Sales:
(856) 437-5753 – Call Anytime
Service:
(217) 803-4059 – Mon–Fri 9–5 EST
Skip to main content
Skip to navigation
Memorial Day Sale: Save $100 on Your Next Pair of Hearing Aids — Plus a FREE TV Connector with Qualifying Pairs! Contact Us or Order Online — Savings Applied Automatically at Checkout.
TL;DR: Phonak AI hearing aids use a dedicated deep learning chip called DEEPSONIC to separate speech from noise in real time, delivering up to 10 dB improvement in signal-to-noise performance and up to 45% reduction in listening effort. This is not a software update on existing hardware. It is a fundamentally different processing architecture. We fit the full Phonak AI lineup remotely, using the same manufacturer software traditional clinics use. Take our free online hearing test to get started.
If you have ever lost a conversation in a restaurant, struggled to follow a meeting while others seemed to manage fine, or arrived home from a noisy event exhausted in a way that has nothing to do with how much you enjoyed it — you understand what speech-in-noise difficulty actually costs.
Hearing loss compounds this in ways that go beyond immediate social discomfort. Research published via PMC found that people with hearing loss experience cognitive decline at a rate 30 to 40 percent faster than those with normal hearing, a finding that has driven significant urgency around earlier intervention and more effective hearing technology. The link between untreated hearing loss and accelerated cognitive decline is one of the strongest arguments in hearing health research for acting sooner rather than later.
These hearing aids are the most direct technological response to the speech-in-noise problem currently available. This guide explains exactly how the technology works, what it delivers in real-world use, and how we fit it remotely with professional precision. For a full overview of current Phonak models, our Phonak hearing aid innovations guide covers the complete lineup.
Before explaining what these devices do, it helps to understand why speech in noise is genuinely hard to solve, and why previous generations of hearing aids fell short.
The fundamental challenge is that speech and noise often occupy the same frequencies. A voice across a dinner table and the ambient conversation, clattering cutlery, and background music in the same restaurant all produce sound energy in the 500 Hz to 4,000 Hz range most important for speech intelligibility. Telling them apart requires more than simple amplification. It requires a sophisticated understanding of what speech sounds like relative to non-speech.
Earlier hearing aids approached this using directional microphones and digital noise reduction algorithms. These work by pointing a beam of sensitivity toward the front and reducing amplification of sounds arriving from other directions. In controlled conditions they perform well. In real-world environments with multiple simultaneous speakers, reflective surfaces, and unpredictable noise sources, they reach their limits. The brain has to do significant work to fill in what the hearing aids cannot resolve, which is what produces listening fatigue.
The current Phonak lineup addresses this with a fundamentally different architecture.
The breakthrough in the current Sphere Infinio Ultra is the separation of AI processing from amplification processing into two dedicated chips working in parallel.
Previous Phonak hearing aids, along with most current competitors, run AI and amplification on a single processor. This creates a tradeoff: the more computational work the chip does on speech-noise separation, the more power it consumes, and the more it diverts resources from other processing tasks. Manufacturers manage this tradeoff by using AI selectively, activating it only in certain environments or limiting the depth of analysis it can perform.
The Phonak Audéo Sphere Infinio Ultra changes this with the DEEPSONIC chip, a dedicated AI processor with 53 times more processing power than the previous generation Phonak processor. The DEEPSONIC chip handles one task only: running Spheric Speech Clarity, Phonak's proprietary deep learning model. Its output feeds into the ERA chip, which handles amplification and standard sound processing. Both chips run simultaneously rather than sequentially.
This means Spheric Speech Clarity can run at full depth in every listening environment without compromising amplification performance or battery efficiency. The AI is not rationed.
Spheric Speech Clarity is the deep learning model running on the DEEPSONIC chip. It was trained on a large dataset of real-world speech and noise combinations, and its job is to analyze the incoming audio stream and separate the speech signal from the noise signal before amplification occurs.
The performance numbers from clinical testing are significant. Devices running Spheric Speech Clarity deliver up to 10 dB improvement in signal-to-noise ratio compared to hearing aids without this technology. To put that in context, a 10 dB improvement in SNR is roughly equivalent to halving the perceived level of background noise relative to speech.
A 10 dB SNR improvement is a substantial, perceptible difference. Users typically describe the experience as conversations becoming followable rather than strained. Listening effort reduction matters as much as the SNR improvement for quality of life across a full day.
Listening effort reduction of up to 45% has been measured in independent clinical testing. This matters because listening fatigue is cumulative. It builds through the day and affects concentration, memory retrieval, and mood by evening. A 45% reduction in listening effort does not just mean a better experience at dinner. It means finishing the day with more cognitive reserves intact.
Spheric Speech Clarity activates automatically when the hearing aids detect a demanding noise environment. Users do not have to do anything. The system identifies the acoustic conditions, activates Spheric processing, and the improvement happens in real time.
The Sphere Infinio Ultra runs Spheric Speech Clarity on top of AutoSense OS 6.0, Phonak's automatic environment detection platform. AutoSense OS 6.0 continuously analyzes the acoustic environment and classifies it across scenarios including quiet, speech in noise, music, outdoor, and restaurant, adjusting the full processing profile of the hearing aids accordingly.
The relationship between Spheric Speech Clarity and AutoSense OS 6.0 is important to understand. AutoSense OS 6.0 manages the full hearing aid processing profile across all environments. Spheric Speech Clarity is specifically optimized for demanding speech-in-noise scenarios and activates within that AutoSense framework. In quieter environments, standard AutoSense processing handles the job efficiently. As noise levels rise and speech-in-noise demands increase, Spheric processing takes over.
