Using multiple interfaces in ProTools - even identical interfaces possible?

[And#PT]

Hello, in the Mac's audio MIDI setup I can combine several audio interfaces into a new main device and then use this as a playback engine in ProTools. The inputs and outputs of these different interfaces are then available there. So I can connect a Steinberg and a Behringer interface together, etc. But what happens if I want to bundle two identical interfaces? I tried this with 2 Behringer UMC1820 that are currently being manufactured. This means you would have at least 36 inputs and 40 outputs available for just EUR 500. However, I couldn't get it to actually work. I can combine both UMC1820s into a new main device in the audio MIDI setup. If I then select this new main device in ProTools, all inputs and outputs are also correctly listed in the I/O setup in ProTools. I can also route the signals accordingly. But unfortunately there is noticeable noise on some channels, and the signal comes out much quieter on the second interface. All settings etc. are correct, I rule out an error on this page.

Is it possible that ProTools (or the Mac already in its audio MIDI setup) can't handle it when two identical, identical interfaces are addressed? The same drivers are used. Does this bundling or combining of several interfaces into a new main device actually only work if you use different interfaces?

(MacOS 10.15, PT 22.12)

Andi

[Farhoof]

Is the volume set the same on the devices individual pages in the audio midi setup?

I don't use aggregate devices, but I assume the individual setup affects the aggregate device as well.

[And#PT]

Yes, all settings in the audio MIDI setup are identical.

Alternatively, I selected one of the two connected UMC1820 directly/separately in the ProTools peripherals (then only 18/20 I/O). The following happens there: only the first interface (top in the peripherals) always works, even if I select the second (bottom) one there. This means that a signal that is routed to Out1(1) of the second interface still comes out to Out1 of the first interface. I suspect that ProTools (or MacOS?) cannot distinguish between the two interfaces due to the identical name (and identical drivers). Is this assumption correct?

It is not possible (for me) to give the identical interfaces a different name, which might help.

Andi

[devinemke]

Configure the 2nd device via the Pro Tools Aux I/O system.
You will need macOS >= Big Sur + Pro Tools >= 2022.9.

[Darryl Ramm]

Originally Posted by And#PT

Yes, all settings in the audio MIDI setup are identical.

Alternatively, I selected one of the two connected UMC1820 directly/separately in the ProTools peripherals (then only 18/20 I/O). The following happens there: only the first interface (top in the peripherals) always works, even if I select the second (bottom) one there. This means that a signal that is routed to Out1(1) of the second interface still comes out to Out1 of the first interface. I suspect that ProTools (or MacOS?) cannot distinguish between the two interfaces due to the identical name (and identical drivers). Is this assumption correct?

It is not possible (for me) to give the identical interfaces a different name, which might help.

Andi

macOS aggregate I/O is built into a low-level in CoreAudio, it will not have routing problems, and neither is Pro Tools likely to. CoreAudio and Pro Tools do not rely on the names you see to identify devices (which all internally have unique CoreAudio audio device object identifiers), these systems were not designed by total idiots, and aggregate I/O and Pro Tools will work with aggregates of identical devices (this is not unusual at all for some folks to do), ... within the standard caveats of using aggregation with Pro Tools.

And those standard caveats are... you should avoid aggregate I/O where possible, it can introduce timing and performance issues, aggregate I/O in Pro Tools was introduced to allow aggregating Mac built-in speaker/output to other interfaces, and Avid used to document that was the only supported configs, no idea what they say now. The quality/robustness of the aggregate device is largely dependent on the quality of the device driver. Vendors like Motu deliberately put effort into making sure their drivers work well in an aggregate device. RME drivers work well in aggregate because well RME driver quality.

You are trying to aggregate cheap Beringer interfaces that rely in low-performance USB class compliant drivers??... I would have low expectations that will work reliably/well (eg at low HW buffer sizes).

When aggregating interfaces like this you should have one box driving the clock into the other. Did you do that?

If you are having signal routing problems it is most likely something you are doing wrong, stop assuming the software is broken and focus on what you are doing, possibly how you are using the UMC mix control hardware mixing in each of these boxes... that knows nothing about the aggregate I/O and works as if the boxes are entirely separate and can easily lead to confusion with aggregate setups

And hopefully you are not trying to do this with Pro Tools Intro or Artist as you will just run into their IO limits.

[Darryl Ramm]

Originally Posted by devinemke

Configure the 2nd device via the Pro Tools Aux I/O system.
You will need macOS >= Big Sur + Pro Tools >= 2022.9.

I would hope there is no fundamental reason Aux IO works any better here than aggregate I/O and it may introduce latency/ADC issues. But it is a good idea and worth trying maybe just because it gives a different view of I/O and might help with any routing confusion.

The hardware mixer/mix control setups in both boxes should be excluded first as a possible factor.

[EGS]

Originally Posted by And#PT

... But what happens if I want to bundle two identical interfaces? ...

Dunno about two 1820's, but you can daisy-chain two identical Presonus 2626 interfaces on Mac or Win:
https://support.presonus.com/hc/en-u...c4Ny42MC4wLjA.

[Darryl Ramm]

Originally Posted by EGS

Dunno about two 1820's, but you can daisy-chain two identical Presonus 2626 interfaces on Mac or Win:

Nope. These are low-cost class compliant USB interfaces, heck they are USB 2.0 class compliant devices which pretty much guarantees low-performance. So not going to be great interfaces at low H/W buffer sizes, and really likely to be not great when aggregated. And being class-compliant USB they are not going to do anything fancy like chaining as supported by many *Thunderbolt* interfaces from UAD, Apogee, and Presonus.

What I forgot to mention before is the "proper" way of expanding these interfaces is to use ADAT. It's great that they have ADAT and great that they have a stand-alone mode to enable chaining an 1820 off ADAT on another 1820. (edit: correction--the 1820 does not seem to have a stand alone mode).

[EGS]

Originally Posted by Darryl Ramm

... It's great that they have ADAT and great that they have a stand-alone mode to enable chaining an 1820 off ADAT on another 1820.

As far as I can tell (I hope I'm wrong), the 1820 does NOT offer a stand-alone mode, i.e. it canNOT be used as an ADAT slave expander to another 1820.

[Darryl Ramm]

I think you are right, I was going by some comments posted online by Berringer, cough, on Twitter/X. When you look at their (horrible brief) doc and other comments its clear they only talk about using their dedicated ADAT preamp product with this interface.