So, is it possible that I use the RME with say 4 Behringer ADA8000? Simple as that? Of course involving any setup needed in RME software.
But in Cubase all I will see is the inputs/outputs from the RME?
It is as simple as that.
You will have to use clock sync to have them all synchronised but that’s simply setting each to clock off the ADAT input to each from the Digiface (it has no BNC clock port) and then either slaving the Digiface off Cubase (as clock master) and then slaving the four ADAT boxes off the Digiface - or using the Digiface as clock master and slaving Cubase and everything else off that. Even if you don’t use the outputs from each ADA8200 you will need to install an ADAT cable for output to each unit just for sync. If you just plug them all in without the ADAT out sync from the Digiface you will get complete digital breakdown and distorted audio!
The ADA8000 it should be noted is superseded by the ADA8200 and I believe with big improvements: better A/D & D/A Cirrus converters, lower latency (0.55msec now vs 1.43msec), lower distortion and the addition of Midas mic pre’s. Downside there’s no direct link between inputs and outputs and it also has global Phantom across all inputs - they’re all on or all off. The ADA8200 doesn’t have the heat failure problems the ADA8000 had - consumption has been reduced to 15W from 25W and also they changed from a linear power supply with a transformer, to a switched mode, overall the unit running cooler and quieter. The restriction to 44.1 or 48KHz is the same for both models.
Perhaps if the budget allowed a step up to say the Focusrite Scarlett OctoPre - up to 192KHz, Phantom assignable to Ch’s 1-4 and 5-8 as groups, direct throughput of inputs to outputs from mic pre’s but no D/A - this is an input only device. You can use one as ADAT master clock source, setting the Digiface to sync off ADAT, and connect via BNC daisy chain to the other units for clock sync.
Edit: looking again at your post you would want the D/A so the above Scarlett would not work for you.
If you do need the outputs as well the Clarett+ OctoPre has A/D and D/A but it’s more expensive again.
That being said there’s nothing wrong with the ADA8200, a good sounding unit (I’ve installed a couple of systems with these as a part) and if you don’t need the higher sample rates offered by other more expensive brands then go for it. I’d be a bit wary of the earlier model 8000 as if it doesn’t have issues it soon will especially with four of them racked together.
Cubase will handle up to 256 physical inputs. More important is how your computer will handle a high simultaneously recorded track count. What you see in Cubase is all 32 channels (from the Digiface) as ADAT port channels which you can then label as you wish - the ADA8200’s are not seen as themselves as they are not interface devices but simply converters tacked onto the Digiface.
32 channels of I/O is nothing for current and earlier USB3 - FireWire would struggle, FireWire 800 not so much but would handle it and it’s way slower than USB of today.
USB3 depending on version 3.1, 3.2 gen 1x1 thru 2x2 has throughput from 5, 10 and 20Gbit/sec so using USB for audio is definitely not a drawback. The actual speed downside for large track count recording is drive/storage access speeds - SSD’s top out at 500MB/sec (4Gbit/sec) but if you have a newer computer with NVMe and a drive to match you can have read speeds of around 3.2GBytes/sec (25.6Gbit/sec)
So for most computers the USB throughput is faster than drive read speeds and bucketloads faster than the write speeds of the contained drives.
You need the combination of drive read and write simultaneously for audio production - these speeds, along with interface speed, dictate system latency. With direct monitoring this is not such a problem however, but a high track count with lots of plugin inserts does require decent speeds all through or you will crash. Recording a large number of tracks with plugins in place requires a fast system as you can’t hear the track inserts via direct monitoring and you need to keep the latency down for the performance monitoring or the performer physically is out of sync with what they hear through the cans. For recording where reverb is required I just do a send from the track(s) to an FX track with a reverb inserted and that comes through the cans as well as the direct signal coming via direct monitoring - the RME TotalMix has all you need for that - but it has no inbuilt FX though so you would need to do as suggested above if you want reverb for vocalists etc.
Hope this helps.