In 2024, Sonos introduced its new flagship soundbar, the Sonos Arc Ultra, which packed in more speaker units than its predecessor could manage at the same size, for more impressive Dolby Atmos sound.
The key technology that made this possible was a new super-compact bass speaker design, which Sonos calls Sound Motion. Sonos acquired this technology from a company named Mayht back in 2022, and the Arc Ultra was its long-awaited debut in an actual product. (Long-awaited for those of us who follow interesting speaker design innovations, at least.)
Sonos was surprisingly quiet about this technology (though the company at the time was in the process of recovering from its troubled major app update). But I’ve always wanted to hear more about the development process of Sound Motion and Arc Ultra – so I spoke to Naphur van Apeldoorn, Senior Manager of Hardware Development Engineering at Sonos, to find out more about Sound Motion’s development in the Sonos Arc, and its future at Sonos.
The reason that Sound Motion is something of a revolution is that it’s capable of moving a lot of air in a much smaller space than conventional speaker drivers, thanks to a flat, dual-sided design, rather than the single deep cone shape of a standard driver (though van Apeldoorn still refers to the moving membranes of the Sound Motion speaker as “cones”).
The Sound Motion speaker consists of four magnetic motors, one in every corner. They work in pairs, each one connected to the motor diagonally opposite by an aluminum rib, and the cone that pushes the air is also connected to the rib.
The magnets move up and down “like pistons”, says van Apeldoorn, and he notes that in this system, a “bigger portion of the transducer actually pushes air compared to a conventional transducer,” meaning that it’s highly efficient in terms of sound produced for the size.
“To give a reference, the single Sound Motion woofer in Arc Ultra has the same output as every single woofer in [the original] Arc, and Arc had eight woofers,” he explains.
This makes it especially great for putting bigger bass in a smaller space – van Apeldoorn explains that the Sound Motion unit in the Arc Ultra can move “almost a liter of air,” which puts it somewhere in the region of a 10-inch to 12-inch conventional subwoofer driver.
The system is also force-cancelling. That means that because you have two cones, each pushing air out in the opposite direction to the other while playing the exact same audio signal, most unwanted vibrations that would leak into the rest of the soundbar’s body are cancelled out because equal vibrations are being created in the opposite direction at the same time.
A sizable problem
Van Apeldoorn previously worked with Mayht, the company that originally developed the Sound Motion tech (known at the time as HeartMotion), and joined Sonos in the acquisition. “I joined Mayht as the first mechanical engineer of the team… once they had working prototypes, they approached me to join a team to see whether we could bring this to mass production.”
Van Apeldoorn’s role was to work out how “instead of building, like, tens of units, we’re talking hundreds of thousands of units – how do you design an assembly line?”
That question travelled with his team into the Sonos fold, but it wasn’t the first question. The first question was how to fit a square peg into a round hole.
“This was the first mass-producible version back at Mayht,” he says, showing me the picture above. “We call it T3. About three days after the acquisition, I flew with one of the founders to Santa Barbara, and we came into a room with like 30 engineers, designers, product management. And they had the old Sonos Arc on the table. And we showed up with this T3 device, and they were like, ‘Well, we want to have that performance, but inside that soundbar.”
The issue is obvious: the T3 was significantly larger than the Arc’s cross-section.
“There’s no way if you flip the transducer that it can fit inside that soundbar,” continues van Apeldoorn. “So the question was, can we translate this to a different shape, because a square or circular version is quite easy, but if you want to go for a racetrack design, that imposes some additional challenges to the technology.”
“We made more than 900 digital prototypes. We made almost a prototype every day”
One of the main “additional challenges” is that the design of the Sound Motion speaker requires magnetic coils on each side of the transducer to create its motion, and “the longer the distance between those two voice coils, the more difficult it is to make the transducer stable,” van Apeldoorn explains. And in the Arc Ultra, the magnets are at the far opposite corners of the rectangular Sound Motion design.
Why not put them closer? There was simply no space in the design to increase the depth of the speaker by adding the magnets on the shorter side, so they had to go at the ends. “We want to make it as stretched as possible. That’s a space restriction, but it also shows how adaptable this technology is,” van Apeldoorn adds.
“How did we solve that? We made more than 900 digital prototypes,” he says, wryly. He shows me a video of a huge line of variations on the Sound Motion design, stretching across a whole room. “This was just the first month of the acquisition. We made almost a prototype every day.”
“During the day, the engineers would run simulations and CAD design. In the evening, we put them in the 3D printers; next morning the 3D print came out. You hook it up to [a test machine], and those 3D-printed transducers, they maybe survived five to 10 minutes, but it’s all you need for those measurements in the beginning. So that allows us to iterate very quickly on those transducers without worrying too much about, like, injection-molding tolerances.”
From prototypes to reality
Developing these prototypes to build towards a real production plan is where being a part of Sonos came into its own. “What Sonos really helped with is that they have an awesome material collection and library, so we could load those into our simulations and then see what material is good for us,” says van Apeldoorn. “So the material used is very similar to what we used in the Sonos Era 100 and Era 300 [speaker] cones, for example. We could redesign knowing all the [material’s] specs.
