With the imminent arrival of InfoComm 2016, the professional audio industry is rolling out the booths, demo rooms and display trucks to show off what’s new in professional audio technology. Among the myriad sights and sound, many manufacturers will be displaying their latest and greatest loudspeakers.
Spanning the range from very affordable to very high-end, loudspeakers fill many niches and have many flavors. As loudspeakers are our most flawed products and also the one most expressing an artistic sensibility, choices and opinions for specific loudspeakers tend to be very strong. Simultaneously, the pace of innovation in loudspeakers has been slow compared to most other areas of the audio industry. Essentially, any modern loudspeaker would be (at least vaguely) recognizable to those who built the first transducers a hundred years ago.
This month, we look at the growth of a range of technologies that are being integrated into new loudspeakers. We also posit the future of loudspeakers as they could exist for the professional audio industry. There is a good chance that your favorite speaker company is already doing some of the things in this article. We do not predict a complete re-imaging of the loudspeaker as we know it (such as a modulated compressed air subwoofer) but rather a synthesis of potential improvements to collectively present a conception of where speaker technology could head in the future.
Computers and Product Design
Computer simulation has been widely used in the design of professional loudspeakers for more than a decade, and this behavior will continue. Simulation tools are used in professional audio to simulate many items, including: the magnet structures in loudspeakers, the performance of electrical components, the movement and radiation of transducers, the design of acoustic waveguides, the design of plastic molds, the loads on rigging components, the thermal performance of drivers and amplifiers and the spreading of sound over the audience area.
The power of these simulation systems continue to improve, which allows the designer to investigate more properties more thoroughly. The tools have also become more cost-competitive for smaller manufacturers, and the tools have become easier to use. Further, computing power to run simulations has improved drastically and now is available to be purchased in increments from online vendors. There are even fully cloud-based simulation tools like SimScale (simscale.com), with one variant shown in Fig. 1, to the right. These tools, broadly known as computer aided engineering (CAE), will probably continue to be adopted, more and more, for use by the industry.
Simulation tools are only as good as the boundary conditions that can be assigned to solving the underlying system. Sometimes these are very robust, and other times they are only a moderate approximation of real life. Because of this restriction, CAE tools are commonly used hand-in-hand with physical prototype models for test and validation. Prototypes can range from one-off test PCBs from a board house to CNC-milled rigging hardware to 3-D printed horn prototypes to vacuum cast heat sinks. These prototypes validate underlying assumptions made in simulation and can confirm performance and ergonomics before release to production. Rapid production of prototype parts has matured greatly in the last decade.
Of course computers also play a role in collecting performance data after prototypes are built, too. This can take the form of commercial data measurement tools like LabVIEW, or simple scripting on DIY hardware like an Arduino. And the data collected from the measurements can be analyzed and displayed with commercial tools like Matlab or Mathematica, or one could increasingly use open-source tools like Sage, R, and Jupyter.
Supporting the simulation and prototyping steps above are software tools that deal with the resulting data. Product data management (PDM) tools collect, organize, provide version control and manage changes for the collateral that is produced during the design process. Manufacturing resource planning (MRP II) tools help manage purchasing, stocking and release of work inside the factory. Product lifecycle management (PLM) tools glue all of the documents, processes, and people together through design and beyond. As the richness of the data that can be produced grows, the tools for managing data should also become more robust.
New Materials
Future loudspeakers, whose designs have been thoroughly simulated and vetted before production, are able to consider the use of more diverse material systems that require more up-front investment than plywood. Plywood still offers an appealing mix of acoustic performance, and ease of use, but there are other options in metals, plastics and composites. Once the designer has confidence in the product, then the expense for tooling is justified. Traditional molds for injection molding or die-cast parts can run to six figures, so the manufacturer needs to be sure of both the product and the market before making the investment.
Material cost for individual parts made via molds can be quite attractive, but the tooling costs must be amortized over a number of units. For a high-volume utility self-powered loudspeaker, the tooling is easy to justify, but for products with a more limited commercial footprint, the business case may not be there. Rapid prototyping methods have opened up the possibility to create some interesting new molded parts at a more reasonable cost. As an example, there are structural urethane parts that are made in silicone molds, and these can be produced directly from 3-D printed parts. As another example, it’s possible to build vacuum cast parts in aluminum or magnesium alloys via custom forms created directly from computer models. This allows creating intricate parts that historically had to be made with very expensive metal dies.
Loudspeaker vendors are increasingly sophisticated about utilizing materials intelligently where they offer the best performance. Thinner plywood parts can be strategically strengthened or stiffened with supporting elements; rigging hardware can be more tightly integrated into the enclosure; and a cabinet’s inner structure can be precisely shaped into monolithic structures, rather than being pieced together out of numerous subcomponents. Fewer components and the attendant attaching hardware simplify assembly, while reducing cost.
Advanced Motors
Arguably the invention and popularization of NdFeB magnet materials was the biggest breakthrough for loudspeaker drivers since the dawn of sound reinforcement. Since that happened in the 1980s, robust simulation of electromagnetic structures has enabled more complicated motor structures to drive speaker cones. These include drivers with multiple magnetic gaps, multiple voice coils, and voice coils that employ tailored winding along their length.
