Case studies about real time digital audio processing using highly available low cost computational platforms

Presenter: André Jucovsky Bianchi

In this masters work we explore different possibilities of real time digital audio processing using platforms that are highly available and have relatively low cost.

Arduinos are minimal structure for interaction with ATmega microcontrollers and are generally used as control interface for other eletric or eletronic devices. Because it has pins capable of ADC and DAC, it can be used to capture, process and emit analogic signals.

GPUs are parallel processing cards whose structure evolved from the traditional graphic processing pipeline. They have hunderds of processors that operate in parallel in its own memory. Thus, they can operate in many channels at the same time, or use some audio processing algorithms' inherent parallel properties.

Mobile devices are becoming more present and, because they have the capability of capturing and emiting audio, they can be explored as platforms for audio processing in real time. In this context, it is interesting to analyse the performance of different mobile devices for common tasks in audio processing.

In this seminar, we will present the platforms described above and results obtained when performing real time audio processing in each of them.

When: May 8th, 2013