“When you have a more powerful transducer in a smaller enclosure, your air pressure goes up as well, so we have to have a lot of ribs and reinforcement to prevent the cone from bulging, to be able to survive the long pressures over the lifetime of the product. So the cone also had to be reengineered for all the forces during the playback of music,” van Apeldoorn explains.
“We also want to make the transducer as light as possible – or at least, the passive parts – so we applied an algorithm with optimization that removed a lot of material, but we kept the same stiffness, and could improve the performance of this cone.”
Once they had a transducer design they liked, it could be tested in a proposed Arc Ultra design alongside the other speakers planned for inclusion, and it turned out that everyone was underestimating what they had on their hands.
“It was a kind of funny moment with our analysis and mechanical engineering team where the first prototype we made was leaky everywhere. You could hear the air coming out everywhere on the enclosure. Then I made a calculation on how much internal pressure we would be seeing with this transducer, and people were like…” van Apeldoorn smiles. “It’s not like they didn’t believe the calculations, they just found the result absurd that we were seeing there. So it took some iterations to get it fully airtight – that was a big structural change with this transducer.”
I asked whether van Apeldoorn’s team made any changes to pull back on the power of the transducer to help with the air pressure challenges, and he waved such an idea away. “No, no, we could solve those. We actually boosted some of the frequencies, so that’s pretty nice.”
“The first prototype we made was leaky everywhere. You could hear the air coming out everywhere on the enclosure…”
The next step was to work out how to mass-produce the new-look Sound Motion speaker designed for the Arc Ultra – but this was why van Apeldoorn was hired in the first place.
“Most of the work my team actually did on the Arc Ultra development wasn’t necessarily on the core Sound Motion [design], but we invented and designed new machines for magnetizing the magnets to inject more tension, for example. With 3D printing, a lot of things are easy, but the scaling step up – that’s really where our core focus was, and the technology really matured from that perspective, but the core [design principle] is still the same.”
Van Apeldoorn gives examples of how Sonos’ tooling team really helped to ensure that the long, flat design of the Sound Motion transducer could actually be flat without any warping, and noted that the suspension brackets used around the magnets to keep everything in place, even as it moves, are unique to this speaker completely.
“There was no company in the world who could make those,” says van Apeldoorn. “So we designed them ourselves, the whole manufacturing process and everything, the quality control, reliability.”
What’s next for Sound Motion
Naturally, I asked van Apeldoorn what products we can expect to see Sound Motion appear in next. Equally naturally, he declined to give me any specifics – but he was quite happy to talk about the overall potential for how it could be used in the future.
“Arc Ultra is the first implementation of how we want to push boundaries, but there are many other different form factors and possibilities there. We have distributed motors in all corners, so you see here four motors on [the Arc Ultra speaker’s] design. But some of the earlier prototypes have like six, eight, 10, or even more motors. So that’s something we experimented with … if you want to have more performance, you can have more motors in parallel, and that scaling allows you to really increase performance,” he says.
Would it make sense to build a full-size subwoofer using the technology if it can scale up this way, or would there be diminishing returns? “There’s definitely some room to scale the technology up there. We made some back at Mayht, for example,” van Apeldoorn muses. “I think the only diminishing return you would have is from a design perspective – at a certain point, it’s still just big.”
“We’re exploring all kinds of applications, whether it’s more bass, more compact form factors, more energy efficiency”
I also ask whether the technology only makes sense to replace bass drivers, as in the Arc Ultra, or whether it could be useful for higher frequencies too. Once again, the question is whether you’re saving any size, says van Apeldoorn. “You can play to higher frequencies, but the question is how much can we gain there. If you look at laptops, TVs, mobile phones, you can see that the mid or high frequencies, [the speakers] are already pretty small. So I feel like there’s less gain there, but we really lost that low frequency extension in products becoming slimmer.”
“We’re working on acoustic innovation, and we’re exploring all kinds of applications, whether it’s more bass, more compact form factors, more energy efficiency,” he says, which immediately gets me excited about the idea of a Sonos Roam 2 successor that includes this tech.
“The big question then becomes: what is the next experience we want to deliver to people,” says van Apeldoorn. “If you see the whole music industry now, Apple and such are providing Dolby Atmos content, you see that more and more movies too, you see more high-resolution audio coming up… These are the trends we’re also looking towards.”
Naphur van Apeldoorn
Naphur joined Sonos as part of the Mayht acquisition in 2022, and is now part of a newly formed research and development team working on the next generation of transducer technology. Naphur received his Master of Science in Mechanical Engineering from Delft University of Technology, specializing in Biorobotics and Haptic Interfaces. After graduation, he began his career in computer vision, exploring the intersection of art and technology by digitally reconstructing lost cultural heritage. With the rise of deep learning, he applied his image processing skills in the field of radiology while living in Japan. Intrigued by Mayht’s technology and after meeting the founder in Tokyo, he decided to return to hardware engineering, bringing with him the fast development approach from software.
