Skip to content

hifiberry/hifiberry-dsp

Repository files navigation

Python package PyPI version PyPI license PyPI download month GitHub contributors

HiFiberry DSP

Software for HiFiBerry boards equipped with DSP. This package can be used to read/write data to HiFiBerry DSP boards using the Beocreate TCP server.

The software comes "as-is". There is no individual support for this software. Feel free to post in the HiFiBerry forum for questions.

Installation

Before installing the dsptoolkit, you need to have a working Python 3 installation (>=3.10) and a working pip. You also need to enable SPI as dsptoolkit needs SPI to communicate with the DSP. Check out https://www.raspberrypi-spy.co.uk/2014/08/enabling-the-spi-interface-on-the-raspberry-pi/

You can then install the toolkit by just running

bash <(curl https://raw.githubusercontent.com/hifiberry/hifiberry-dsp/master/install-dsptoolkit)

sigmatcpserver

This server runs in background and provides a TCP interface (port 8089) to access DSP functions. It is compatible with SigmaStudio. That means you can directly deploy DSP programs from SigmaStudio and change parameters online.

Command line utility

The dsptoolkit command is the main command line tool to communicate with the DSP TCP server. The command line parameters are documented in doc/dsptoolkit.md

REW integration

The software can be used to push filters created by Room Equalisation Wizard (REW) to the DSP. Have a look at the guide indoc/rew-basic.md

DSP profile format

DSP profiles can be generated directly in SigmaStudio. However, to enable the full potential of DSP Profiles and allow DSPToolkit to directly control the DSP program, you need to add some additional metadata to the XML file. The process to create a DSP profile is documented in doc/dspprofiles.md