Casual? I don’t know how casual one can be about synchronization.  At least, not when your sync has gone south.

The sync issues involved with digital audio wordclock and SMPTE or MTC timecode are confusing to some people.  They are two very separate sync issues, and can be the source of nightmares when these very separate issues have to occur simultaneously.  Two separate yet simultaneous syncs. Get ready.

I was working in a studio with three O2Rs, three ADATs, and an analog 2” Otari deck.  When I first entered the scene, the 2” 24 track analog machine was the master deck, the ADATs chasing the timecode on the 2”.  Most of the recording was being done on the 2” and transferred to ADAT, but the vocals always were done on the 2” deck.  Ultimately, each project had tracks on both ADAT and analog.  And, every project was in sync hell.

Some mistakes had previously been made in locking up the projects, and I don’t know WHAT in the world was going on there.  But the ADATs were getting their sync reference from the SMPTE timecode on the 2”, while they were also acting as the wordclock master to the O2Rs.  I’ll give you a minute to process that one.

The main issue here is not who was the wordclock master, and not just the fact that a digital device (ADAT) had to lock to timecode that had some inherent variance (the analog deck), but that the ADAT and the O2R had to also have their wordclocks synchronized.  Let’s break this down.

Two wordclock square-wave signals, perfectly in sync

When transferring digital audio, the sampling clocks of both devices must be in sync.  If both devices are running on their own internal clock, they’ll pass audio just fine, but the final result will often be harsh and grainy.  If one is set up to sync to the other, and this sync becomes compromised in some way, other audio artifacts will begin to arise.

When we talk about a digital recording device, we’re talking about sampling.  Sampling is going to happen at a certain determined rate, but will also begin at a particular time.  This time is either generated by an internal clock within the digital device, or is determined by a clock signal that it receives from an external device. 

This clocking information is contained in the ADAT Optical signal, as it is in a S/PDIF signal or an AES/EBU digital audio signal.  On some devices, clocking info can be handled by a signal stream that is independent of the audio information, and is generally considered the most stable vehicle for this information.  And wordclock is its name-o.

Again, the clocking info is contained in the digital audio stream, and not all devices have a wordclock option.  Some folks use a BRC to control the ADATS, which does have wordclock I/O.  The O2Rs, bless their little CPUs, are versatile enough when it comes to their clock options.  You can have them run off their internal clock, or slave to either S/PDIF in, AES/EBU in, ADAT Optical in, or wordclock in.  Generally, a device will have a setup page, a sync page, or a master clock settings page in which this selection would be made.

When you have chosen an external device as the master clock, the O2R will run at the sample rate and according to the sampling clock that it sees at the selected input.  If you select “wordclock” as your timing master in the O2R, as in this example, and the signal at the input breaks down and becomes “invalid,” the O2R (or any digital audio device) will start popping and clicking and may stop passing audio altogether.  How’s that for a segue back to a story?

Two wordclock square-waves, tragically out of in sync

So there I was, the O2Rs a-poppin’ and a-clickin’ and then a-not passing audio.  Sometimes, when the 2” was coming up to speed, the O2Rs would pass the audio signal, then all of a sudden there was no meter activity and no sound.  Other times when the 2” speed would vary, such as a point where the tape was stretched slightly, we’d ‘just’ hear that popping or clicking sound that has become the trademark of clocking problems.  Every once in a while, meter activity would disappear, then a moment later audio would cease.  I hope I never witness this again.

We racked our brains, called every tech support number we had including my Cuisinart (they were very helpful, really), switched wordclock master to the O2Rs, got different cables, tried a wordclock distribution amp.  The theory was, the O2Rs are cascaded, so wordclock goes from one to the other.  The distribution amp made wordclock hit them at the same time.  Still, no joy.

Out of all of this, some words out of Alesis tech support kept ringing through my head.  No one from any company had an answer, which was confusing enough.  It was their technology, after all.  In the midst of my desperation, the Alesis guy said sheepishly, “The ADATs like to be the master.”  Mainly, it was the situation that they were most used to and able to support, but those words haunted me, until one day a light bulb went off in my head and I exclaimed, “The ADATs like to be the master!”

To Be Continued!! Tune in Next Week for Part Two!!

JD Mars is the producer of Digital Pro Sound. JD is a recording engineer with major label credits, with a background in technical writing from his sojourn as tech support manager at Midiman/M Audio.