Enabling ANC in Open-Ear Earbuds Beyond Apple with GreenWaves GAP9 AI+DSP Processor

Open-ear earbuds are an emerging design that has recently gained much visibility with the launch by Apple of their AirPods 4. The open-ear design provides comfort for extended use and enhances situational awareness by allowing ambient sound to enter the ear canal. As people wear earbuds for increasingly longer hours, open-ear designs are gaining traction, particularly in Asia.

Active Noise Cancellation (ANC) is now an anticipated feature in most earbuds. However, implementing ANC in open-ear designs is a real challenge, which Apple claims to have solved for loose-fit designs for the first time with the AirPods 4 thanks to their proprietary H2 processor. The effectiveness is half that of their sealed AirPods Pro 2. The primary difficulties in implementing Active Noise Cancellation (ANC) for open-ear designs are multifaceted:

1. Noise Cancellation: ANC involves both passive and active methods. In an open-ear design, passive cancellation is by definition almost nonexistent, allowing more external noise to leak into the ear.

2. Anatomical Variations: Each person has a unique ear shape, causing variations in how external noise leaks into the ear.

3. Environmental Sensitivity: The effectiveness of ANC is highly sensitive to environmental changes. For example, a gust of wind can create additional noise, and the ANC system must quickly adjust to compensate for it.

As a high-gain closed-loop system with such highly variable externalities, the system’s stability margin must be closely monitored and adjusted within a very tight timeframe to avoid howling departures. To achieve this goal, a neural network trained on corresponding situations has been proven to be far more effective than classic DSP solutions.  Considering the targeted ANC bandwidth, the latency of the adaptation loop must be sub-millisecond, which in turn poses an architectural challenge for the processor. 

By uniquely meeting this sub-millisecond latency requirement, the ultra-low-latency shared memory AI+DSP processor GAP9 enables brands, beyond Apple, to implement active noise cancellation (ANC) in open-ear designs.

Specifically, the GAP9 runs a highly reactive neural network that monitors the howling margin every millisecond and dynamically triggers the re-parameterization of the ANC graph running on GAP9’s ultra-low-latency Smart Filtering Unit.  

The GAP9’s architecture allows for the seamless integration of multiple advanced features in addition to ANC such as AI-ENC, AI-transparency or AI- F2F denoising. 

A processor is only as valuable as the software that runs on it. Our strategic partner Orosound Labs is among the very few companies that master AI-controlled active noise cancellation and reduction technologies, among other leading-edge features. Orosound Labs has been recognised by Frost & Sullivan as the first to market in delivering effective active noise cancellation and uplink noise reduction for open-ear headphones. Our collaboration has culminated in a full solution integrating their advanced features on GAP9. This joint offer drastically shortens the development of high-end hearables including open-ear designs.

ANC in open-ear TWS running on the GAP9 processor is commercially available and represents a crucial feature leap for vendors that defines the next generation of open-ear designs.