Signia MaX: Four DNNs Power a New Generation of Hearing Aids
Signia’s newest hearing aid platform is called MaX, short for Multi-adaptive Xperience, and the big headline is easy enough to understand: instead of relying on a single deep neural network to handle sound processing, Signia MaX uses four.
That’s noteworthy, because deep neural networks have quickly become one of the most important technologies in premium hearing aids. Oticon has used an always-on DNN for several generations. Phonak took a different approach with Audéo Sphere Infinio, adding a dedicated DEEPSONIC chip specifically for AI-based noise reduction. Other manufacturers have been moving in the same general direction.
Signia is now taking the idea a step further by giving several different neural networks different jobs, then having them work together as one system.
Four DNNs, Four Different Jobs
Before getting into the individual features, it helps to understand what Signia is trying to solve.
Most hearing aids are already very good at recognizing speech and reducing background noise. The challenge is that real conversations are rarely as simple as “speech in front, noise behind.”
Imagine dinner with five people at a busy restaurant. Someone across from you starts talking, then the person on your right jumps in. You answer them, a waiter walks up from the side, glasses are clinking behind you, and there’s music playing overhead.
A hearing aid has to make a lot of decisions very quickly.
Signia’s approach with MaX is to divide those decisions among four specialized DNNs. One focuses on the people you’re talking with. One recognizes your own voice. One analyzes the broader sound environment. And one handles unwanted noise.
Those four networks are coordinated through Acoustic Intelligence, rather than operating as four unrelated features.
The first is RealTime Conversation EnhancementAI, which is probably the most important of the four. Its job is to identify active speakers and help the hearing aids follow a conversation as it moves around you.
That matters most in group settings. Traditional directional microphones can work extremely well when the person you want to hear is sitting directly in front of you, but group conversations are more complicated. The person speaking keeps changing.
Signia says the latest version of RealTime Conversation Enhancement processes 75% more data points per second than before. On MaX 7, the hearing aids can also use three Dynamic AI Focus Streams to track important speech sources. MaX 5 gets one Focus Stream, while MaX 3 does not include that feature.
That tells you pretty clearly where Signia is putting its attention with MaX: difficult, multi-speaker conversations.
Signia Is Also Giving Your Own Voice Its Own DNN
The second network powers Own Voice ProcessingAI.
If you’ve ever tried hearing aids before, you may already know why this matters. Your own voice is one of the first things that can sound strange when you start wearing amplification. People often describe it as boomy, hollow, echoey, or simply unnatural.
Signia has been working on this problem for years with its Own Voice Processing technology. With MaX, the hearing aids now use a dedicated DNN to recognize when you are speaking and adjust the processing accordingly.
The goal is simple: when you talk, you should still sound like yourself.
I’ve always thought own-voice quality was one of the more underappreciated parts of a hearing aid fitting. We spend a lot of time talking about speech-in-noise scores and directional microphones, but if someone hates how their own voice sounds every time they speak, they’re probably not going to be thrilled with their hearing aids.
Signia cites internal data suggesting its newer own-voice processing can contribute to fewer hearing aid returns. As with any manufacturer-provided data, I’d like to see how that holds up as MaX gets into the hands of more patients, but it’s a sensible area for Signia to focus on.
The Other Two DNNs Handle Your Surroundings and Noise
The third network powers Ambient Scene AdaptationAI.
Hearing aids have classified listening environments for years. They can recognize that you’re in a quiet room, a noisy restaurant, a car, or another common environment and change their settings automatically.
MaX takes a more continuous approach.
Rather than simply dropping you into a fixed listening mode, Ambient Scene Adaptation is designed to keep analyzing the environment as it changes.
That distinction is important because real life doesn’t happen in neat little listening programs. A restaurant might be quiet when you sit down and loud 20 minutes later. You may walk from a quiet lobby into a crowded ballroom. A one-on-one conversation can suddenly turn into a six-person conversation.
The fourth DNN, Dynamic Noise ControlAI, focuses specifically on noise. Instead of treating every unwanted sound the same way, it attempts to identify the type and context of that noise and decide how aggressively it should be reduced.
I like the thinking behind that.
The goal of a hearing aid shouldn’t necessarily be to make the world as quiet as possible. Some environmental sound is useful. Sometimes you want the room to still sound alive; you just need enough separation between speech and everything else to follow the conversation.
