Ignorant Word Clock Question

[Gilliland]

The only point on which I might comment would be the one about the exception being a flawed/defective clock. Maybe this is dancing with semantics a bit, but the mobo clock in your average PC-based DAW platform is not going to be what I would consider flawed or defective; at least not for it's designed purpose of clocking general purpose computer curcuitry. But it certainly isn't something I would consider to be operationally stable enough to call a quality house synch signal.

This has been an excellent conversation, I agree. And you're certainly right that the circuitry used in most PCs to govern the CPU and support chips would not be designed to the same purpose as a clock that is intended for audio. While they are similar, they have somewhat different goals. But it's also worthwhile to note that it is not particularly difficult or expensive with current technology to design a clock that is both accurate and precise in audio terms.

What's difficult and expensive is propagating that clock over a distance to multiple targets. That's where a Big Ben or other quality clocking device comes in handy. But even then, careful attention must be paid to the clock signal path - even more so than the audio signal path.

This is, after all, HOME recording dot com. Most of us are not at the point where we need to worry about a clock bus for our home studios. There are other ways to spend our money that will have a much bigger impact on our work.

 

[SouthSIDE Glen]

MD & Gilliland,

Yep, I agree with everything that you guys have said.

Just one more association that's maybe not so much a technical issue as a more pragmatic one.

Let's say that someone has a project studio setup that includes a digital console in the class of a Yamaha 02R or Tascam 2400, two gold channels through something like the UA 2192 or similar, a few channels of MIDI keyboard and control, and a decent DAW running Cubase or the like. Maybe toward the higher end of home recording, but certainly not unheard of.

Let's say, just for the sake of discussion, that the owner decided that the optimal configuration for him with the equipment at hand was to master the sync from the 2192. Coolness. On a config of that class, good house sync is almost certainly a benefit. Until he decides it's time to replace or remove the 2192 from the chain.

One advantage of a seperate, dedicated clock outside of the actual signal chain is that one can remain flexible with the chain itself without having to worry about where the clock is based. And on a system of that class, an decent external clock like a Big Ben is only about 10% or less of the total system price, a defendable investment.

Or look at a lower budget example. A guy wants 8 channels mic in going to digital, but is on a limited budget where a Rosetta 800 or better for the A/D alone is just a wet dream. However, for that same price he can afford 8 channels of Presonus or MOTU or the like along with a Big Ben for superior clocking, and still have enough left for a couple of channels of gold conversion (or a decent DAW w/Nuendo), all clocked from a quality master. Would that not be at least arguably a defendable option?

(Assumption on both of the above is that budget for stuff like microphones, monitoring chain, signal processing, etc. is a seperate issue and, as they say, "all else equal.")

Now I'm off to dream of sugar plumbs and BMWs and 4-digit clocks while waiting for Santa to deliver a lump of coal for my old, coal-burning home studio .

G.

 

[bigwillz24]

So basically are you comparing the two converters on internal and maybe the Motu clocked off the Rosetta as a third?
I just run my RME's here so just curious.
Wayne

Exactly,

I've ran the S/PDIF out of the Rosetta into the MOTU. Then I ran a BNC cable from the word clock out on the Rosetta to the in on the MOTU. Just been playing with it like that.

 

[MessianicDreams]

Let's say, just for the sake of discussion, that the owner decided that the optimal configuration for him with the equipment at hand was to master the sync from the 2192. Coolness. On a config of that class, good house sync is almost certainly a benefit. Until he decides it's time to replace or remove the 2192 from the chain.

One advantage of a seperate, dedicated clock outside of the actual signal chain is that one can remain flexible with the chain itself without having to worry about where the clock is based. And on a system of that class, an decent external clock like a Big Ben is only about 10% or less of the total system price, a defendable investment.

