SoundGrid 301 Part 3: SoundGrid System Components in Depth


Please note: Completing SoundGrid 201 is a prerequisite to taking the SoundGrid 301 test.

Every SoundGrid system needs the following two components:

  1. A host computer with an Ethernet connection, running a host SoundGrid application such as SoundGrid Studio, SuperRack or eMotion LV1
  2. A SoundGrid-compatible audio interface

With these two components, you have the basis of a SoundGrid system that can network audio between hardware devices as well as software devices. However, this setup does not provide low-latency monitoring as there is no SoundGrid DSP server connected.

A SoundGrid setup that has only these two components will nearly always be found in a studio environment, not a live sound environment.

For optimal use of SoundGrid in live sound and broadcast, you will need a SoundGrid system that consists of at least the following four components:

  1. A host computer with an Ethernet connection, running a host SoundGrid application
  2. One or more SoundGrid-compatible audio interfaces
  3. A SoundGrid-compatible switch; and
  4. A SoundGrid DSP server

This is the most popular way to work with SoundGrid in a live setting.

You can also add more components to expand the system’s capabilities:

  1. One or more computers with an Ethernet connection, running a SoundGrid driver
  2. A second server for processing redundancy

The Host Computer

A SoundGrid system’s host computer runs the SoundGrid host application. The host application lets you manage your hardware inventory, clocking and patching.

With the latest versions of Waves’ eMotion LV1 live software mixer and SuperRack plugin rack, you can use a feature called “I/O sharing” to share hardware and software I/Os between separate SoundGrid systems over the SoundGrid network, via the computer’s Ethernet port or adapter.

The SoundGrid DSP Server

The SoundGrid DSP server is the processing unit for the SoundGrid network. It is designed to perform low-latency audio processing. SoundGrid servers are all based on the following internal parts:

  1. An Intel chip
  2. A network interface card
  3. An operating system written by Waves on top of a Linux kernel

Only one server at a time can be used for DSP processing in a SoundGrid setup. You cannot double up your DSP power by adding a second SoundGrid server.

In the studio, there is no reason to add another server for redundancy. Of course, if you have different rooms and systems, you might want a second server to be used separately in a different room.

In live sound & broadcast, when using the SuperRack, MultiRack or eMotion LV1 host applications, a second server added to the inventory will automatically become the redundant server, meaning it will take over if the first server fails.

Make sure that your redundant server can take the same load as the main one. If the first and primary DSP server is a SoundGrid Extreme-C, make sure the redundant server is also a SoundGrid Extreme-C.

In SuperRack and MultiRack, you can use the Test Redundancy feature to make sure that your setup is reliable. This feature will check whether your second server is up to the task of performing if required.

SoundGrid-Compatible Switches

In the support pages on, we have a list of qualified 1Gb network switches to use with your SoundGrid network. Waves can only support switches that have been tested and qualified, so please check the list to make sure the switch you have chosen is supported.

SoundGrid requires a private Ethernet network. It cannot run alongside other types of networks such as internet, wi-fi, or other AV networks, as it requires the entire bandwidth.

Waves Audio supports up to four daisy-chained switches between any two points in the SoundGrid network. In most cases two switches will cover all your needs.

Daisy-chaining SoundGrid devices directly is not recommended. Instead, each device should be connected directly to a switch. This applies even to SoundGrid devices with two Ethernet ports, such as DiGiGrid IOS-XL or DiGiGrid IOX.

Most switches use what is called a “store-and-forward” algorithm. In practice, your SoundGrid-compatible switch gets the frame (your audio) into its port, stores it, then forwards it on to its destination. This process takes only a few microseconds and is completely inaudible. Each switch you place in the path of your audio adds more delay. This is the reason for the four-switch limit.