Word Clock need on ADAT audio interface or not?

[Christianturton]

I really fancy getting a Mackie D8B and have seen one with Apogee card in it, I can an audio interface and the RME Digiface looks ideal accept it’s doesn’t have BNC wordclock - the D8B word clock apogee card only has BNC and the RME doesn’t- do you know if there is any other way of syncing or would I need another card? And can you recommend something with 24i/o on ADAT. TIA

[Christianturton]

Also someone has just told me word clock is not need when connecting via ADAT light pipe?

[Phil.c]

As far as I know, optical light pipe will only send audio and not clock signal.

[Y-my-R]

Basically, it's this:

All digital audio connections (e.g. ADAT optical, TDIF, AES/EBU, S/PDIF etc.), do inherently carry a clock signal, by the nature of how digital audio works. If, for example, sending a signal at 48000 Hz via a digital connection, there will be 48000 "snapshots" of your audio (aka "pulses") of a certain size transmitted over the digital cable every second. How many such digital "snapshots" are sent every second, is what your "sample rate" is. 44100 per second for the 44.1 kHz sample rate... 48000 per second, for the 48 kHz sample rate.

The "size" of each such "snapshot" or "pulse" is your bit-depth, btw.. The size, or much more the amount of information included in each snapshot/sample, dictates the available dynamic resolution btw. A snapshot/pulse in a size of 24-bit will be able to recreate quieter signals before they disappear into background noise. If the snapshots are in a size of 16-bits per second (or less), the audio signal will disappear into background noise even earlier - thus, less dynamic range is reproduced from the loudest reproducible signal (always at 0 dB digital... with artifacts happening as soon as 2 samples/pulses in a row hit the 0 "ceiling") and down to the quietest reproducible signal before it disappears into quantization/background noise.
(Maybe this helps to clear up the common misconception that the higher dynamic range of 24-bit would help make things "louder" or "punchier" - it doesn't and it just as "loud" as a signal at 16-bit etc. What's different is, that the signal can get quieter without disappearing into noise. So, with a higher bit-depth (aka snapshot size), you increase your dynamic range DOWNWARDS to be able to play quieter stuff before it becomes just noise).

Anyway... the "pulses/snapshots per second" are equivalent to your sample rate, and present on all digital audio connections.

While MOST digital audio devices and audio interfaces for the computer CAN use the clock signal that is inherently transmitted over digital audio connections (by it's "pulse rate") for synchronization purposes, the D8B CANNOT synch to such a signal - but it can (and always will) transmit it over such digital audio connections, and other devices can sync to it!

So, if you'd use the D8B as the clock master and the RME interface as slave - no problem. You set the D8B to internal sync, connect an ADAT (or TDIF or S/PDIF or AES/EBU) cable from the D8B to the RME audio interface, and configure the RME interface to sync as slave to the incoming clock from that audio connection.

The other way around is more difficult. If you want the D8B to run as word clock SLAVE to anything, then you need to have a word-clock cable connected to the BNC word clock input on an Apogee clock card installed into the D8B (...and then set the D8B to sync to the external word clock signal).
The D8B can NOT sync to the inherent clock that is transmitted along with other digital audio transmissions, such as over ADAT, etc.
I always thought that this was a stupid design choice. The clock rate is there, anyway (i.e. "pulse rate")... and if the clock runs pretty stable (e.g. the RME clocks are pretty decent), there should be no reason why the D8B can't expect this to be a stable enough clock to sync to it... but it doesn't, and that's by design. Sigh...

So, long story short:
If the D8B should run as Master, you can sync other devices included Audio Interfaces off of it easily, using any digital audio connection such as ADAT or S/PDIF.

If the D8B should be Slave, then you NEED to have separate word clock feeding into the D8B via the BNC word clock connector. No way around it.

One way to get around having the D8B act as clock master (...the clock on these Apogee cards isn't awesome from what I understand, and the "internal" clock card the D8B came with is worse AFAIK, not to say generally poor), is to get an external master clock that has at least one BNC word clock out to sync the D8B to. Then you could sync the D8B as slave to the master clock... and the audio interface as slave to the D8B (and thus indirectly from the master clock) from any other digital connection, such as by using the S/PDIF output on the D8B and connect it to the RME and have it sync to it.
(Alternatively, if you don' need to send audio over the S/PDIF cable, you could also have a S/PDIF output on the master clock - which many have, but they usually only carry clock and no audio - to the audio interface such as the RME. If both/all devices are synced off of the same master clock, they all run in sync so no audible artifacts/dropouts should occur).

Anyway... I feel I typed more than necessary and might have made it murkier than necessary by doing so, sorry. I still hope this generally clears it up a bit... but I can try to clarify if any of what I'm saying didn't make enough sense.

Best of luck!