Well i think for the first case, I would check whether the 2192 sounds any better with an external clock (an ADC of that quality really shouldn't). If it does sound better, then buy the external clock - if it doesn't, then don't. I would suggest using the 2192 as the house clock, it's clock should perform just as good as any external clock within a reasonable price range. It would be fairly easy to have a flexible master clock signal. a simple BNC patch-bay would solve this, with all the units which one would want to source clock from appearing on the patch bay, and 1 input which leads to a T-bar daisy chain BNC connection through the digital devices.

Or look at a lower budget example. A guy wants 8 channels mic in going to digital, but is on a limited budget where a Rosetta 800 or better for the A/D alone is just a wet dream. However, for that same price he can afford 8 channels of Presonus or MOTU or the like along with a Big Ben for superior clocking, and still have enough left for a couple of channels of gold conversion (or a decent DAW w/Nuendo), all clocked from a quality master. Would that not be at least arguably a defendable option?

this is the typical kind of set-up which benefits greatly from a good master clock - i would still question getting a master clock if there are a few "gold conversion" channels, and would probably sync off of these. I think the most important thing here would be to have the best converter running off it's internal clock, and sync everything to that - like that your top converter will have the best possible environment to do its job. i do think that it is possible to make a fairly cheap converter sound like a converter twice or 3x it's price by having a good clock input.

Exactly,

I've ran the S/PDIF out of the Rosetta into the MOTU. Then I ran a BNC cable from the word clock out on the Rosetta to the in on the MOTU. Just been playing with it like that.

you should really make sure that both are properly clocked, and not both set to internal clocking, eliminating any potential for digital issues. Using embedded signal or BNC shouldn't really change significantly, as long as cable runs are short. For long cable runs, you're much better off using BNC as the actual frequencies passing thru the BNC will be much lower that embedded (we're talking 44.1 or 48Khz vs a signal in the Mhz). As we know the capacitance of the cables will basically mean in long cable runs they act like a LPF, which can be dangerous for digital signals.

There is some discussion RE BNC vs embedded, however I don't believe the audible difference, if there is any (these are to do with reflections & phase mostly), are worth really considering.

I would set the Rosetta as the master (this is the converter and the point at which precise clocking is crucial), and then either get the MOTU to get clock sync from the embedded signal, or from a BNC connection

 

[SouthSIDE Glen]

Well i think for the first case, I would check whether the 2192 sounds any better with an external clock (an ADC of that quality really shouldn't).

this is the typical kind of set-up which benefits greatly from a good master clock - i would still question getting a master clock if there are a few "gold conversion" channels

The general assumption in both examples is that the chains perform at the very least no worse on the external clock than they would using an in-line clock for overall system sync. Of course it would not make sense to use an external clock that made things worse .

And I didn't quite type what my brain was thinking on that second example (note the timestamp on that post ). Those two channels wouldn't necessarily be gold conversion; they could be simply gold channels in the form of top shelf preamps, or (as I alluded to) even a brand new PC workstation. I was using examples of what could be done with the extra money versus the cost of 8-channels of gold conversion alone.

But the point I was trying to illustrate with both examples is that it can be the most flexible studio design in a cost-effective way (and often in a performace-boosting way, as in the second example w/o gold conversion) to treat the clocking as a seperate studio function and subsystem from the signal chain itself, much in the way that the body's nervous system is seperate from the circulatory system.

Sure the 2192 could do a perfectly fine job controlling the house in the Yamaha project studio...as long as you have a 2192 as part of your signal chain inventory. Sure two channles of gold conversion could control the second Presonus HR system very well, as long as you commit to spending that money on two channels of conversion instead of something else.

External clocking as a seperate control system than the signal path itself frees one's budget and creative flexiibility up on the signal path design itself. And those were to examples of how it could be both cost-effective and performance positive (or at the very least, performance neutral) to do so.

Maybe I have a bias because I come from a history of working in both audio and video, and not just audio only, but for me the idea of "house black" or just discrete house sync control of any protocol in general as a seperate subsystem seems a natural part of studio design, even if it is an entirely optional one on the project and hobbyist level. And with a proper order of battle in equipment choice, a quality external word clock can be integrated into a small studio system quite cost effectively.

G.