rooms such as this, it is often surprising to the uninitiated just how much area needs to be changed to achieve any significantly noticeable change in the general acoustics. Except for microphones and instruments in close proximity to the variable panels, rotating only three or four of them would usually be virtually unnoticeable in the context of a change in the overall acoustic of the room. It really takes a change of 20–25% of the overall surface area of the room to have any readily noticeable general effect. In a 15 m 10 m 5 m room, accepting for now a hard floor, it means changing something around 100 m2 of surface area for a worthwhile change in acoustics. Once over that threshold, however, the effect, in which ever direction it is acting on reverberation time, begins to develop rapidly.
The rotating wall-section technique takes up more permanent space than the method of varying the wall surfaces by movable panels, either hinged, as in Fig. 3.3, or attached by some sort of hook system, but with the panel system it is more difficult to achieve the intermediate situation, and it may also be less rapidly adjustable. Obviously, the hinged panels need a free space to swing out, which is perhaps not too much of a problem in a very large room, but in smaller rooms it could mean dismantling a whole drum kit just to make an experimental change. On the other hand, if the changes of acoustics can be pre-planned, the hinged/hung panel system does cause less permanent loss of space inside the isolation shell than does the rotating system of Fig. 3.5. As usual, compromises exist in each case.
3.2.1 Small room considerations
Small recording rooms, of less than around 150 m3, tend to be more difficult in terms of variability. They tend to flip-flop from one state to another, passing through some strange sound characteristics in the intermediate stages. What is more, in a small room the number and positions of people and equipment occupy a much greater proportion of the room volume, and hence may themselves have a great bearing on the acoustics. The variations in the surfaces of small rooms usually have to be judged once the room is ready to record, with all personnel and equipment in place. As mentioned earlier, with wheeled or hinged panels, changing things after the instruments have been positioned can become a very disruptive and slow process in a congested room. I usually prefer to sub-divide smaller areas into sections whose acoustics are likely to be the ones most appropriate for the majority of the music which the studio expects to be recording. A technique such as that shown in Fig. 3.5 would be unlikely to be a good choice. The overall loss of space due to the movable sections would be a much greater proportion of the space in the isolation shell, and their effects would not be so subtle because they would be very close to the instruments or vocalists.
The important point to remember is that the movable panels of Figs 3.5 and 3.6 will not scale. They cannot be built one half of the depth in a room of one half of the volume. The effects of the panels are related not to room sizes but to wavelengths, and as the frequency ranges of the instruments in a small or large room remain the same, some of the dimensions of the variable wall sections must remain the same. If an absorber needs to be 1 m deep, then in a room of half the total surface area, although only half the surface area of absorbers may be needed, they still need to be 1 m deep. I have been in studios where small scale versions of these devices have been tried, but they have tended to be effective over much narrower frequency bands, and their sonic effect has usually been very unnatural. For so many reasons, recording spaces tend to be like boxers; a good big one will almost always beat a good little one.
In the smaller rooms, it is almost impossible to achieve situations whereby the first reflexions back to the instrument are of the late type (echoes). There are many more early reflexions returning to the musicians and the recording microphones, so the reflexions of this type produce tonal colouration rather than a sense of spaciousness. This can be off-putting for the musicians, whether they hear the sound directly or via the microphones and fold-back systems, both of which can affect the way that they play. All in all, it has been my experience that smaller rooms are better suited either to fixed design concepts, or to being provided with the means of gross state changes. Subtle variability rarely seems to achieve its aims in small spaces, although proprietary diffusing systems, such as those produced by RPG, do offer solutions to many of the small room problems. They can at least provide a degree of reverberant life, without a build-up of troublesome early reflexions, but in situations which need reflexions for spaciousness, they can do little to help. Even diffusers suffer from wavelength limitations, and for good low frequency diffusion, they currently need to be relatively large in their front-to-back dimension. However, some interesting research is afoot, whereby actively driven end-walls, at the back of the pits, may be able to simulate in a shallow pit the effect of a deep one. Time will tell! Nonetheless, in experienced hands, the idiosyncratic nature of small, variable rooms can be put to good use, especially in the more ‘creative’ environment of modern music. In a situation where ‘different’ sounds are being sought, there can be some interesting possibilities, but one has to be careful that similar room-sound characteristics do not build up on themselves to become overbearing in the final mix. The following story may help to highlight the point.
