Here are a few pointers regarding routing between a console and a WFS processor. The console is there to preprocess all the signals from the microphones, recorded or synthesis tracks and send them to the WFS processing. After this the output signals have different options depending on the console’s possibilities.
When using WFS for live stage work (concert, opera, theatre or dance piece…) it is advisable to go through a console. This way you can take advantage of the built-in signal processing (Eq, dynamics…) and the ability to group several tracks (for a drum kit or the sends to a reverb for example).
After this the signal is sent to the WFS processor through the post-fader direct-outs of the input channels. This way it is also possible to control the audio level of each track with a fader.
The type of digital multichannel protocol depends on the console, the computer and its soundcard. In any case we will try to achieve the lowest latency.
Some consoles have USB ports for multichannel sends and receive with computers. It is generaly the cheapest option if the internal routing allows patching these I/O where we need them: post fader direct-outs for sends to the WFS processor and either to some direct input to the mix channels (cascade/merge/ext.input or insert) or to unused input channels.
However the USB layer has its own rather noticeable latency on top of the driver’s buffers.
Then you have audio-over-IP protocols, ie. Dante, AVB, Ravena, AES67… These are pretty fast, but the asynchonous nature of the network exchanges the native drivers such as Dante Virtual Soundcard (DVS) have large buffers (4 to 10ms).
There are some dedicated PCI-express cards with very low latencies.
this option can be very interesting if the dispatch to the amplifiers use the same protocol.
MADI with either a PCI-express or Thunderbolt soundcard will probably be the most stable and fast option for a large number of channels. But it is also the most expensive.
There are some MADI-USB sound cards too, but once again the USB layer has its own latency on top of the drivers buffers.
It’s also possibleto go through ADAT, but given the number of channels quite a few fibers will be needed.
AES50 interfaces are rares for Midas and Behringer consoles.
It might be easier to convert back and forth to an other digital audio protocol with a greater choice of sound cards on the computer side.
After the WFS processing there are several alternatives depending on the capabilities of the console and other necessities:
- Some consoles have a lot of output buses, others have a limited number.
- Some consoles have restrictions as to what can be routed direction into a bus. You might also find direct inputs into a bus, ie. cascade/merge/ext.input, insert returns…
- You might also want to mix some inputs in a classical way and others through the WFS process.
- In case something goes wrong with the WFS process you might need a quick solution to go back to a standard mix with all inputs going through the console’s summation.
Some consoles can mix a signal into the an aux or group. Eg.: DiGiCo SD ; Allen&Heath dLive, iLive, GLD ; Midas Pro ; Yamaha 01V96*, DM1000, …
In this case it is very easy to mix the return from the WFS process with a regular mix through the desk. You’ll have to make sure you don’t mix the same track through both so as not to get some horrible phasing.
this is a very elegant routing solution, but it may not be the easiest when something goes wrong with the WFS processing. In this case all tracks sent to the WFS have to be reassigned to the classical mix path. A copy of a stereo mix assigned to the mix channels left muted until an eventual crash can be a very handy back-up.
An alternative is to use the insert returns on the mix channels. If the computer’s sound card has a built-in mixer you can even merge the classical mix with the WFS.
In the eventuality of a problem with the WFS process all you need to do is to deactivate the insert to return to the normal mix. As in the previous case if you mix WFS and classical mix you might have to keep a classical mix of all tracks sent to the WFS assigned to the output channels. Keep it muted as long as you don’t need them for back up.
Other consoles do not have this possibility and will require going through some input channels for the returns of the WFS process. This will increase a bit the latency of the whole system since you have to go a second time around through input channels.
And finally if you have too many speakers to feed, you may not have enough channels on the console. In this case the outputs of the WFS process is sent through converters to the amplifiers.
This requires there are some Eqs and dynamic processing in the WFS processor. In cas of a problem with the WFS processor there is no easy way to reassign the console’s outputs to the processor…
Wave Field Synthesis DIY 