Stage Acoustics for Symphony Orchestras – Just Black Magic? Part II

HOW ACOUSTIC CONDITIONS AFFECT PERCEIVED CONDITIONS

Not much research has previously been carried out regarding which physical or acoustic factors control the ability to hear the other players clearly well balanced with sound from one’s own instrument. The most significant contributors to this subject are Jürgen Meyer and Anders Christian Gade and their research groups over the last four decades. This section focuses on how level and time differences may relate to perceived balance between instruments within the orchestra and conditions in general. Due to a lack of relevant scientific investigations it is only possible to draw up possible relations based on knowledge about masking effects, the precedence and the cocktail-party effect.

The subjective relevance of the within-orchestra sound levels and delays

If the delay of direct sound from different players is consistent the players appear to be able to adapt to the delay of sound from instruments across the stage. It appears likely that orchestral players are able to adapt to (more or less subconsciously) and not get rhythmically disturbed by delays of the direct sound within approximately 60 ms, which is the approximate maximum delay they will experience. This would imply that any sound events within 60 ms may be perceived as the ‘direct’ or the ‘immediate’ orchestra sound among the players. But the different delays from different instrument groups can contribute to perceptual temporal masking, where the sound from the instruments that arrives first contribute to make later arriving sound inaudible. Such an effect appears difficult to adapt to and overcome, since the masking effects origins much from the biological construction of our auditory system.

The significant level balance and time arrival differences between different instruments (Figures 4 and 5) appear to make it difficult for the players to hear all the other players clearly, well balanced with the sound from one’s own instrument. The string players can be expected to easily struggle to hear each other, since the mutual sound between string players can be perceptually masked by louder and/or more proximate instruments which are less delayed and attenuated. The sound from the instruments at the back will both arrive first and at a higher level and will easily mask the sound of other string players. Additionally, the percussion and brass players may struggle to hear the string players due to high direct sound levels from players within their own group or their own instrument.

These conditions of sound level and propagation delays are likely to make it difficult for the players to balance correctly. With significant attenuation or masking of sound at frequencies above 500 Hz, over-harmonic sound components from in particular strings are significantly reduced in level and the attack/transient sound of instruments are less audible. This is likely to also lead to intonation problems and acoustic timing cues being less audible. The intonation problems may be exaggerated by level differences between instruments on stage differing significantly compared to what the audience hears. This could be related to perceived pitch being affected by the actual sound level. If the sound of one’s own instrument is very loud relative to the others players, the player could potentially misjudge the pitch compared to what the audience hear when the over-harmonics of the sound is not clearly audible. The problem with mutual hearing across the stage might be possible for the players to learn to cope with to some degree by following visual cues, but visual cues will only be useful for timing, not balancing or intonation. These considerations have been made solely on differences of the within-orchestra sound level and delays (excluding the acoustic response from the stage enclosure and the main auditorium). The following two sections discuss the possible importance and role of the stage enclosure and main auditorium on perceived conditions on stage.

The subjective relevance of the staging conditions and the stage enclosure

The differences of within-orchestra levels between players may be addressed by the three following aids: the use of risers for the different instrument groups, screens or increased distance to reduce or block the direct sound from excessively loud instruments and reflections provided by the stage enclosure. The use of increased distance to and screens in front of loud instruments will decrease the direct sound level significantly, particularly at higher frequencies for screens. But with large distance between players there will be increased problems with time delays. Risers will normally contribute to increase direct sound levels at higher frequencies, but with a certain height of the risers the direct sound from brass sections will to some extent propagate above the heads of woodwind and string players at high frequencies. Three major factors are involved with the placement of reflecting surfaces: the level, delay and direction of the reflections introduced. The efficiency of a reflecting surface to improve the inherent level balance problems appears to depend how the reflecting surface reflects the different instruments groups in terms of both level and time arrival (delay). Referring to figure 4, any reflecting surfaces closer to the percussion and brass compared to the string players can make the skewed perceived level balance worse. Referring to figure 5, any reflecting surfaces that lead to a reflection of percussion and brass arriving before the arrival of the direct sound from string players far apart can also contribute to the perceived conditions getting worse (due to perceptual masking in the time domain). Such reflections are within this study called ‘competing reflections’ since they compete with the later arriving string sound at low level. On the opposite, reflections that contribute to compensate for the skewed perceived level balance are called ‘compensating reflections’. How to accomplish effective compensating reflections taking into account the delay of the direct sound on stage is discussed below.

