I’m confused about the Midihub’s USB MIDI functionality.
I currently have 6 USB devices connected to an Elektron Overhub powered USB hub. This connects directly to my Mac. I use Bome MIDI Translator to route USB MIDI as required. This is just the app running on my Mac. I don’t own a Bome Box.
That all works fine.
I’m hoping to go standalone with this configuration. Will the Midihub allow me to do so? So can it act as a standalone hub for routing MIDI over USB?
doesn’t this need a computer?
Connect up to 7 units to a single host computer
says their website. This also suggests that’s true.
Bear in mind that when Midihub is working standalone, it needs to get its power via the USB cable. So to use Midihub with your USB devices you need some box will send power* and take your 6 USB inputs and allow you to configure (either in itself or via computer just for set-up) which of Midihub’s 4 USB inputs (USB A-D) their messages will be routed to.
Once there, the world’s your oyster…
(*) actually Midihub uses so little power that I found that my charge-pack would switch itself off when powering it alone. I now use a cheap usb splitter and power both a keystep37 & Midihub from the pack:
this gives another option of powering both the USB box & Midihub this way and running a MIDI DIN cable into Midihub from whatever box to carry the messages from the 6 devices.
welcome to the forum; I hope some of that helps!
Midihub does not host USB devices. But for a portable setup, you could set up a Raspberry Pi (which has 4 USB ports, but you may add a USB hub) to automatically forward the USB MIDI data to where it needs to go, including through Midihub’s DIN-5 ports.
I did get a simple router working on a Pi with some Python library but it wouldn’t talk to my Waldorf Quantum. Every other bit of gear in the studio was happy. It would find and open the Quantum’s ports, but then they’d ignore each other. I didn’t know enough about the library or Python under the hood to investigate.
You should just use
aconnect for forwarding MIDI data between devices. There’s
aconnectgui to make it easier to use. Also check out
amidiminder here in the forums for a daemon service to automatically set up the connections.
Thanks, I’ll look into those
The Python library was MIDO. The Quantum has two separate layers with independent MIDI channels. I assumed there was something about that implementation that MIDO wasn’t seeing.
aconnect would forward the entire 16 channel MIDI stream, so it’s up to you to set up the hardware to transmit on the necessary channels. If you need channels to be changed, then indeed other software based solution (or a good Python script) would be necessary
Hey there, @dcjams !
I was intrigued by this cos I’ve been playing around a little with
mido (mostly via
isobar). When I read that your Quantum wasn’t responding I thought I’d take a look at Iridium (as the Os is much the same) with mido via usb (being a bit old-school I’d only used DIN cables from Midihub previously…)
Here’s what I found:
- both the above give ‘Iridium’ occurring twice in their lists, which is odd & I thought might cause a problem, BUT
- with a Mac* it does cooperate when the only patching with the Iridium is via mido
- this is also the case when an Iridium pad is hit and the script sends back a transformed note after short delay,
BUT I had to go to Global>MIDI>Outputs and turn on ‘USB Computer Out’
*I’m not near my rPi for a month or so , but I’ll give it a go when I am reunited!
In the meantime, I’d be most grateful if you’d give a quick update on your progress with mido/ aconnect as I’d like to explore the rPi/mido/Iridium combo myself.
Thanks for your reply! The Iridium’s a great synth. It wasn’t out when I got my Quantum. The Quantum’s huge but it is very beautiful looking
I was planning to have a tinker again at the weekend. I’ll let you know what I discover.
It is a gorgeous beast. Yet to see one in the flesh.
Look forward to hearing how you you get on, Jim.
Cheers & fingers crossed,
I just ran the exact same code from my Pi on my Mac and it found two port references to the Quantum. The code doesn’t actually distinguish between them as the Pi only found one port so I was never looking for two. I think the code will assign the second port as the active Quantum port.
Regardless, the exact same code that failed to communicate with the Quantum on the Pi, works absolutely fine on my Mac.
Maybe I’ll just get an old Mac Mini and run it headless! You can get the earliest Intel ones (Core Duo etc) for not much more than a Pi on eBay!
That’s interesting, what output do you get on your Raspberry Pi when the Quantum is connected to it from the following commands?
On the rPi, did you try
amidiminder mentioned by Giedrius, as a matter of interest?
amidiminder is not found
amidi - l reports only a single reference to the Quantum:
IO hw:7,0,0 Quantum MIDI 1
aconnect -lio same:
client 44: 'Quantum' [type=kernel,card=7]
Linux raspberrypi 5.10.103-v7l+ #1529
It’s connected via a hub with lots of other devices. I’ve tried connected directly to the Pi, all four ports. It’s the same.
My Mac doesn’t know what amidi or aconnect are, so can’t compare
So Quantum is supposed to appear as 2 USB MIDI ports?
If you read up the thread, a simple MIDI router in Python using MIDO found a single port for the Quantum and wouldn’t communicate with it, but would with every other USB MIDI device. The exact same code running on MacOS did work. The only difference being MIDO found two unique ports for the Quantum.
… and, I don’t know if it’s supposed to appear as two ports or not, but it does under MacOS and doesn’t on my Pi
My guess, Jim, is its a Waldorf glitch, cos I’ve not noticed this (duplication name in mido) with any other attached USB device and, unless there’s some other handle like a USB_MIDI_device_unique_id, nothing can attach to the ‘other Quantum’ anyhow!
I notice, btw, on the Mac that software such as MIDI Patchbay/SysEx Librarian/Ardour all mention just one instance of ‘Iridium’. So I’d hypothesise that the different OSs are handling the Waldorf glitch in different ways which mido then gets confused by on rPi.
Wonder if Waldorf are aware of this?
This does hint at OS level issue, Quantum USB MIDI peculiarity, or different MIDO library (and possible others) versions. It would be interesting to see the output of this command, executed soon after plugging in the Quantum synth to the Raspberry Pi:
dmesg | tail -n 20