Signia Trained Acoustic Intelligence on More Than 40 Million Sound Samples
Signia says Acoustic Intelligence was trained on more than 40 million sound samples, including 11 languages, six music genres, thousands of voices, and a wide range of real listening environments.
The system can also reportedly decode changes in a conversational environment in about 0.01 seconds.
Those numbers are interesting, but what matters is whether the extra processing actually helps someone hear better.
Signia has released its own study comparing MaX with several competing premium hearing aids in noisy group conversations. According to the company, 97% of participants heard better with MaX than with the key competitors tested.
That’s an impressive headline, but I’d put an asterisk next to it for now.
This is Signia’s own research supporting a Signia product launch. That doesn’t mean the findings should be dismissed, but I wouldn’t crown MaX the best hearing aid for speech in noise based on manufacturer testing alone.
I want to see independent lab results. I want to see customer reviews. I want to see return rates after these hearing aids have been on the market for a while.
Those things usually tell us more than a launch presentation does.
Still, the early results are promising, and the technology behind them is legitimately interesting.
Connectivity Is Much Better Thought Out
The other major upgrade with MaX is something that won’t get nearly as much attention as the four DNNs: Bluetooth.
Signia calls its new connectivity system OneConnect, and it gives MaX support for Bluetooth Classic, Bluetooth LE Audio, a telecoil, and eventually Auracast.
That combination should make MaX compatible with a much wider range of phones and audio devices without requiring consumers to understand the increasingly messy world of Bluetooth protocols.
For many newer phones, LE Audio will be the preferred option. Bluetooth Classic provides broader compatibility with devices that don’t support LE Audio. The telecoil remains useful in venues with hearing loops.
MaX is also built for Auracast, although there is one important caveat: Auracast is not enabled at launch. Signia says it will be activated through a firmware update delivered through the Signia One App or fitting software by March 2027.
So I’d call MaX Auracast-ready rather than Auracast-enabled today.
Pairing is also simpler. On compatible phones, pairing the hearing aids through the new Signia One App can establish both the app connection and audio streaming at the same time.
That sounds like a minor improvement until you’ve spent 40 minutes helping someone reconnect their hearing aids to a phone.
Battery Life, Charging, and the Signia One App
Despite all the extra AI processing happening inside MaX, battery life remains solid.
Signia rates Pure Charge&Go MaX for up to 30 hours of use without streaming, up to 28 hours with five hours of LE Audio streaming, and up to 22 hours with five hours of Bluetooth Classic streaming.
The hearing aids take about four hours for a full charge.
Signia also has two new chargers (shown further below on this page). The Desktop Charger 2.0 is the straightforward bedside option, while the Portable Charger Pure 2.0 contains its own battery and can provide up to four additional hearing aid charges. The portable charger also supports Qi wireless charging.
MaX uses the new Signia One App, where users can adjust volume and programs, change directional focus, access Signia Assistant, and receive support from their hearing care provider.
TeleCare has also improved. A hearing care professional can now make certain programming adjustments and send them to the patient asynchronously. In other words, both people no longer need to be on a live remote appointment at the same time. The adjustment arrives through the app, and the wearer can install it when convenient.
As a hearing aid specialist, I like this kind of feature. Not every small programming change should require another drive to the office.
My Take on Signia MaX
What interests me most about MaX isn’t simply that Signia managed to cram four DNNs into a hearing aid.
More AI doesn’t automatically mean better hearing.
The more important idea is that these neural networks are specialized. Signia isn’t asking one AI model to understand every part of a complex listening environment. It has separate networks concentrating on conversation, your own voice, ambient sound, and noise, with Acoustic Intelligence coordinating the results.
That approach makes a lot of sense to me.
It also feels like a sign of where the hearing aid industry is headed. We’re moving beyond the question of whether a hearing aid uses AI and toward a much more interesting question: what exactly is the AI doing?
MaX looks promising, particularly for people who struggle with group conversations in background noise. The connectivity package is also one of the more complete Signia has offered, battery life remains strong, and I’m glad they accomplished all of this without turning the hearing aid into something noticeably larger.
Now I want to see what happens outside Signia’s laboratory.
If the independent testing and real-world customer experiences come close to the early data, MaX could end up being a very meaningful release for Signia.