Google AI has unveiled Lyra, a new low-bitrate audio codec for speech compression. With the ongoing challenge of developing codecs that increase quality, using less data while minimizing latency, Lyra aims to make voice communication possible even on the slowest networks by leveraging advances in machine learning (ML). Using models trained on thousands of hours of data to create a novel method for compressing and transmitting voice signals, features or distinctive speech attributes, are extracted from speech every 40ms and then compressed for transmission. On the other end, a generative model uses those features to recreate the speech signal.

ADVERTORIAL: Read about how Dielectric is helping broadcast engineers demystify when and where problems may arise in the RF infrastructure. Dielectric’s RFHawkeye is the broadcast industry’s first native IP-connected RF monitoring system for antennas, transmission line, and most external associated RF components. Broadcasters not only have an IP connection to remotely monitor antenna and line system performance and condition; their predictive capabilities are optimized with a graphical user interface that presents meaningful, actionable system data. Read more here.