Audio Interface Recommendations ~$500

Yes, that is helpful!

The game for me is not so much to figure out the Mackie Control protocol as it is to find MIDI sequences generated by that protocol that do something (hopefully useful) in RME’s TotalMix (aka “TM”). Toward that end, there are a number of folks who have uncovered useful MIDI sequences. Many of them are dependent on the state of the GUI (e.g. move the slider that is currently selected on the GUI). These sequences are problematic when producing a MIDI sequence that has no context for the state of the GUI.

This 2008 document by Jeremy Lee does point to an intriguing use of Pitch Bend that has not be explored by anyone else. I will give this a try and see what it does on TotalMix.

Thanks Torsten!

Hi @ClintGoss,

I guess that your main problem is that the selection commands are relative.

mesns: the same midi command can e.g. control the input level acting on output 4 or 7, depending on whether output 4 or 7 was selected before. This makes the midi command sequences in Cantabile difficult to handle.

I faced the same problem and first solved it with an Arduino. But in the long run this is impractical.

Since I had more ADAT I/O available than I needed with the Digiface USB and the Fireface UFXII, I solved the issue by connecting one of the unused ADAT ins to an unused ADAT out with an optical link cable.

This sounds absurd at first sight, but only until you realize that you actually have a whole matrix of faders in front of you. I’ll try to illustrate this with an example:

Suppose I want to control the volume of a synth at input 5 to output 7.

Then I would first have to select output 7 via midi command and then I could adjust the fader of input 5 via midi.

With the alternative method it goes like this:

  1. assuming ADAT I/Os 9-16 are bridged

  2. input 5 is mapped in TotalMix to ADAT Out 9

  3. the signal appears automatically at ADAT In 9 because of the bridge

  4. ADAT In 9 is mapped to output 7 in TotalMix

  5. now the control can be done with a single midi command, which firmly controls the fader of output ADAT 9.

Since no more channels have to be selected, in this example it works for 8 channels at the same time.

By the way, with this scenario you can also create stacked fader groups and e.g. build effect sends (pre- or postfader).

Supposedly there is also a possibility in TotalMix to create the said bridge via software. But I have not succeeded so far (but this is completely irrelevant for the principle).

Maybe now it becomes clear why I like the Digiface USB with its 4 ADAT strips.

Greetings, Volker


Well that is brilliant! Creative use of wiring … Hope when I’m a semi-senile WindSynth player I can remember how this all works …

One though however: Could the physical optical link cable be replaced by clicking the [Loopback] on that channel in TotalMix?

Regarding using the loopback channels in place of the optical cables: that was exactly what I suspected, but so far haven’t gotten to work.

What does work is using the loopback function to feed an output signal back into Cantabile, for example. I like to use this to feed a portion of an external audio stream (such as a hardware synth controlled in volume) back to a reverb or delay plugin within Cantabile.

Since these types of effects usually have a delay or predelay anyway, you can shorten that delay by the ASIO runtime, leaving the plugins virtually latency-free (which is why I don’t miss the reverbs and delays included in the more elaborate RME cards on the Digiface USB).

The routing from Out to In, however, I have unfortunately never managed (even if I find references to it from time to time). If someone knows a way: feel free to post it here.

Greetings, Volker

1 Like

Right. Forgot about the fact that [Loopback] does not actually loop back to the Inputs of TotalMix.

I’ve learned to imagine that TotalMix has a 4th bank of ports: In addition to Input ports, Software-Playback ports, and Output ports, there is a hidden bank of Software-Input ports (the opposite of Software-Playback). Hardware signals coming in to the UCXII go to both the Input and Software-Input ports, but [Loopback] outputs get routed only to the corresponding Software-Input ports.

Some questions:

  • So I have a UCXII - can I just add a wire from ADAT-Out to ADAT-In??

  • Could I do the same with the SPDIF ports? (wire from out to in)

  • Since these are digital ports, is there any latency being introduced by this scheme? (i.e. should not even need to go through the ADC and DAC)?

Sorry for hijacking this thread, but this is a rather critical topic to me and I don’t know exactly how to extract an ongoing conversation into a new topic …

O.K., I’ll try to answer that.

  1. from my point of view the same applies to the ADAT connections of the UCX as for the Digiface or the UFX. However, there is only 1 pair of IOs available here. So you would have to decide if you want to allocate the ports for the purpose. Of course, an ADA8200 will not work then. Generally you have to make sure that there is no feedback loop programmed in TotalMix - then it whistles.

  2. S/PDIF is also possible. One chinch cable is sufficient. But then there are only 2 channels instead of 8. In principle even analog IOs could be used in the same way. In contrast to ADAT or S/PDIF there will be additional noise and converter latencies (even if I consider them rather academic).

  3. as far as the latencies of the process are concerned, I see it similar to you: the signal probably does not leave the digital domain here. So there should be no converter latencies to speak of. Probably there won’t even be latencies for any PLLs - but I can only guess. For me it would be academic, but since I think I know that latencies are extremely important to you, my recommendation would be: measure!

Greetings, Volker


RME are cool; their manuals include info on converter / totalmix sample latency.
Whilst the AD/DA latency has improved significantly between the original BF and the latest BF Pro FS the digital latency has remained the same:

The term describes the digital path of the audio data from the input of the interface to its output. The digital receiver of the Babyface can’t operate un-buffered, and together with TotalMix and the output via the transmitter, it causes a typical delay of 3 samples. At 44.1 kHz this equals about 68 µs (0.000068 s), at 192 kHz only 15 µs. The delay is valid for ADAT and SPDIF in the same way

BF Pro FS:

Digital receiver buffers plus TotalMix and output transmitter shoiw a typical delay of 3 samples. At 44.1 kHz this equals about 68 µs (0.000068 s), at 192 kHz only 15 µs. The delay is valid for ADAT and SPDIF in the same way.


I just did the ADAT loopback test, and the results are as reported by RME documentation. The cost of a physical ADAT cable routed through TotalMix should be 6 samples (tested at 48kHz).

Here’s the diagram which gives the specifics … hope I’ve got all the specifics straight: