LOW FREQUENCY SOUND IN ROOMS

Room boundaries reflect sound waves.

 

For low frequencies (typically where the room dimensions are comparable with half wavelengths of the reproduced frequency) waves reflected multiple times can create resonances (peaks and dips in sound pressure level) at particular locations in the room. These resonances are often called room modes or standing waves. It is very desirable to control or even eliminate the influence of room modes, as these will colour the reproduced sound. How could we do this?

 

First, we need to understand the nature of room resonance modes and of boundary reflections and in particular the difference in behaviour between them.

 

Room mode resonances are sealed chamber behaviours. They can have very large resonance Quality (Q) factors with long establishment and decay time constants. Any loss of energy from the chamber (room) will lower the level of trapped energy and lower the resonance effectiveness significantly. This can occur through lossy boundaries or through room leaks.

 

Boundary reflections are a function of the amount of reflecting surface. As for room modes, multiple boundary reflections will occur in sealed rooms, but unlike room modes, boundary reflections still influence sound for very open structures. Slap echoes are an example of boundary reflections and can occur between two flat parallel walls even when there is no roof and so no room modes are sustained.

 

We can use this difference to advantage. If we can somehow couple the initial low frequency energy to the listener and then remove it before room modes have a chance to be excited we should be able to improve bass delivery.

 

To understand how this could work, we first consider the theoretical case of a football shaped (rotated ellipsoid) room, say 6 metre long. Such a room would have two reflective foci.

If a point source were placed at one focus point and the sound measured at the other symmetrically opposite focus point, the sound would radiate out spherically from the source in three dimensions, be reflected by the ellipsoidal boundary and exactly converge back down to a point at the second focus. Figure 1 shows modelled wave-front snapshots of this happening.

 

Figure 2 shows the sound pressure intensity of these components at the second focus during the propagation time. The earlier arriving sound energy spike shown in Figure 2 is the spherical wave part radiating directly from the source and past the second focus prior to being reflected by the boundary, and is almost 40 dB lower in intensity than the totally reflected wave about to arrive. After arrival at the second focus, the wave-front then diverges and later converges on the first focus (the source location) again.

 

This theoretical study clarifies some important principles.

1. The source and listener locations and the boundary shape are all important when considering the initial delivery of low frequency sound in reverberant environments.
2. It is possible to deliver 100% of the low frequency source energy to a location in a room with careful choice of the room construction shape and the source and listener locations.
3. The reflected wave can be precisely delivered to a specific location in a room by the choice of room geometry and both source and listener location.
4. The direct wave component is not necessarily the dominant low frequency source in rooms.
5. The direct and reflected wave components of low frequency sound can be separated by small time intervals (less than 10 msec) in rooms of quite large dimensions ( 6 metres in this case). If acoustic low frequency events separated by less than 10 msec are generally perceived as being simultaneous, the direct arrival sound will not be discerned as a separate event. It would be “post-masked” and would simply add to the perceived level.

 

For the ellipsoidal room example above, as time continues the reflected wave then diverges from the second focus until it intersects the boundary a second time and is reflected back to the source. This cycle of expansion/ reflection/ convergence would continue forever, but for dispersion and absorption by the air and the walls.

 

This behaviour is not a volume resonance as it is independent of frequency. Independently, the ellipsoid room supports certain resonances at particular frequencies independently of the early reflection behaviour described above.

 

If the source is now moved away from the first focus, the boundary reflection behaviour is altered.

 

As an example, Figure 3 shows the arrival wave near the second focus when the source was moved 300 mm above the first focus. It can be seen that the first point of convergence has now moved approximately 300 mm BELOW the second focus.

Figure 4 shows the acoustic energy levels at the centre of this arrival region over time. The early arrival (spherical) wave is still present though slightly lower in level because of the extra distance travelled. The reflected wave energy is now also significantly lower (-20 dB) in peak intensity.

 

This shows:

1. The boundary shape, source location and listener locations are all important for initial low frequency energy delivery
2. Even in large (6 metre) rooms placement can significantly affect delivered energy levels. In this case a movement of 300 mm lowers the delivered energy by 20 dB. 300 mm is the diameter of many present day woofers!
3. There appears to be a pattern to location of the source and the location of the first point of energy focus that is centred about the gas centroid (centre) of the room.

 

Very few people would have an ellipsoidal room but it should be apparent from this theoretical study that the room modes and the source to listener coupling warrant separate consideration. This forms the basis of new techniques and apparatus for control of room modes for low frequencies in reverberant environments.

 

 

Real rooms

The use of boundary reflections applies to all manner of room shapes. Figure 5 shows progress of a wavefront from a source to a first point of energy convergence.

Figure 6 shows the acoustic energy at the centre of this arrival region. Again there is a directly propagating spherical wave arrival before the main boundary-reflected wave-front.

The main reflected wave intensity is only approximately 20 dB down on that for the perfect ellipse study previously. This is comparable to the result obtained in the ellipsoid room when the source was relocated away from the ideal focus location by only 300 mm.

 

Better bass by placement

HuonLabs analysis of reflected sound fields in all manner of room shapes has provided a rule that helps determine where to place a low frequency loudspeaker (woofer) in any room for best results.

 

First you need to know the location of the centre of mass of the air (centroid) in the listening room. The rule then says that for best early coupling the listener and the woofer should always be placed on a straight line through the centroid and at equal distances from it. As this method of coupling uses boundary reflections, it is relatively insensitive to holes and gaps in the boundaries. HuonLabs refer to the approach as Wave Focus (WF).

