Benefits of adding a 10MHz Master Clock to a digital system


As a long time DCS stack owner (first Paganini, currently Vivaldi V2) I’ve been a convert to the value of dedicated clocking systems. In the context of a DCS setup this means including a dedicated clock unit that provides a combination of 44.1KHz and 48KHz signals to each of the other units in the system (the transport, DAC and up-sampler in the case of a full four box stack).

The DCS clocks are pretty darn expensive boxes and while they (like all DCS gear) benefit greatly from upgrading power cords, 75 ohm interconnects and feet I had never really thought that adding a further reference clock would bring any benefit. However piqued by the following recent review of the Vivaldi One in which the addition of both the Vivaldi clock and a Cybershaft 10MHz reference brought great benefit I wondered if I’d been mistaken and if an external 10MHz reference could add even more to my system.

Some on line research quickly made it clear that the Cybershaft OCXO clock used in the review is unobtanium so the question is who else makes a good clock? Online commentators seemed to think quite highly of the Ref10 from Mutec so that’s what I opted for from an Amazon seller. The Ref10 is a solidly made but very utilitarian box. It provides up to 8 outputs all on BNC, a mix of 6 at 75Ohm and 2 at 50Ohm so will match with whatever you have.

The DCS is 75Ohm so as soon as I had the Ref10 I installed it using a generic 75OHM BNC and the stock power cord. Notwithstanding recommendations to leave it on for 48 hours to stabilize the first impressions were very favorable. With the addition of the reference clock I was hearing further into the recording and also appreciating a surprisingly large increase in rhythmic consistency and bass realism (i.e. bass seemed crisper and less boomy with the elimination of a sense of overhang in the 80-120Hz range that had plagued my system before)

What then amazed me was how much further improvement upgrading the power cord (to Marigo Iridium V2) and BNC (to Marigo Apparition Extreme 75Ohm) brought - even more air and clarity. The final icing was optimizing the support under the Ref 10 with Marigo RHZ feet on a Marigo platform. I've now had the Ref10 in my system for several weeks and frankly it's probably the first time I'm really beginning to believe that digital can give my analog setup a real challenge -- with how I have it now I feel as if I'm hearing two increasingly similar takes on the same absolute -- digital converging on all the best in terms of space and realism I've found in analog, while the latter (with the recent addition of Ron Heydrich's latest power cord on my turntable PSU) getting the speed stability and rhythmic integrity that digital can do so well.

Overall in the context of a megabuck DCS setup it was surprising how much benefit a relatively modestly priced ($3595) professional grade add on brought. That I then added nearly $9K of tweaks to it probably speaks more to my obsessive nature (and budget I guess) but overall consider me a full convert to the benefits of 10MHz master clocking in any attempt to scale the heights of digital reproduction.
128x128folkfreak
I recently acquired the Mutec Ref 10 and I am having great success with it hooked up and synchronizing my dCS Vivaldi One w/Vivaldi 2.0 Clock via its Master Reference input, two (2) SOtM sNH-10G Ethernet Switch Hub w/clk. (in parallel) into their 75OHM inputs and SOtM’s tX-USBultra special edition USB Regenerator via its 50 Ohm input.  I immediately heard a noticeable difference with receiving greater clarity and insight into the music and the music is more layered or dimensional.  I am also looking to update my 50 and 75 Ohm BNC Cables to SOtM digital BNCs.  Overall going to the Mutec Ref 10 or an external clock is Highly recommended
Post removed 
My clock journey continues. I ended up replacing my OP21A with a Stanford Research Systems PERF10 and custom power supply by PLIXIR. Conversations with dCS engineers helped me to understand better how the Vivaldi Master Clock behaves with the DAC, and any other clocked components, and how it interacts with an external reference. OCXO crystals, no matter how low noise, are unlikely to significantly improve the performance of Vivaldi Clock. dCS uses VCXO instead of OCXO for a reason. What can improve its performance is referencing to an atomic clock or GPS-disciplined clock that has lower noise/jitter than Vivaldi, *and* is more stable in the long term. dCS even note this in their collateral about the Clock: "Vivaldi Master Clock may be slaved to an external reference (such as an atomic clock or GPS reference) if increased accuracy is desired." In that category, you want something that can actually make the Vivaldi more accurate. Whether you can hear that accuracy is a different matter, of course. 

STUTE, as a rubidium clock, might offer this accuracy, but not necessarily. When Abendrot writes this stuff on its website—"Through the exclusive technology of Abendrot, STUTE produces the coveted 10MHz signal ideal for digital audio."—one's radar ought to start beeping. There is nothing "coveted" about a 10mHz signal, and it is certainly not the  ideal for digital audio. Any engineer could tell you that a signal that did not need two be recomputed for digital audio's multiples would have been preferable to 10mHz, which is just a legacy of the communication industry. So, that's just marketing blather. Further, Stute's claimed accuracy of ±5×10-11 is no better than quite a few out there [same as PERF10], while its phase noise is good, but not even as good as the PERF10. So, does that mean the STUTE doesn't sound better? Maybe, maybe not. It does suggest there might not be too much reason for its ridiculous price tag. BTW, the Merging +CLOCK-U has even better specs for an OCXO, doesn't require an external reference,  and costs a bit less than the STUTE. I have heard the difference it makes.

For us "poor" Vivaldi Clock owners, the SRS PERF10 is the bargain of the options out there. But for something that really nails it, consider a GPS-disciplined rubidium with extremely low phase noise. Almost as good as a cesium, but much cheaper. Here's an example: https://synreference.net/catalog/item/10-Channel-GNSS-Locked-Low-Noise-Rubidium-Audio-reference-Kron...
I run an Antelope Audiophile 10m directly into a Zodiac Platinum Dac. Before buying the admittedly expensive clock I experimented with a cheap Chinese Ocxo clock bought on Alibaba, which very much validated the upgrade.

The step up to the Antelope clock was significant; most evident were improved spatial resolution of individual instruments in the soundstage, snappier leading edge and longer reverb.

After the initial trial I replaced the stock BNC cable with Shunyata sigma BNC as well as using a Nordost Brahma power cord. In addition I use Final Daruma footers with Black Ravioli pads on an isolation mat and have the clock covered in EMI/RFI rejecting fabric. (Be cautious, though about them running hot...) For such an expensive addition, the amount of fine tuning needed is astounding.

High end digital replay is very much about optimum clocking, particularly when upsampling. I second the view that it now gives my analogue setup a run for its money.
Well, several manufacturers I’ve read from, including Mytek, state categorically that due to the distance from an external clock to the DAC chip itself the external clock would always have higher jitter than the internal clock built into the Brooklyn and higher end DACs. Keep in mind that Mytek DOES provide for an external clock syn to their DAC’s but they state this is only for use when synchronizing multiple different devices such as in a recording/mastering studio. Barring that situation they recommend the internal clock as being categorically better.

I’d really have to have a mind-blowing experience with a clock to go against that!