The digitization of audio systems in the last decades, coupled with the performance improvements of digital electronics, have made it possible to design complex digital audio systems around Linux Systems-on-Chip, instead of traditional DSPs. Using Linux as an OS gives access to a range of tools that make it possible to develop new connected audio features at a lower cost. However, the analog to digital audio transition brought up new issues, such as clocking and synchronization, which cannot always be efficiently solved in software. In this talk, we will explore how different hardware/software partitioning impacts the performance of Linux digital audio systems, and show different co-design options to mitigate these issues.
Philip-Dylan Gleonec, Savoir-faire Linux