First, if you are on Windows, you should make sure to follow these few steps, if not done already.
There are several ways to do what you’re aiming to, but some are better than others :
The easiest way (but not recommended) would be to enable Loopback on your interface, and set the loopback channel as the input in Discord. Some interfaces have both a dedicated Virtual channel and an option to merge it with the 1/2 input. If you want to use a mic without having to monitor it through the output, it’s this second option that should interest you.
It will transmit everything that goes into the interface output (in short, everything that you hear) plus the mics that are plugged into 1/2, but then, that would be problematic since, when other people are talking into Discord or such apps, they would hear themselves in return.
I let you imagine if two or more users are using loopback at the same time, that would create an endless feedback loop, which is not what you want.
A workaround can be to have a Focusrite interface with S/PDIF connectors and use physical loopback (put the S/PDIF Out into the S/PDIF In) and do the appropriate routing in the Focusrite Control mixer so that only the outputs of Cubase and your Mic go into the loopback, and select the S/PDIF Input in Discord. That way other people won’t hear themselves, but it is a bit more complicated to setup and you need an appropriate interface.
The better way is to install Sonobus and Virtual Cable.
- In Cubase, add a VST instance of Sonobus on your Output Channel or in the Control Room.
- Add another instance of Sonobus on your Mic Input Channel.
If you have a USB or wireless microphone, then use a standalone instance instead, but then only set the input (no need to set an output).
- Open Sonobus as standalone (you should now be having three instances), and select CABLE Input as the output (do not set any input).
- Connect the three Sonobus instances together.
- In the Control Room instance, mute your Mic instance and vice versa so they don’t feed back into each other (this is critical).
- In Discord, select CABLE Output as the Input, and the interface output as the output (same as Cubase).
- Start voice chat with a friend and test if that works.
This way you have free control over your Mic and Cubase levels with the standalone instance of Sonobus, without the hassle of loopback.
For the microphone, you can eventually add other plugins in the chain before Sonobus, so your voice can benefit from gate, compression, etc, but keep in mind that these effects will be printed to the recording in case you record the mic in Cubase, so don’t forget to bypass the effects accordingly.
It’s good to note that Sonobus itself doesn’t affect the chain, it acts like a true bypass and only “captures” the signal that passes through to send it to other Sonobus instances, and “outputs” the signal it receives from other instances.
Additional note : In Discord when you do a mic test with Virtual Cable, you may hear a destroyed signal in return, but apparently that’s the mic test that is broken, and not your actual signal, as people will still hear it clean (a quick search on google can confirm this). Also you must disable all the enhancements like echo cancellation, noise reduction, auto gain, etc.
The best way is to ask your friends or clients to download and use Sonobus instead of Discord, Zoom or other platforms. As you can stream audio up to 32 bit PCM, they can benefit from the best possible audio quality if your - and their - connection is fast enough.
Even though Sonobus features a chat room, you can still use Discord or Zoom in parallel for this purpose and for webcam/screen sharing, but then all group members must mute their audio in these apps.
The setup is the same as the above method, except that you don’t need the third standalone instance. You should only make sure to mute every channel in the Mic instance so that the audio from other people only goes out from the output channel or control room.
Just saying that Discord is terrible due to the simple fact that it only uses mono audio.
On the other hand, Zoom does have stereo audio, but for music stuff where audio quality is a priority, then Sonobus is the way to go.
And not to mention that with Sonobus you can also remotely record other people directly inside your DAW, just like VST Connect as mentioned by Martin. It only lacks MIDI features for now.
As suggested by @Martin.Jirsak, VST Connect is the right choice when it comes to remote recording and listening that requires critical stability and performance.
It is very well integrated into Cubase as it takes the network latency into consideration when recording, so that the recorded data is always aligned to the grid even if the performer and the producer live on different continents.
The only drawback is that it can only be used by two persons, the producer and one performer, and it doesn’t work both ways at the same time unlike Sonobus — but then Sonobus doesn’t have latency compensation, so you choose.
Optional : Gig Performer is an awesome software when it comes to mixing and routing multiple audio sources, specifically when used in conjunction with Sonobus, Virtual Cable and ASIO4ALL. You can add plugins wherever in the signal path, making it the ultimate sandbox
Please let us know if it works