I’ll preface this post by saying that I don’t have much hands on experience with high end clocking. I’ve been in plenty of rooms using very expensive clocks but haven’t had a chance to A/B those. Most of this post will be relating to the theoretical aspects of clocking. Now, I’m the first person to say that theory has very little bearing on what’s actually audible but I hope that this post can make you question certain products on the market and their marketing. They may sound fine but the marketing promotes the wrong ideas.
For anyone in the mixing or mastering field, I strongly recommend reading the following article. This post is largely based on that article. Unfortunately the images on that article don’t seem to be available anymore.
https://pinknoisemag.com/pink-papers/pink-paper-002/
What do you want from a good clock (and cables for clock transmission)? Low jitter. What is jitter? Basically variations in the clocking. One cycle of the clock may be a little early or a little late. This can be caused by the clock, distortions in the cable, problems in the receiving piece of gear, or any retransmitters in the path. Jitter results in the sample being played (or recorded) ahead or behind where they should be which distorts the audio. This normally translates to high frequency noise and lack of clarity.
As you can see, we need clocking (for the sake of simplicity this includes cables and everything in the clocking “path”) that’s as stable as possible. No variations (jitter) caused by any kind of instability or distortion.
Nowadays there are countless clocking options on the market. Some better than others. The article linked to includes comparisons of various clocks on different converters (and their internal clocks as well). One of the biggest selling points when it comes to clocking is “atomic” rubidium 10MHz clocks such as the Antelope 10MX. On the surface it seems like a great idea. It’s worth pointing out that these units are little more than an off-the-shelf oscillator in a box. The oscillator unit itself costs $1500. How Antelope justifies $6500 for the 10MX, I’m not sure.
These 10MHz clocks will supposedly stay in sync for years which on the surface seems like a great idea. Essentially if you set up 2 separate units, years later they will still be in sync with each other. The problem with this is that in audio we generally don’t care about things staying in sync for that long. If in 1 hour the clock has drifted by 1 sample, no one will notice unless maybe you’re recording a single take for multiple hours. Even then, I believe the drift from non 10MHz clocks is much less than that.
The problem is that these clocks aren’t made to be stable in the short term. They may not drift but they have higher jitter. It’ll make sure to put out a pulse at every interval but that pulse may not be at the perfectly correct location in time. 10MHz clocks do not belong in audio. Audio uses require clocks which pulse at the perfect timing interval regardless of if that eventually drifts by a little which wouldn’t be noticeable. This is how you get the lowest jitter.
Learning this should make you question Antelope’s product line. Their top clocking option is the 10MX atomic clocks. These units are used to improve both their crystal clock units (ie driving a non 10MHz clock with a 10MHz clock) and their interfaces which accept the 10MHz clock signal. I have no doubt that adding the 10MX to their other products greatly improves how they sound. If 10MHz clocks aren’t what we want yet they improve the sound of their units with the non 10MHz clocks what does this say about the quality of the clocks they’re throwing into their products? I believe they’re purposely shooting their own products in the foot to make the 10MX be a nice upgrade (and gives them $6500…). As a result, I stay away from all Antelope products.
A couple of engineers I know who own the Antelope clocks (10MX or the older 10M) have switched over to non 10MHz clocks which cost a fraction of the price and they say sound better. These units include the Black Lion Audio MKIII XB and the Grimm Audio CC2. In my own system, I don’t really have the option of using an external clock but if I did, I’d be using one of these two units.
One thing I haven’t touched on is whether something like a DAC being used on its own will sound better with an external clock. As far as I can tell, it seems that different units behave differently so in certain cases they can be improved and in others it’ll sound best with internal clocking (and any other great slaved to that unit).
One last thing to mention is that clocks transmit a square wave which the receiver has to look at and determine when to “switch” to the other position. A cable (or anything else in the signal path) can affect how this square wave behaves by rounding off the sharp corners. This makes the receiver have some level of “confusion” knowing exactly when to switch position. Again, this results in jitter. While it’s just sending 1’s and 0’s, the quality of those 1’s and 0’s is very important to having good clocking which is why the cables used for clocking (and any digital connection carrying clocking) can make a big difference.