Early reflections of percussion and brass from the back wall and ceiling/overhead reflector can appear before the delayed direct sound from the strings on the opposite side, and hence be competing reflections. From Figure 5 we see that the direct sound lags by up to 48 ms relative to the direct sound from percussion and brass. Any additional reflections from percussion and brass appearing before the direct sound from the strings on the opposite side will make it even more difficult to hear the string players on the opposite side, due to masking effects. A ceiling/overhead reflector must be about 11 m high to avoid the ceiling reflection from percussion and brass appearing before the direct sound from the strings at the opposite side, assuming the brass players are on a 1 m high riser. Avoiding overhead surface at low height above the string players may also be beneficial to avoid the sound of one’s own instrument becoming too loud for the string players or the conductor. A reflecting back wall can be beneficial for the horn players, but may be made absorbing behind percussion and brass to reduce the level of competing reflections. On the contrary, close to parallel outwards sloping reflecting surfaces at the sides of the orchestra (close to the outmost string players) will introduce reflections that may effectively compensate for low within-orchestra sound levels from the strings. Such reflections will have minimum delay and maximum level of string sound (due to minimum propagation distance and proximity to the string players). By having side walls with vertically tilted sections at 2–3 m height, the reflection path will not be obstructed by the orchestra and the side walls will not lead to late arriving reflections on stage. The level of this reflection can therefore be approximately 10 dB above the attenuated direct sound and take perceptual precedence over the direct sound (the precedence effect). It appears important to minimize the delay of the side reflecting surfaces so the sound of strings at the opposite side is not delayed much more than 46 ms. If this reflection is delayed above approximately 60 ms it can represent an excess delay of the string sound that the players are not accustomed to. With delays above 60 ms the compensating reflection from the sides therefore may end up being detrimental to clarity of the sound of others. If having an orchestra width of 16 m and reflecting surfaces at the sides at 2 m distance from the orchestra, the side reflection will appear within approximately 58 ms across the stage. If we would like this reflection to appear before any overhead reflections from percussion and brass, the overhead reflecting surface must be increased further, from 11 up to approximately 13 m.

In general it appears important to avoid any strong late arriving discrete reflections within the stage enclosure, which could be detrimental to clarity of sound and perception of rhythm and clarity. A narrow stage enclosure with outwards sloping and/or vertically tilted sections will not easily lead to late arriving reflections, while a high ceiling and/or wide enclosure easily can. For a high ceiling and wide stage it appears relevant to which degree the reflecting surfaces contribute to scatter the sound spatially (sound diffusion).

The above consideration suggests that a narrow and high stage enclosure will be more beneficial than a wide and high. Avoiding reflecting surfaces close to the percussion and brass in combination with side reflecting surfaces close to the orchestra at the sides appear to lead to better perceived level balance and clarity of sound. Additionally, with a moderately deep stage (typically 12 m) without any flat reflective back wall, neither the level nor delay of the percussion and brass will become extreme. Another benefit may be that the reinforcing reflections of strings will appear much in the same direction as the direct sound. This may make it easier to localise the sounds of different instruments more correctly. By reference to the cocktail-party effect this may therefore improve the ability to listen to specific instruments. Narrow side walls will also provide a reflecting surface close to the double basses, which is beneficial for raising the low frequency sound level of double bass, for both the orchestra and the audience. Jürgen Meyer studied acoustic conditions for the conductor. He found that overhead reflecting surfaces above the string players could to lead to the string players becoming too loud for the conductor, whereas overhead reflectors that reflect woodwind down to the conductor may be beneficial.

