Clocks & ADATs ... a question (not a problem for once!)

[BigKahuna]

I got my ADATs back from service and they work great now. One thing I noticed is that they defaulted back to a clock of 48k INT.
The result was a lot of clicks and pops every so often when trying to record (from a DA7 connected w lightpipe). I nearly blew a gasket! However, it was easily resolved by switching the clock source back to 48k DIGital.
My question is ... what is the main difference between the two?
By choosing DIG, my guess is that it's reading the clock from the lightpipe input (read DA7) instead of within the master LX20 ... but why wouldn't they just be the same considering both are set at 20bit and 48kHz?
Also, is this "digital" clock different than "word" clock? if so, how? what's the difference?
Those are probably dumb questions ... but what can I say?
This feels like a fundamental digital recording concept that I'm still lacking understanding of. I really appreciate anyone shedding some light on it for me.
Thanks!

[Track Rat]

What ever you're sending digital audio to has to be slaved to the source which is the clock master. For instance, when I record to my ADATs using their analog inputs they use their internal clock as a timing master reference. If I then want to send that audio from the ADATs into the computer for editing, the ADATs are still using internal clock and the computer is set to ADAT sync. Once that audio is in there the I switch the computer to internal clock. To send that audio back to the ADATs they have to be set to Dig so they will slave to the computer. As you know the first indication that there's a clocking issue is the clicks and pops. All digital signals have clock info imbedded in them. Things like minidisc and DAT recorders will generaly automaticly lock up to this incoming signal without being told to. On the other hand ADATs and the like have to be told who's the master and who's the slave. Word clock is a very stable distributed clock for syncing up digital devices to each other. I know this is over simplified, but I think that's the jist of it.

[BigKahuna]

So ... if I read that right ... it's more of a syncing thing than actually setting a resolution. Not a "transport sync" but just a "sample sync". Please forgive me if that doesn't make sense.
Even though both the board and the ADAT are set to 48kHz ... I have to tell the ADAT when the sampling "starts" so it can match up to where the board "starts". Sorry if that seems too simplistic ... but sometimes I need to be told like I'm a 3 year old.
Thanks for your response, Track Rat.

[Track Rat]

Digital devices use crystals in their clocks for sample rate timing. In the real world no two crystals are the same so for any two devices to talk one has to be master (sending sync information) and one the slave (recieving and locking to sync information). If two digital devices are in a system,one transmitting (or playing back) and one is recieving (or recording) and both are using there own internal clocks those differences manifest themslves as popping and clicking.

[mnicoletti]

Hi, Big Kahuna,

Do not apologize for your questions...they are perfectly valid. Roger Nichols once said, "Never turn your back on digital".

The analogy I use for customers is that a digital setup is like the body: it uses ONE HEART for its timing. You can imagine how things would be if you had more than one heart trying to impose their rates. Even if they were both at 72 BPM, if they are not perfectly in sync, you would feel very stange to say the least.

As Track Rat aptly said, digital systems need one clock Master (heart) to run the system. You had two Masters running at 48K. Since their clocks were close, you heard ticks. If the clocks' timing was farther apart you would likely get bursts of white noise. Since the dig console is Master, and does not have ADAT 9 pin, you need to put the ADAT in DIG CLOCK so it clocks from the Optical, as you discovered.

If your system grows appreciably, you may want to look into "house sync": one clock source/distribution box so that all clock comes from one "point" source. This is because if you daisy chain enough units together, you will get some noticeable delay (in samples) by the time you get to the last unit.

Some things, like physics, you just don't change with technology.

Mike