I have a weird use-case where I have my Windows desktop and a Mac Mini where I want to use simultaneously so that the audio is played on my Audeze Maxwell headset over USB. Currently I switch the connectivity of the USB port for the headset using a Ugreen USB switch, but I want a solution where I can get the audio from both machines into the headset simultaneously.
Previously I tried quite a a few combinations of solutions, none of which really worked out.
Piping the audio from one machine to the other over 3.5mm or optical and having the receiving machine handle the mixing for both machines. This doesn’t work well as the receiving machine needs to be on at all times which isn’t necessarily possible.
Same as option 1, but I used a dedicated SFF / NUC-like machine with an Intel N105 chip. The audio worked, but the quality was surprisingly poor with noticeable delay. I had also connected my dedicated XLR microhpone connected to this system, and running any sort of VST’s like an equalizer or noise gates on Windows were enough to overload the system and cause a lot of audio popping.
Same as option 2, but I used a Digiface USB with the Digiface RME software on the SFF system. This reduced the load on the system and decreased the latency of the audio, but the same issues arose with the VST’s and audio quality.
Are there any solutions to this issue, or should I stick with switching which machine my headset is connected to using the USB switch / a KVM?
In general, any audio interface with 4 or more input channels will do, the two I linked cost more but are digital and will allow for mixing the inputs to different outputs without the need of both PCs to be on, the rode one will give you more options, both will provide all channel input separated and without mixing to your main PC in case you want to use a daw(multi track mode)…
This will still hold true, if you use the headset in USB wireless mode the machine to which you connect it will need to receive the main mix… and be on at all times, if you use Bluetooth you will be able to use the rodecaster, same if you use the trrs cable…
So even that Rode mixer requires a primary PC to handle the audio. Seems like my previous plan of having some dedicated Windows PC handle the audio mixing, where I pipe the audio from other machines to it.
The main issues for me are:
Having a powerful enough machine to handle VST’s for both the microphone and the headset output, without lagging / crackling the audio, and while remaining as low power as possible
Get audio from other machines with low latency
Get audio from other machines without crackling / EMI
I previously fixed issue number 3 by using optical cables, but that introduced a lot of latency like issue number 2.
I fixed issue number 2 by using the Digiface USB, but it it kept having sync issues between inputs that would essentially bring back issue number 2.
And issue number 1 would only be solved by getting some decent PC hardware to host this setup, but I’m not keen on running another computer in my room with my homelab since it’s already hot as-is. Not sure where to go from here on this
What about a cheap single board computer like a PI5 with a couple of DACs attached and your headset. That could be always on without the heat / noise and give you other options for capture.
I confess I’ve not tried it but it could be a neat solution. My situation is a little different in that I needed my mic to work in two separate VMs for different client zoom calls. I have up in the end and just got two mics.
Maybe it would be better if you laid down in plain English what you want to achieve, with the whole audio chain ?
You have an USB headset that you want to use (simultaneously, one at a time)? On two different machines.
You are doing vst processing on (mic/audio out)? On (one of ,both PCs, one at a time)?
You want to keep the headset connected using the 2.4ghz dongle, and use both the output and the mic input?
If the use case is the headset connected to one PC at a time, then switching manually the USB dongle is the least resistance/least cost/more stable path
Switching the dongle using a KVM will lead to the interface sometimes not being detected
Connecting the USB dongle to an audio interface directly will not work because the USB dongle is in reality an audio interface…
I did try this with that Celeron N105 which is more powerful than a Raspberry Pi (not to mention that the Pi is ARM-based and isn’t going to run Windows, which has more support for audio routing / mixing and such over Linux). The N105 couldn’t handle some basic VST’s like equalizers or noise gates, and I would expect a Raspberry Pi 4 or 5 to do any better.
Fair enough, let me outline the pain points that have led me to this idea and what my chain currently looks like.
I currently use my Windows desktop as my primary computer. I’ve been moving my general tasks over to the Mac Mini since it’s been more stable and performant. Essentially, if it’s not a Windows-specific task like gaming or programming for Windows specifically, I’ve been doing it on the Mac Mini.
I have a MacBook Pro for work that I connect to a dock, and use an Audeze Maxwell headset. For work calls, I use the bluetooth mode and for my desktop and Mac Mini I use the USB connection of the headset. The Audeze Maxwell does not allow you to play audio from both USB and bluetooth at the same time.
I have a very weird setup overall. I have 3 monitors (I’ll just call them Mon1, Mon2, and Mon3) where Mon1 and Mon2 are connected to a Level 1 Techs 2-monitor 2-system DP KVM, whereas Mon3 is on a Level 1 Techs 1-monitor 4-system HDMI KVM. The DP KVM is just between my Windows PC and the Mac Mini, while the HDMI KVM switches between the Windows PC, Mac Mini, and Macbook.
I do this so that I can use one of the systems through the DP KVM while using another system on the HDMI KVM, such as when I am doing work on the work Macbook while listing to music or videos on either the Windows PC or Mac Mini. The headset USB dongle is connected to the DP KVM so that it switches to the active computer.
The problem with this setup is these two issues:
Windows will always try to use the most recently connected audio device as the audio output and input devices. This would be fine if I wasn’t using something like the Elgato Wave Link software, where I need the Elgato virtual audio devices to remain as the default instead of the actual headset device.
I can’t get audio from multiple machines at the same time. I’m fortunate enough that I have a 1st-world-problem of having a job that I can do from home in my own setup, but I’d like to have audio available from both whatever the primary machine is (the Windows PC or the Mac Mini) while also being able to hear any audio coming from the work Macbook. Another good example is that with software engineering at a big company, I get stuck in a lot of meeting where I need to be in those meetings but don’t actually need to participate. Being able to hear both those meetings and run music or videos in the background would be amazing for my sanity
Ok, thanks for the detailed explanation .
You mentioned a separate mic? Or are you using the headset microphone as well? Also, you mentioned audio plugins/VSTs , do you run them on both machines?
So this is another silly thing. I do have a dedicated mic connected through an Elgato Wave XLR, which is also plugged into the DP KVM.
I end up running the Elgato Wave Link software on both my Windows desktop and Mac Mini (each running the relevant VST’s), but I’m not allowed to install it on my work Macbook (and don’t really want to use this mic on there anyways).
Ideally I would have just the headset microphone go to the work Macbook , but the more we talk about it the more I’m convinced that I just need to build some more powerful system to handle this audio routing. Unfortunately the Elgato software works best on Windows, and using that has it’s own headaches even with debloating and such
.