The issue I had with the periodic dropouts was....... Ozone was not authorised and it was doing its demo mode 'prevent you from really using me' thing.

Why I think I missed it was:
a) No dialog box requesting authorisation was displayed when I opened the project.
b) The dropouts were occurring even when the plugin was bypassed.
c) I thought the dropouts were meant to be every 10 seconds if not authorised.
d) Mercury retrograde, but now near its end.

I must have accidently accepted the demo mode after I installed it with the new computer build a couple of weeks ago.

The good news
So I get to continue to use TWO Firefaces at 192k. REALITY bites!!!

Sorry, but there was a particular person on my Cubase thread for this issue that kept trying to tell me that my problems were because I was using:
a) Two audio interfaces (he didn't know that the RME driver could do that, but still discounted that later on), and
b) 192k, and put unnecessary pressure on the system.

The reality that it all worked previously did not seem to faze his need to keep pushing his theories.
My motto is that 'REALITY trumps theory EVERY time!'.

Some people just can't get their head out of their own a*** when dealing with things outside their reality
The strange thing was that the supposedly 'helping me' contributor kept trying to push that it was stupid for anyone to use multiple audio interfaces. Well many RME users have been doing just that for years. Mine was to have I/O in two different locations (mix area and the studio).

So you are all obviously fools if you dare to do that. You need to go get yourselves certified and locked up IMMEDIATELY!!

Is that true?

If so, with what TI chipset(s)?

Reaper test setup:
a) i7-4930X 3.4GHz, Asus P9X79-E WS, 32GB
b) Syba SD-PEX30009 PCie 2x1394b+1394a, TI XIO2213A
c) SIIG NN-FW0012-S1 PCie 3x1394b, TI XIO2213A
d) Input into 800 on ports 9 and 10 (mics).
e) Inputs into 400 on 3 and 4 (line).
f) Both channels set to out on the 400, ports 1 and 2.

Basic testing with Reaper shows:

a) Selecting any port on the 400 does not throw the syncing out at all (unlike with Cubase).

b) If both Firefaces are on the same card (either Syba or SIIG), then the recorded sounds are very distorted.

c) If the Firefaces are on separate cards, 800 and 400 either way round, the sounds are all perfect, with NO periodic short dropouts (like there was with Cubase).

OK, me thinks there might be some funnies going on with the RME driver and the XIO2213A, as they are the common components, though the disparate behaviour with Cubase, Reaper and Amplitube is cause for concern.

I have reported these results on the Cubase Issue thread as well.

a) No Host flashing

b) No repeated dialogs

c) Very short audio dropout every three seconds, with lots of crackles (but no ASIO CPU spikes).

This is basically what I got when I used the SIIG card driving the FF800, and the Asus motherboards VIA 1394a chip driving the FF400, though without the crackles.

Note that Amplitube has no such audio issues in this scenario either.

To me it is looking like Cubase does not like something about the dual 1394 audio interface setup, but particularly goes troppo if they are both on the same card.

At any rate, all the scenarios involving the new TI chip do not work when there are two audio interfaces. Again, none of these issues occurred with the old TI chip on the Gigabyte motherboard.

With Amplitube, selecting ANY connection from either Fireface gives no Host flashing or dialog boxes.

With only one Fireface connected and in Cubase, selecting any inputs or outputs is Ok as well.

With both connected and in Cubase:

a) Selecting only FF800 inputs or outputs is OK.

b) Selecting ANY (even only) FF400 input or output results in Host flashing (on both Firefaces), and repeated mismatched sample rate dialog boxes.

None of this occurred with the older TI chip.

Also 'is the same' is not the point here. If you change the interface ports then the driver is reset and your settings might no longer be the same. That's why I asked whether master and slave are still configured accordingly.

When I connect my Fireface 800 AND 400, either individually to the card outputs or daisy-chained, they work with IK Mulitmedia's Amplitube 3 configured with IO from both cards.

However, when I open Cubase 7+, the Host LEDS on both interfaces flash for a couple of seconds, and when I open a project (just open, not play), I get:

a) constant flashing of the Host LEDs on the Firefaces, and

b) a never-ending stream of dialog boxes telling me that there were different sample rates and asking me to select how I wanted to handle it. Neither selection was palatable and neither solved the problem.

This occurred with either Firewire card with both:

a) Gigabyte GA-X58-UD4P, i7-920, 12GB, Win 8.x Pro, 3081 RME driver

b) Asus P9X79-E WS, i7-4930X, 32GB, Win 8.x Pro, 3081 and 3084 RME drivers.

When I used the GA-X58-UD4P's on-board Firewire from its TI TSB43AB23 chip, these problems didn't occur.

I have raised this issue in the Cubase forums as well.