How To Program Embedded Audio and Sensor Data With Almost Zero Latency!Sunday, March 20, 2016
Bela is an open source platform for high performance ultra-low latency interactive audio that has a responsiveness of only 1 millisecond action to sound latency. Bela is capable of latency under 1ms, which is faster than any musician can detect, and faster than MIDI on MAC (5ms), iPhone 6+ (9ms), Arduino to Max (11ms) and PD (Pure Data) on Raspberry Pi (19ms).
Bela is able to achieve ultra low latency audio and sensor processing because the audio code does not go through the operating system (OS) like on most audio systems. Bela's custom audio environment does not have to compete for time with all the other processors on the board; hence, you get super low latency and no glitches because the processor is able to respond fast enough.
Bela is built on the Beaglebone Black and combines the power of an embedded computer with the hardware connectivity of a microcontroller. It's easy to use and has everything you need for audio and sensor processing in one self-contained package, making it ideal for embedded projects.
Bela is pretty much plug and play. Once you plug in the board into your computer, launch the on-board IDE on your browser, you can start coding and compiling in C++. When you are done editing, you can detach the board from your computer, power it with a battery and embed it in your project.
Bela runs on the Beaglebone Black with a a 1GHz ARM processor and gives stereo audio in and out (I/O) of 16-Bit/44.1kHz, as well as onboard speaker amplifiance (1W 8 ohms), 8 high quality 16-Bit analog in and 8 high quality 16-Bit analog out, and 16 digital I/O. As Belas runs on Linux you also get access to networking, storage, USB and other peripherals.