One of the important part of the video games is the more realistic surround sound, which went through a significant development, due to the spread of the Virtual reality, in the past few years. Creating a 3D audio with better quality is computationally intensive, therefore the goal of my thesis was to create an FPGA design, which is capable to perform these calculations instead of processor.
In the first part of the thesis I review how the 3D audio works and how a modern development environment create a 3D audio. Nowdays AMD and NVIDIA also provide hardware acceleration for real-time processing. AMD TrueAudio Next filters in a frequency domain using an uniformly partitioned convolution algorithm. It offers a solution to perform a large size Fourier transform efficiently and with low latency.
In the second part of the thesis I present the created FPGA design. It was developed with Xilinx Vivado HLS which allowed to create hardware design in high level C language. The final acceleration card is able to process 20 channels of audio using 5 second impulse responses.