I was once involved in the mixdown of a tape containing five acoustic guitars from a television recording. The guitars had been recorded via pickups on a very large, important live recording before an audience of 10,000 people. During the mixdown for TV, it was realised that the spirit of the occasion had somewhat dominated the event. Of the five guitars, only two were generally usable, but nonetheless, the programme looked great. The band in question were platinum sellers in their home country, and it was not long before their record company began pressing for a live album from the event. Well, what will stand for a one-off TV broadcast and what will stand critical repeated listening, without the distraction of a picture to watch, are two different things. It was decided that all five guitars would need to be replaced. This is not quite the musical travesty that it appears, as it was something of a celebration concert, with rather an excess of guitars, but the record company wanted to keep the overall feel of the live event which had an unusually ‘electric’ atmosphere. The rhythm track was excellent and the vocals were powerful, so the job seemed worthwhile.
Using microphones for recording the acoustic guitars at such a live event was not practical, but on the studio albums of the band, the guitars had all been acoustically recorded. A studio was booked to make the overdubs, and it was decided to record the replacement guitars with microphones, and not with the pick-ups used in the live recording. The engineers recording the replacement guitars chose a position in a relatively live room where they considered the correct balance to exist between the direct and ambient sound, giving a fullness to the guitars which the direct injected recordings lacked. Although each guitarist brought his own guitar, when it came to the mixing stage it became painfully apparent that all the recordings had been done in the same place in the same room; and what was more, with the same microphones. Each guitar track on its own sounded very good. Each guitar track when played with the backing track sounded perfectly acceptable. Nothing at all had stood out in the individual guitar sounds to cause anybody to question the sounds during the recordings. But, when all five guitar tracks were added to the mix, the recording room was clearly evident, and sat very unhappily with the live instruments from the concert. I do not blame any of the recording personnel for this, as they were unfamiliar with the room and had only been given notice of the overdubs four days before the record company said that the tapes needed to be at the CD plant, ready for a Christmas release, but it was a salutary lesson of how the idiosyncratic nature of any room, other than a dead or neutral one, can insidiously build up if used to excess.
A later chapter deals with the subject of vocal rooms, and here again, the unnatural acoustic of many of the typical rooms, particularly some of the older ones, can stand out in an unpleasant way. This is particularly true when listening on headphones, or in many modern control rooms with dead monitoring conditions, when the control room ambience does not swamp the perception of the low level ambience in the recordings. The moral of this story is that if you do have variable acoustic conditions in a smallish room which will be used for many overdubs on the same recording, then use whatever variability is available, even if this perhaps means that each individual recorded sound is not at its considered optimum when listened to in isolation. Most of the characteristic sound of the small rooms such as those under discussion here are not frequency aberrations per se, but characteristics of the time response. It is not just the frequency content of the room sounds which stand out, but the time-performance of the resonant modes and reflexions which give rise to the idiosyncrasies of the overall decay. As such, with the problems being in the time domain, there is little hope of using the equalisation controls of the mixing console to correct the problem. Any attempts at such ‘corrective’ equalisation will seriously detract from the frequency balance of the direct sounds, and in many cases, the effect of the medicine will be worse than the illness.
Such room problems can usually only be solved by acoustic changes in the rooms themselves. If small rooms cannot be significantly changed acoustically, then at least the positions of subsequent recordings should be changed. If we are to continue our medical analogies, although the room-related problems may not be terminal to a recording, once in the recording system, they do tend to be incurable. In smaller rooms, where the room sound is noticeable, it is of paramount importance to monitor it carefully at the recording stage, and if any sign of a characteristic acoustic build-up becomes evident, steps must be taken to ensure that further recordings on any given song are done with a varied acoustic. This may mean moving the musician(s) and/or the instrument(s), moving the microphone(s), or making use of any acoustic variability. Although changing a microphone type may change the direct/reflective pick-up characteristic, it is not likely to be as effective as the former measures, as the offending time characteristics are quite likely to still find their way into the recordings. Nevertheless, idiosyncratic acoustics do have their place when they are available for effect, so perhaps we should move on to the next chapter; a discussion of live rooms.