Some questions about Overhub

[Noisebuddy]

I’m intending to get the Overhub for some reasons. First, i want to sync my Elektrons sample-accurate to my DAW. But i also want to trigger my Blofeld and my Radias using the Hub. It’s basically a transmission from sync-signals, note-on/off and CC-messages. So here is the question, since the Overhub has seven usb-to-hardware-ports and one host-port, i have to configure so called MIDI-ports in my DAW (one port has 16 midi-channels). It will look like this:

port#1 > just syncing the AR (and from there the A4 and the OT via MIDI-trio)
port#2 > Blofeld (can be driven via USB)
port#3 > Radias (can be driven via USB)

my fear is, that the concept of midi-ports over a single hub isn’t well understood by the machines as long as i don’t use a real midi-interface with more than one connection like some of the iConnectivity MIDI Interfaces.

am i right?

[SoundRider]

First I would say … consider MIDI and USB as more or less independent technologies or interfaces. USB is only a carrier of MIDI information.

I have several units with MIDI over USB. If I connect them to my computer, I find inside my DAW a TR-8, System-1, Elekrtones via OB, you name it, as dedicated “MIDI-Ports”. This does not depend on how this gear is connected via USB. I can use different USB-Ports of the computer as well as one Overhub. The Overhub is particularly designed to manage a broader bandwith for data transmission. But this processing power is only required, if more then one unit is sending/receiving audio-data.

If your instruments conncect well via USB … and maybe in combination with a dedicadet driver … they should be listed in the MIDI section of your DAW and you are ready to go.

But you mentioned syncing as well. Now there is a problem, which is caused by the OS of a computer and the low timing priority for MIDI, particularly MIDI-clock. Since neither Mac nor PC have a “real-time” OS, there may be always timing issues caused by the computer, if it shall be used as the master-clock . Sometimes only an external clock-device can help to overcome this. I would try to use the OT for this task prior to buy an expensive outboard master-clock.

[Noisebuddy]

But you mentioned syncing as well. Now there is a problem, which is caused by the OS of a computer and the low timing priority for MIDI, particularly MIDI-clock. Since neither Mac nor PC have a “real-time” OS, there may be always timing issues caused by the computer, if it shall be used as the master-clock . Sometimes only an external clock-device can help to overcome this. I would try to use the OT for this task prior to buy an expensive outboard master-clock.

thank you for the informative answer

Sync-accuracy for the Blofeld and the Radias is not needed because they are only used for long evolving or complex pads and atmospheres, something which is to complicated to achive within the OT-sequencer. So this two synths are not primary used for in-sync baselines.

This brings me to the next question which brings me to the Overbridge-plugin. It is supposed, amongst other functionalities i don’t need, to sync the the AR/A4 tight to the DAW-clock, is that right? And if this is the case, is there any known interference with other signals like note-on/off and CC-messages for other synths, when running simultaneously thru Overhub. My notebook has USB3 - this shouldn’t be an issue!

[SoundRider]

This brings me to the next question which brings me to the Overbridge-plugin. It is supposed, amongst other functionalities i don’t need, to sync the the AR/A4 tight to the DAW-clock, is that right? And if this is the case, is there any known interference with other signals like note-on/off and CC-messages for other synths, when running simultaneously thru Overhub. My notebook has USB3 - this shouldn’t be an issue!

The “tightness” of MIDI is not an issue of OB, it’s an issue of the OS. Since OB is software, which is executed by a “non-real-time” OS, it suffers, or may suffer, because the computer MIDI-clock is “not-tight” by itself. This inaccuracy will then be passed to the gear too. I would say that your Elektrons are absolutely “tight” to OB, and OB “tight” to the computer timing … BUT … the computer timing is not “tight”.

I don’t get precisly what you mean with “interference”, but I would say, if you send MIDI on a track out to gear-A and MIDI on another track to gear-B via USB and Overhub, it should not get mixed up. I have never experienced this problem, and I think, this is because of the MIDI implementation of the typical DAW, where gear-A and gear-B are independent receiver using independent “virtual” MIDI-Ports.

If you have a clean system environment running, without too much of background tasks, stealing away CPU time, you should also experience a good enough synchronization. Just plug your gear in and do some testing. I expect, it will work just fine.

If you run into synchronization issues, you can always swap in a tight external clock, make it the midi-clock-master for everybody else and synch your DAW from this source.

[invisible_acropolis]

The “tightness” of MIDI is not an issue of OB, it’s an issue of the OS. Since OB is software, which is executed by a “non-real-time” OS, it suffers, or may suffer, because the computer MIDI-clock is “not-tight” by itself. This inaccuracy will then be passed to the gear too. I would say that your Elektrons are absolutely “tight” to OB, and OB “tight” to the computer timing … BUT … the computer timing is not “tight”.

yep… I dont think people really appreciate just how old and subsequently arcane the MIDI protocol truly is

Given that MIDI messages are at least ten bits long including framing bits, and that their transmit rate via 5-pin DIN is limited to 31.25 kbits per second

thats already a problem… back when MIDI was invented, the most RAM a computer had was 1MB… thats it! and CPUs ran at 5-10 Mhz… not 2 or 3 Ghz (2000 or 3000 Mhz)… and the codebase for an OS (for example MS-DOS) was itself less than 1MB. The entire OS that the computer ran from was less than 1MB!!! why does this matter? Because MIDI was designed to run perfectly on machines like this. Its expecting a very regular interval of bits turning over and so forth, mathematically designed for precision on computers of that era, where the number of data messages being parsed is a few thousand per second - compared to today where the messages being parsed are in the millions per second. Computers today are so vastly different than computers in the early 80s - it is like comparing a toaster to a full service 4 star kitchen with a head chef and prep cooks.

Thats why you may hear people talking about still using their Atari ST or Amiga computers as their sequencer. If you use a computer sequencer to run your MIDI rig, its pretty much just as tight as using hardware these days - because that is how it was designed to work. It was not designed to fly through all kinds of busses and kernel functions, etc. etc. Yes, MIDI still technically “works” on modern machines - but not as intended.

Plenty of people will talk about how its “good enough” - and hey if it works for you, then go for it, live it up and have a cocktail to celebrate. But dont be surprised when you have problems. You should expect problems. Constantly. As a matter of simple fact.

[Noisebuddy]

The “tightness” of MIDI is not an issue of OB, it’s an issue of the OS. Since OB is software, which is executed by a “non-real-time” OS, it suffers, or may suffer, because the computer MIDI-clock is “not-tight” by itself. This inaccuracy will then be passed to the gear too. I would say that your Elektrons are absolutely “tight” to OB, and OB “tight” to the computer timing … BUT … the computer timing is not “tight”.

ok, this time i got it

the last time i used OB with FL syncing the AR was a complete mess

but the other way round … syncing FL to for example the AR, could work