The considerations above does not necessary rule out overhead reflectors below 13 m as an alternative. Some designs of overhead reflectors that reduce the negative effects may exist, but the initial conditions set up by narrow side walls appear to supersede the initial conditions set up by a low overhead reflector. For an overhead reflector, dividing the reflector into smaller parts (reflector array) as well as avoiding fully flat reflective surfaces, can contribute to avoid the negative effects. The results from the investigations of existing stages suggest that overhead reflectors at a height below 13 m may not be needed at all if the stage enclosure is sufficiently narrow and high. With wide and high stage enclosures an overhead reflector array partially acoustically transparent can provide some useful compensating reflections. The results suggest that such an array can contribute positively to improve perceived clarity of sound on certain stages. But without any side reflecting surfaces close to the orchestra the level of compensation may not be high enough for the string players, compared to possible competing reflections also introduced by the array. Overhead reflector arrays may also be used for increasing the level of woodwind instruments, but depending on the reflector design it could result in significant variations of acoustic conditions between different positions on stage, particularly for the bassoonists who have their instrument pointed upwards.

The subjective relevance of the acoustic response from the main auditorium

In general, if performing in a very acoustically dead space with a low level of reverberant response it will easily lead to an enforced playing style. The players need to put a lot of effort in, without receiving any audible response back towards them or enhancement of their sound. The acoustic response received back from the room is found to be referred to as ‘bloom’ or ‘resonance’ among the players. Most orchestral works are based on the instruments being enhanced by the reverberant response from the room. The level of the reverberant response appears relevant, assessed by for instance the acoustic measure G_l. If the level of the reverberant response from the room is low, the within-orchestra sound or early reflections from the stage enclosure may perceptually mask the acoustic response from the main auditorium. If the level is high, the overall sound levels will become too high and the reflected sound will dominate, leading to poor perceived clarity of sound. Such level differences will somehow be monitored by the acoustic measure C₈₀ (the ratio between reflected energy arriving before and after 80 ms). For a symphony orchestra it therefore appears to be an optimum level of the reverberant late arriving acoustic response, controlled by the total within-orchestra sound levels. (For other ensembles other optimum ranges may exist.)

Not only the level of the late reverberant response appears relevant, but also the quality and direction of it. The quality of the acoustic response will be shared with the audience and contributes to general enjoyment of the sound produced. If the reverberant acoustic response appears mainly from the main auditorium, the direction of this sound component will deviate significantly from the direction of the within-orchestra sound. The players will in such a case be to a high degree able to hear the reverberant sound as the audience hears it and with reference to the cocktail-party effect they can be able to easily separate this sound from the stage sound. This appears to contribute to what the players refer to as ‘projection’ (reaching through to the audience). This enables the players to adjust their articulation to fit the reverberant response from the main auditorium. Hearing the acoustic response from the main auditorium clearly as a separated entity is reassuring for players since it enables the players to hear what the audience hears – to a certain extent. The response from the main auditorium may also be helpful for assessing level balance and intonation, since the main auditorium’s response of the orchestra sound will have the level balance and pitch relations between the different instruments closer to what the audience hears (who will after all be the final judge of the performance). Differences of measured G_l on stage and within the stalls section of the audience area may provide an indication of the dominating direction of late acoustic response on stage.

A low solid and flat ceiling parallel to the stage floor at low height above the orchestra appears to be a worst case scenario for a symphony orchestra, especially if this is combined with long distance between flat parallel side walls on stage. A stage enclosure close to this description was reported as leading to a loud stage sound, where the balancing problem imposed by the orchestra remains and where the combined early and late acoustic response from the stage enclosure ‘overpowers’ the reverberant response from the main auditorium. On the contrary, this stage enclosure was reported to work fine for chamber groups. Such smaller ensembles will not lead to high sound levels accumulating within the stage enclosure and the acoustic communication between players is easier to start with. By exposing the stage more to the main auditorium, the stage sound will be more influenced by the acoustic response from the main auditorium. If having reflecting surfaces at the sides that effectively provide compensating reflections for the string players, there may not be much more need for reflecting surfaces close to the orchestra.