Drivers have seen great increases in linearity over the last decade, but they remain the weakest component in the overall reproduction chain. While stronger magnetic materials would be desirable, at this point, many transducers have the strength of the motor limited by primarily the ability of the steel that makes up the driver to carry more magnetic flux. Future transducers could end up using “returnless” magnetic structure, where the field lines are carried between multiple magnets directly through the air, and not through a guiding steel structure. The air does not have the flux limitations, and this might open up alternative configurations that yield stronger magnet structures. Stronger magnet structures offer flexibility to transducer design engineers. Do they want to improve driver efficiency, use a stiffer suspension to reduce box volume or save cost with a smaller motor? A stronger motor provides flexibility on these parameters.
Closed Loop Control
Drivers are ripe to have closed loop control, which is to say that the movement of the motor structure response would be analyzed and corrected based on deviations of the intended response. The idea of closed loop control is by no means new, but never have the sensors and hardware to implement control loops been more ubiquitous and affordable. Costs could come down to a point that enables widespread adoption with control loops offering improvements in control of distortion, frequency response, and the changes in voice coil resistance that occur as the assembly becomes heated by the input power.
Computing Power
Whether any future loudspeaker has closed loop control of its voice coil or not, there is a good chance that it will have substantially increased computing power, either onboard or on a tightly coupled external system. This computing power will allow for some immediate benefits for speaker performance. In the past, the equalization filters applied to loudspeakers have been of a type known as IIR (Infinite Impulse Response) biquads. Computationally, these filters are fairly efficient, but they do not provide very fine granularity of control of frequency response. By contrast, Finite Impulse Response (FIR) filters, which historically have been thought of in the context of phase linearization, can be used to create an extremely nuanced loudspeaker response, with or without independent phase manipulation.
IIR filters still offer plenty of uses, especially for behaviors at low and mid frequencies. Present and future loudspeakers will use a hybrid combination of the two filter types to produce an arbitrary desired magnitude response. This powerfully enables the loudspeaker company to focus on consistent coverage and low distortion. The final tonal palette of frequency response can then be shaped in essentially any way, using computing power to match the room and listener.
Behind the scenes, there is a sea of change taking place in the choices that the design engineer can make to obtain the necessary computing power. Historically, the math of filtering was executed on a dedicated DSP chip, and other functions within the controller, amplifier or powered speaker were managed by separate microcontrollers. Recently, though, microcontrollers have arrived that provide sufficient DSP capability to handle a common level of loudspeaker DSP filters without the need for an external dedicated DSP chip. Microcontrollers based on the ARM Cortex M4 hinted at this capability, and new devices based on the ARM Cortex M7 (see Fig. 2, at left) can be as capable as dedicated low-end DSP chips.
It does not necessarily make financial sense to use a single high-end microcontroller, rather than the combination of a simple controller and dedicated DSP, but the future path seems headed towards more integrated solutions. Another powerful growing reality is that it may become cheaper to implement most of the filtering done today with passive crossover components within a digital footprint, even for lower-end loudspeakers.
Power of Integration
With all the progress noted above, many aspects of how loudspeakers go together are shaped by physics and the need to cover a broad frequency spectrum. Twenty years from now, there is a very good chance that cones and horns will take forms broadly similar to those today, even if performance improvements have been executed in myriad subtle ways. What will change drastically, in my opinion, is the tightness of integration between electronics, electrical and mechanical components.
Consider a world where every powered speaker-on-stick in inventory has wireless Wi-Fi and Bluetooth connection to both the rest of the P.A. system and phones or other mobile devices. Loudspeaker controls are no longer dominated by knobs and switches that can be broken or moved by the audience. Instead, loudspeakers are programmed via the mobile device, and the settings for any given event are saved for permanent instant recall. All processing is executed at the speaker itself, and can be restored anywhere there is a power outlet. For low criticality events, signal is provided wirelessly to the speaker over a TCP/IP network and/or Bluetooth. Audio can be bridged locally into any speaker via a wireless microphone that can interject or intermix with the main feed from the mixing board.
Changes to equalization, volume, choice of input, auto mixer, etc. are handled from the mobile device and registered at the speaker level, rather than at any system DSP level. This future speaker-on-a-stick is able to utilize its integrated microcontroller to be aware of its orientation in space, and to automatically change its processing depending on whether it’s a stage monitor, delay speaker, main speaker, front fill, etc. This speaker can record and report relevant details about its operating conditions and make autonomous changes to its response depending on temperature and humidity of the event environment. Before and after the event, the speaker can use the mobile devices as a recording tool, and make sure its transducers are still performing to spec. Indeed, the loudspeakers could regularly check and update their response to match a used specified target curve.
For larger arrays, the loudspeaker would be even more aware of itself and its neighbors and know how to tailor its processing based on specified target curves and location in the array. Remote beacons track changing venue conditions and report details to the array, which then makes modifications to match the design profile.
More to Come
One of several final frontiers to consider include speakers that are inherently spatially aware of the environment in which they placed, able to map the space and make decisions based on the local geography. This would free the system engineer to focus on tweaking finer points for the room rather than focusing on gross alignment and configuration issues. If the simulation doesn’t match the real room conditions, such an array would be a powerful ally for provider.
If there are takeaways from this article, the first is that the future remains wide open for the further integration of computer technology in almost every aspect of professional loudspeakers. A second would be that loudspeaker builders are soon going to have a whole range of other features and capabilities to distinguish their products that are not being heavily used today. Tightly integrated systems that bring real advantages (beyond only making noise) represent part of the future value proposition for advanced professional loudspeakers.
