[Introduction]Since the launch of the first-generation AirPods in 2016, after four or five years of rapid development, the TWS headset industry has gradually moved from immaturity to maturity. Relevant data show that the global TWS headset shipments surged by 25 times in the five years from 2016 to 2020, with the growth rate reaching its peak in 2019 and a drop to 24% in 2021. When the industry is facing an “inflection point”, how to break through innovation and leverage new growth for TWS headsets entering the “second half” is an urgent issue for the entire industry to solve.
The immediate needs from the user side have actually pointed the way. The latest user research report in the industry pointed out that in terms of “the potential impact of advanced functional features ON the purchase decision of true wireless earbuds”, active noise reduction ranked first with 71% of the votes. When active noise reduction has become a “just-needed” function of the TWS headset market competition, how should the industry chain grasp the key to the “second half” competition in this industry? Today, we take the latest solution of the world’s leading high-performance semiconductor company Analog Devices as an example to see how high-performance chip and algorithm solutions help downstream customers to introduce products and quickly bring them to the market from the perspective of upstream manufacturers.
To meet the strong demand of TWS headphones, how to build a high-performance noise reduction chip?
ANC implementation principle
In 1978, the noise of an airplane engine prompted Bose founder Dr. Amar G. Bose to create the first noise-cancelling headphones. So far, there are two main ways to reduce noise in headphones: active noise reduction and passive noise reduction. Among them, passive noise-cancelling headphones mainly form a closed space by surrounding the ears, or use sound insulation materials such as silicone earplugs to block external noise. Active noise reduction, on the other hand, generates reverse sound waves equal to the external noise through the noise reduction system, neutralizing the noise to achieve the effect of noise reduction. In contrast, passive noise reduction generally can only block high-frequency noise, and it is not obvious for low-frequency noise reduction. Active noise reduction with more outstanding effects has naturally become the darling of the industry.
However, the seemingly simple principle of active noise reduction faces many challenges in the actual design process. For example, how to accurately sense noise, how to ensure that noise sound waves are completely canceled, and so on. To this end, digital active noise reduction derives three different active noise reduction technologies: FeedForward, Feedback, and Hybrid, a combination of the two. In contrast, hybrid active noise reduction can achieve faster and better noise reduction, and rely on a better listening experience to become the first choice of many downstream manufacturers.
Simple block diagram of ADAU1860
ADAU1860 is a new generation of low-power active noise reduction chip launched by ADI. It is mainly composed of three parts: the first part is the ADC, which collects the external sound source and converts it into a digital signal; the middle part is two very powerful DSPs, one is the ADC. The first one is HiFi-3z DSP, which mainly realizes the recognition of audio scenes, such as uplink noise reduction and voice wake-up. The other is Fast DSP, which is extended to 96-bit instruction set. In addition to completing a set of filter algorithms of ANC, it can achieve more functions such as noise floor suppression and EQ adjustment; the third part is high-performance. DAC, the dual power rail setting of 1.2V and 1.8V on the power supply can further optimize the power consumption according to the different usage scenarios of customers.
In addition, compared with ADAU1860, ADAU1850 introduced in the same period reduces Hifi-3Z DSP, only retains Fast DSP, has three inputs and one output, and includes a digital signal processor. At the same time, the path from the analog input to the DSP core to the analog output has been optimized for low latency. By adding a small number of passive components, the ADAU1850 can provide a complete headphone solution, providing the possibility of more accurate noise reduction at a lower cost. Overall, ADI’s new generation of active noise cancellation chips can provide reliable digital signal processing solutions for many audio and speech applications, including speech enhancement, noise reduction, echo cancellation, level control, acoustic feedback suppression, microphone arrays and Source noise and vibration control, downstream manufacturers can use this to create differentiated features and selling points of earphones, and capture more consumers in the “second half” competition of TWS earphones.
Helping to speed up product development, both software and hardware are indispensable
“A good horse with a good saddle”. In addition to providing high-performance chips at the hardware level, the supporting algorithm solutions and software support are also the key support for the development of an excellent active noise-cancelling TWS headset. Since ADI’s software development tools and some algorithms are implemented locally, it can help customers design better products using ADI’s development tools in a very short period of time.
For example, in terms of algorithms, ADI has implemented adaptive ANC noise reduction, adaptive tuning and spatial audio. ADI has independently developed an adaptive noise reduction system, which is realized by combining software and hardware and taking advantage of the hardware instruction set; adaptive tuning will automatically and real-time calculate and adjust the tuning according to the user’s wearing state. For the audio module, when the user’s earphone leaks, the algorithm will dynamically compensate for the audio loss caused by the leak. As for spatial audio, ADI provides from the processing of the metadata of the gyroscope itself to the realization of the spatial effect of the final sound.
Cadence Xtensa DSP Software Development Tool
At the same time, ADI’s free audio processor programming, development and debugging software Lark Studio tool provides a variety of signal processing algorithms integrated into an intuitive Graphical User Interface (GUI), enabling the creation of complex audio signal flows. Programmed but acoustically proficient customer groups can also use the software to easily accomplish tasks. Among them, Lark Studio is a good online real-time tuning platform. Customers can do pre-development based on Lark Studio, including the completion of prototyping.
Accelerate market penetration, so that ANC is no longer exclusive to high-end
Although the launch of Apple Airpods Pro in 2019 made the public no longer unfamiliar with ANC active noise reduction, in recent years, this function has mainly appeared in mid-to-high-end TWS headphones, and more popular TWS headphones are rarely equipped. The reason is that ANC has many technical barriers in development and production. In addition to the hardware reference design, ADI’s noise reduction solution also strengthens the software reference guide, making it easier for customers to design and mass production, and helping small and medium-sized enterprises to smoothly introduce the ANC solution.
The data of market research institutions have shown that with the gradual filling of the market and the increase of the base, the volume of TWS headphones is gradually slowing down, and the “inflection point” of the industry has reached. As an important product direction of ADI China, digital audio chips have been cooperating with mainstream manufacturers and have been highly recognized, and will further strengthen the cooperation with the upstream and downstream of the industrial ecology to optimize the ANC solution to make it better performance and easier to use. , benefit more market participants, and jointly promote the popularization of ANC function in TWS headsets.