GoodSoundClub - Romy the Cat's Audio Site

In the Forum: Didital Things
In the Thread: Myth of CD clocking = True/False

Post Subject: Digital Clock accuracy and sound qualityPosted by custodian on: 4/24/2015

New member, first post, so go easy on me!
Ive been playing around with clocks in digital audio for a while. Current system uses a DCS Scarlatti stack with the Scarlatti clock. Synching to the Scarlatti clock rather than using the Scarlatti DAC as master does make the sound better to my ears, reducing some edginess that people might call digital glare.
About 6 years ago, I decided to experiment with an external 10MHzclock and I got hold of an ex telecom rubidium unit (available cheaply from your usual auction site). Once Id managed to get it to output a 10mHz ref signal and set up the DCS to accept that reference, I was impressed with the sound improvement.
Now that is quite surprising because, whilst the rubidium clock has a better long term stability, that should not be really important for the music system. What matters is short term stability or phase noise which, for the rubidium is worse than for the OCXO crystal oscillator reference used in the standard DCS clock.
I spent some time with the DCS technical people comparing rubidium with the DCS oscillator and we all agreed it sounded "different" although ultimately the DCS team couldn't agree whether it s better or worse!
Last year I decided to see if improving the absolute specs for the master clock would have an impact so I managed to get hold of an Oscilloquartz BVA block. The BVA clock is an woven controlled oscillator in which the suspension for the crystal oscillator is also made from the same crystal. Complex to produce and expensive, the phase noise is one or two orders better than the standard DCS OCXO.
Results to my ears were exceptional with a removal of more digital grunge from the sound.
Problem with that solution is that since the sale of Oscilloquartz last year, they have stopped manufacture of the BVA clock. I believe that was due to the difficulty in getting acceptable manufacturing yields.
One Japanese company SFZ Sforzato produced a commercial audiophile version based on the Oscilloquartz clock which they retailed for around $40k. Its still listed as available which I assume means they haven't sold many and are still working their way through remaining stocks.
Conclusion is that, to my ears, clock improvements affect sound. Better short term stability is important although rubidium, with worse short term stability, can still sound better that a good OCXO. Best clock is the BVA OCXO which, sadly, is no longer being manufactured.

Rerurn to Romy the Cat's Site