The considerations above suggest that the stage enclosure is not solely for aiding acoustic communication between players, but also for balancing the orchestra sound for the audience and making the acoustic response from the main auditorium audible on stage. In this way the musicians can communicate acoustically with both each other and the main auditorium. On proscenium stages where the stage is enclosed off from the main auditorium, overhead reflectors may have the positive effect of directing the reverberant response from the main auditorium towards the players. The same overhead reflecting surfaces may also contribute positively to raise levels and improve the balance between instrument groups for the audience members at far distance from the stage. Outwards sloping upper or downwards tilted sections of the side walls and outwards sloping ceiling/overhead reflections can therefore appear beneficial to both direct unneeded early acoustic response of the orchestra away from the stage, and at the same time direct the late acoustic response from the main auditorium down towards the players. To achieve this, the reflections from the stage enclosure must not be directed solely towards the audience seats and the seats should not be too absorptive. In such a case the sound directed towards the main auditorium will be significantly absorbed and reduced in level before it reaches back to the stage.

DESIGN OF STAGE ENCLOSURES

Figures 6 and 7 show what appears to be a rough description of beneficial stage enclosures, given an exposed or enclosed enclosure. An exposed stage enables good communication with the main auditorium, but may not always be feasible. The suggested designs are based on the results and considerations mentioned above and with woodwind, brass and percussion on stage risers. Recommendations for values of the architectural and acoustic measures, W_rs, H_rb, D and G_l, are included. The values of G_l are for an unoccupied audience area and for a stage with chairs only.

For the exposed stage overhead reflectors may not be needed. In such a case the top reflecting surface will here be the ceiling of the main auditorium (not included in Figure 6). The top section of the side walls are tilted downwards towards the orchestra for providing unobstructed reflections from the sides. The need for compensating reflections provided by the stage enclosure will depend on the design of stage risers. If the outmost string players are on risers – for instance by use of a circular riser system – the need for compensating reflections is likely to be reduced (not studied in the detail in this project). For an enclosed stage both side walls and the ceiling are outwards sloping and overhead reflectors may be needed for the acoustic communication with the main auditorium. The dashed lines illustrate what appears to represent a problematic layout of the stage enclosure.

All the details of the stage enclosure are not included in Figures 6 and 7, only the overall properties which can serve more as ‘rules of thumb’ rather than exact requirements. The acoustic properties of the surfaces (degree of absorption and sound scattering, diffusion) will have a significant effect on the total level of reflected sound on stage and the spatial distribution of reflections – with respect to compensating and competing reflection as well as the dominating direction of the late acoustic response. For the exposed stage the acoustic properties of the ceiling of the main auditorium will also be relevant. The appearance and acoustic properties will also affect the conditions for the audience. With too many reflections accumulating within the stage enclosure – with for instance by a stage enclosure with hard and non-diffusing parallel surfaces – the sound can have a lack of clarity for the audience.

Examples of venues where the musicians have reported that the overall acoustic impression is good are: Musikverein in Vienna, Symphony Hall in Birmingham, Bridgewater Hall in Manchester, The Anvil in Basingstoke, Philharmonie in Berlin, St David’s Hall in Cardiff and Symphony Hall in Boston. All these venues also have a good reputation among conductors and the audience and they have a stage enclosure that to a large extent corresponds with the two types of enclosures described in Figures 6 and 7.

COMMUNICATION BETWEEN PERFORMERS AND ACOUSTICIANS

How musicians and acousticians relate to acoustics make it challenging for these two groups to communicate ideas and experiences effectively. The musicians will often struggle to put the finger on concrete elements of the stage design if being asked by an acoustician about the acoustics. Musicians have learned more or less subconsciously to not be too affected by acoustic conditions. Musicians are not much trained in natural sciences, and acousticians are not much trained in musical science, so the qualitative descriptions of the musicians can easily be misunderstood or discredited by the acousticians.

