Efficient low latency audio signal compression, which is crucial to wireless headphone user experience, must forgo traditional long block transforms and depends critically on accurate signal prediction capabilities. However, current techniques either ignore prediction altogether (e.g., Bluetooth's sub-band coding) or employ basic and limited prediction, at considerable cost in audio quality and/or battery life. Effective prediction of polyphonic audio is widely recognized as highly challenging and has, until now, resisted many attempts that employed conventional prediction techniques. InfoCoding Labs’ proprietary technology completely solves the problem with an optimal prediction solution that is tailored to polyphonic signals, and which is harnessed in a new generation of highly efficient low-delay audio codecs for wireless headphone applications.
At the heart of InfoCoding Labs' "nextgen" wireless headphone technology are its proprietary "cascaded long term prediction" (CLTP) filters. CLTP filters adapt to and separate the multiple periodic components of the polyphonic music signal, to ultimately achieve the optimal prediction of future signal samples which, in turn, enables the codec to fully exploit all temporal correlations in the signal and hence dramatically enhance the reconstructed audio quality and reduce bandwidth and energy requirements.
Beside the focus on nextgen low-latency compression benefits to the wireless headphone application, the technology itself is fundamental and applicable to a whole slew of application fields where low latency and high fidelity are required. These include a variety of applications involving real-time communication systems (in the personal, commercial, defense and healthcare sectors), live media interactions (e.g., interactive broadcast involving reporters in the field), wireless virtual/augmented reality gear, audio for multi-player networked gaming, real-time distributed music collaboration, and interactive education applications.