This layered approach means the hearing aids deliver appropriate processing for every environment without requiring any manual switching from the user. The transition between modes is automatic and seamless.
The Sphere Infinio Ultra also includes StereoZoom 2.0, an updated directional microphone beamforming system. StereoZoom 2.0 creates a narrow, highly focused directional beam pointing toward the user's primary listening direction, combining input from both hearing aids into a coordinated directional response.
In previous Phonak generations, directional microphone technology and AI processing operated somewhat independently. In the Sphere Infinio Ultra, StereoZoom 2.0 feeds into the Spheric Speech Clarity processing pipeline, which means the directional information enhances the AI's ability to separate speech from noise. The combination is more effective than either technology working alone.
The practical result is that users in the most demanding environments, including crowded restaurants, multi-speaker meetings, and noisy social gatherings, can follow a conversation from a specific person even when surrounded by competing voices.
Even the best on-board AI processing has limits in truly extreme noise environments. The current lineup addresses this with RogerDirect, which allows direct streaming from Roger wireless microphones without an external receiver.
Roger microphones, including the PartnerMic, Roger Table Mic, and Roger Select, place the microphone close to the sound source, which fundamentally changes the signal-to-noise ratio before it ever reaches the hearing aid. A PartnerMic clipped to a conversation partner delivers their voice at a much higher level relative to ambient noise than any hearing aid microphone can achieve at a distance.
RogerDirect is built into all current Phonak models. No external receiver is required. For users in classrooms, meetings, or frequently in extreme noise environments, Roger accessories work alongside the on-board AI to extend performance into situations that on-board processing alone cannot fully resolve.
The Spheric Speech Clarity AI is exclusive to the Sphere Infinio Ultra. The Infinio Ultra R and Lumity carry Phonak's AI processing at different levels.
| Model | AI System | Speech-in-Noise | Roger |
|---|---|---|---|
| Audéo Sphere Infinio Ultra | DEEPSONIC + Spheric Speech Clarity | Up to 10 dB SNR improvement | RogerDirect built in |
| Audéo Infinio Ultra R | ERA chip + AutoSense OS 6.0 | Strong across most environments | RogerDirect built in |
| Audéo Lumity L90 | AutoSense OS 5.0 | Good across standard environments | Roger compatible |
The Sphere Infinio Ultra is the right choice for buyers who spend significant time in demanding speech-in-noise environments and want the maximum available processing power. The Infinio Ultra R delivers strong AutoSense OS 6.0 performance for most real-world environments at a lower price point. The Lumity L90 suits buyers who want proven Phonak AI performance at the most accessible price in the current lineup.
The AI processing in Phonak hearing aids is powerful, but it only reaches its full potential when the hearing aids are precisely programmed to a specific hearing profile. An under-fitted Sphere Infinio Ultra will underperform a well-fitted Lumity in most situations.
We fit every Phonak AI hearing aid we sell remotely using the same manufacturer software that traditional clinics use, with the same programming parameters, the same fitting rationale, and the same licensed hearing care providers. Here is what that process looks like:
Traditional clinics charge thousands more for the same devices you see here, require multiple in-person appointments, and often limit adjustments after your initial fitting. We provide licensed hearing care providers, the same manufacturer programming software clinics use, unlimited remote adjustments, and significant cost savings, all without leaving your home. Bare-bones online sellers ship pre-configured devices with no professional involvement, no licensed fitting, and minimal ongoing support.
Every Phonak AI hearing aid we sell comes with a 60-day risk-free trial.
Current devices in this lineup use the DEEPSONIC chip, a dedicated AI processor separate from the amplification chip. This dual-chip architecture allows the deep learning model, Spheric Speech Clarity, to run at full depth in every environment without tradeoffs in power or performance. Most hearing aids, including previous Phonak generations, run AI and amplification on a single chip. This limits how much processing power can be dedicated to speech-noise separation.
Clinical testing shows up to 10 dB improvement in signal-to-noise ratio and up to 45% reduction in listening effort for the Sphere Infinio Ultra running Spheric Speech Clarity. A 10 dB SNR improvement is a substantial, perceptible difference. Users typically describe the experience as conversations becoming followable rather than strained. The listening effort reduction matters as much as the SNR improvement for quality of life across a full day.
Automatic. Spheric Speech Clarity activates when the AutoSense OS 6.0 environment detection system identifies a demanding speech-in-noise scenario. Users do not switch modes manually. The transition is seamless and continuous throughout the day.
Yes. The Sphere Infinio Ultra, Infinio Ultra R, and Lumity all support mild to severe hearing loss with appropriate receiver options. The AI processing benefits apply across the full range of hearing loss levels. Spheric Speech Clarity improves speech-in-noise performance regardless of the degree of loss being corrected.
Yes, through the myPhonak app and through remote fitting with our licensed providers. The app allows you to see which AutoSense OS mode is active in real time and make manual adjustments to programs and EQ. Our providers can adjust Spheric Speech Clarity parameters, directional microphone focus, and noise reduction settings during remote sessions. Unlimited adjustments are included with every device we sell.
Phonak AI hearing aids represent a genuine step forward for anyone who has found that previous hearing aids helped at home but struggled in the situations that matter most. The DEEPSONIC chip and Spheric Speech Clarity address the core problem of speech in noise with a processing depth that was not previously possible in a wearable device.
The technology delivers its best results when it is professionally programmed to your specific hearing profile and environment by a licensed hearing care provider. That is what we do remotely for every person we work with, and we have been doing it for over 30 years.
Reach out to our hearing care experts and we will help you find the right Phonak AI hearing aid for your life.
