Tamura Kent formally unveils his plans to add new functionality to NetBSD’s audio framework (audio(9) and audio(4)) – specifically, the addition of an audio converter pipeline and in-kernel mixing. These additions along with audio device cloning would make it possible to natively support hardware mixing without the use of a software based soundserver.
NetBSD: Ideas on Audio Framework and Hardware Mixing
2004-12-11 NetBSD 3 Comments
One can do hw audio mixing on the video card
While new to the NetBSD platform, the OSS Live/Audigy driver for Linux systems already supports multiple opens of /dev/dsp, which esentially allows hardware mixing without a software based soundserver.
However, it is a widely known problem that most audio drivers for *nix systems do not support hardware mixing for audio, and I applaud this move forward by the NetBSD platform. No other *BSD systems that I can see has even attempted to rework their audio architecture in a meaningful, significant way for quite some time.
FreeBSD already supports multiple opens of a sound device, you just have to configure “vchans” (or in 5.x, its automatically done for you).