 

So, if you sit anywhere in the room close to the floor, the woofer should “sit” at the corresponding location equally close to the ceiling, for example!

 

There are an infinite number of paired source-listener locations for optimal coupling. Choosing either a listener or woofer location reduces the options to one.

 

The choice of the location pair will influence the overall coupling effectiveness and the time separation of the early arrival sound, but each location pair will be the coupling “best effort”. Also, not all of these pairs will coincide with room mode node/antinode locations.

 

It is now possible to take advantage of the room boundaries to improve the delivered bass level and quality and to simultaneously minimise the effects of low frequency room modes in reverberant listening rooms simply by appropriate listener location and woofer placement, avoiding room node locations for either.

 

The use of the Centroids method provides improvement in bass transient performance and flatness of frequency response but the improved coupling also provides additional sound level for the listener. This can be used for more dramatic effect or the same level can be delivered by turning the bass down. Neighbours will appreciate this.

 

The above examples caution that the zone of “room mode free” bass is very restricted. This is in contrast to the view that low frequency reproduction is generally non-critical of location or placement. In some applications this is acceptable. For example with car sound delivered to the driver of the vehicle, the seated position is quite well defined. In other applications such as cinemas it is not.

 

The general introduction of more woofers at different locations does not help, as (with the exception of some exotic room geometries) the additional units cannot obey the centroid rule for any particular single listener location and will thus worsen temporal smearing.

 

Better bass by annihilation

In the ideal geometry case of the ellipsoidal shaped room earlier, the sound “bounced” back and forth between the two foci.

 

If the original sound source located at the first focus were to emit an anti-phase version of the original source at the appropriate time (when the original wave returned), the returning wave could be annihilated. A listener at the second focus would experience the full output of the source and then no subsequent excitation. As no energy is subsequently available, no room mode excitation is possible from then on.

 

HuonLabs calls this patented approach Wave Focus (WF) Annihilation. The rule of Centroids still applies, but here nature helps out by ensuring that the original source is at a point of energy concentration automatically for any room shape and so is in the correct location for best annihilation.

 

With annihilation, all the benefits of improved level, flat extended response and better transient performance are achieved but in addition, as the bass energy is effectively delivered to the listeners then a large amount is removed, there is less overall sound trespass outside the listening environment. Again the neighbours are happy.

 

Both Wave Focus and Annihilation deliver very accurate early arrival bass. This can sound dry where no bass frequency reverberation is available on the source material. The improved transient performance and extended smooth response more than make up for this on most program material.

 

Better bass for all

It would be very useful to achieve the mode influence improvements over an extended audience listening area. For such an extended audience in a cinema for example, simultaneous arrival time would also be desirable in order to support the visual image. This is now considered.

 

Some geometries of room will better enable the transmission and return of acoustic energy quite effectively and thus offer coverage for large audiences with minimal temporal smearing. One such geometry could create a plane wave that traverses the room, either by using parabolic reflectors or woofer arrays. This wave could then be “caught” and annihilated at the opposite end by another set of parallel wave generating woofers or could be reflected back and somewhat annihilated at the source by the original woofers, remembering that there will be at least some spherical component to the wave-fronts.

 

Unfortunately these configurations would not provide simultaneous wave arrival for audience members. Further, where the round trip delay was greater than 10 msec, multiple acoustic events could be perceived, and particularly where the waves reflect off the rear wall for annihilation on returning to the front.

 

But if the plane wave was launched from the ceiling and reflected off the floor for annihilation on return to the ceiling, the transit time for each audience member would be from head to toe and back and thus would be within 10 msec and so should be perceived as one event. An additional benefit would be the apparent increase in coupling effectiveness of the double pass. By this means, simultaneous arrival time, mode reduced low frequency reproduction could be achieved for the whole audience.

 

If the annihilation were reasonably effective, there would be insufficient reverberant energy to excite perceptible room modes and as a bonus, stray radiated sound would be minimised. The sources could be zoned to give simultaneous arrival for balconies, stalls and the like within the cinema because of the reduction in the stray acoustic energy from each zone. Figure 7 shows a schematic of such an arrangement in a cinema with stalls and a balcony.

 

Add within, cancel without

HuonLabs has extended the concept of WF Annihilation one step further. The techniques of WF and Annihilation ensure improved sound delivery to the listeners but by providing woofers that also couple some inverted phase sound outside the listening environment, sound leakage can be further cancelled externally[1]. The neighbours will be pleased.

 

And as the optimum location can be determined in advance, the woofer it can be installed so that the cabinet is outside the room and only the coupling port is visible. The owners will be pleased too!

 

The approaches of wave focus and annihilation apply generally. This includes set seating applications such as automotive entertainment systems and post-production studios as well as large audience applications including cinemas and other fully or at least partially enclosed environments and venues.

 

We now have a simple rule for improving bass delivery in rooms for any type or brand of woofer that will minimise colouration caused by room modes.

 

We also have new techniques for delivering improved simultaneous arrival of bass sound for large audiences. These approaches do not just make the bass louder. They deliver improved bass transient performance with extended flat frequency response – and happy owners and neighbours all round!.

 

The information contained herein is copyright to HuonLabs. No material can be reproduced in its totality or in part or without the express permission of HuonLabs Pty Ltd. Any reference to this material must quote the HuonLabs source. Trade marks and Patent applications apply to most aspects of the work disclosed here. Contact HuonLabs for further details, product information or licensing enquiries.