From the studies of existing stages mentioned above, the existing acoustic measures did not prove highly relevant for assessing acoustic communication between players on stage. Over the last 30 years, acousticians have given priority to quantifiable acoustic measures for assessing stage acoustics. These acoustic measures are very simplified representations of the complex reality on stage and perception among the musicians. Less quantifiable concepts described in this article have not received much attention. Apparently this may have led to significant limitations on making progress with knowledge of stage acoustics. It is believed that through discussions where the complex and non-quantifiable factors are included we can bring acoustic stage design forward – with the musicians involved. The selection of approaches and methods from the 3-year project benefitted significantly from input from musicians. If the musicians get to better understand the physical and acoustic mechanisms, it may be easier for the musicians to know what to look/listen for on stage and improve their ability to see and communicate relations between physical/acoustic conditions and perceived conditions. Likewise, if the acousticians learn to better understand the musical/performance aspects and the complex factors contributing to perceived conditions, it can be easier for musicians and acousticians to communicate and exchange ideas and experiences.

CONCLUDING REMARKS

The musicians appear to focus on two different sound streams: hearing the others within the ensemble including oneself and to some extent hearing what the audience hears. The musicians will benefit from being part of what the audience hears so the performance can be adapted to the audience\’s perspective. Good acoustic conditions on stage may therefore start with providing good acoustic conditions for the audience that the musicians on stage can be part of. It has been well known that the stage enclosure will affect the conditions for the audience, but the main auditorium’s effect on conditions on stage may have been underestimated. The direct within-orchestra sound has a significantly skewed level balance. By knowing how this level balance effectively can be improved by a stage enclosure, the players can be able hear all other players without loosing contact with the audience sound. If reflections are introduced that ineffectively improve the level balance the stage enclosure can define its own acoustic space, leading to the orchestra becoming isolated from the main auditorium. Based on considerations of level balance between instrument groups and how the instrument groups synchronise relative to each other, a preference for a narrow and high stage enclosure and moderately deep stage has been found. In addition, having the stage highly exposed to the main auditorium and avoiding large parallel surfaces for the stage enclosure appears beneficial. It is difficult to fully prove these preferences scientifically, but they are supported by the results from both subjective and objective studies.

With the current practice the ST measures are used as the main design guide for stage enclosures. These measures only assess the level of reflections returning back to the stage without a full orchestra present, ignoring any information about direction of the reflected sound. Putting reflecting surfaces above the orchestra may be the most accessible location, which may explain why this solution has gained popularity among acousticians. Based on obtained ST_early an overhead reflecting surface will apparently not make a difference from side reflecting surface at similar distance from the orchestra, since the directions of reflections are ignored. This can explain why stages with similar values of ST_early have very different reputation among the musicians. If only studying values of ST there appears to be a risk of missing out on aspects which are highly relevant for the musicians. In such a case there can still be some element of what may be called black magic to the acoustic design of stage enclosures. By taking both the quantitative and qualitative aspects of the stage and stage enclosure into account, it may be easier understand the underlying mechanisms and design functional venues and stages for symphony orchestras and the audience. This corresponds to designing a well functional solution in any complex system. The proposed architectural and acoustic measures may prove to be a good starting point with regard to objective quantifiable characteristics, while the direction of reflections and inclusion of the orchestra represent potential improvement for objective assessments.

Appropriate acoustic conditions on stage can contribute to better artistic freedom. For instance with a well designed main auditorium and stage enclosure combined with a flexible riser system, an orchestra may easily experiment with different orchestra arrangements (like the German instead of the American) and different orchestral pieces without experiencing too exaggerated communication problems. Several orchestral pieces were written for a German orchestra configuration with 1^st and 2^nd violins on opposite sides of the stage. Today several orchestras choose the American and more modern orchestra configuration instead (where the 1^st and 2^nd violins are joined together at the left side of the stage), in some cases apparently because of acoustic challenges. By use of the American configuration instead the left-right ‘dialogue’ between violins that the composer had in mind is lost. From this we can state that well designed acoustic spaces for symphony orchestras will not only improve the working conditions for the musicians and conductor and lead to more exciting performances for the audience. It can also make it easier to perform orchestral pieces corresponding with the intention of the composer.

ACKNOWLEDGMENTS

The concepts and hypotheses above are very much a result of valuable discussions and input from orchestral musicians. The author is also grateful to those who have given valuable input to this article.