For Gustard R26...which bang for the buck: DDC or LHY OCK clock?

@peareye I've had the R26, the Ock-1, Zen Stream, U18 & SU6 for 6 months, and more recently got the Ock-2 and LHY SW-8 switch.

The Ock-2 is a clear step up over the Ock-1, in resolution and refinement, though the Ock-1 is warmer, smoother and perhaps more forgiving of average sources as a result. Still, the OCK-2 is the new incumbent.

My DDCs and Zen have been collecting dust for a while now TBC, I much prefer Roon/HQP via the R26's internal renderer... particularly since I've progressively improved my upstream network path. I do dust the DDCs off every now and then to recheck with new USB or power cables for eg, but there's no question in my system the internal streamer remains so much more musically engaging - more dynamic, weighty and resolving with a broader, deeper and taller soundstage and a calmer background. 

So for the R26 rather than DDCs I'd recommend an OCK-2 and a good switch like the LHY or Fibre Media convertors. 

@soix Yes i2S - a Blue Jeans Cables FE cable which although very well regarded up to say the $200 range I is likely a limiting factor vs say a Tubulus Argentus, which would cost as much again as a DDC which I’m not in a hurry to do. USB cable is an also highly rated if a bit more boutique but well priced Gothic Audio Semperfi Outsider pure silver.

@soix - I bought two BJC Belden FEs - a 1ft and a 2ft, TBH have only been using the 2ft one for convenience as the 1ft would be a pain. That said I'll have to give the 1ft one a go.

I've read conflicting things re optimal cable length for digital cables, though expect it depends very much on the application. The main variable I've found for clock cables at least of which I've tried five is cable quality (conductor, shielding and termination quality) rather than length per se.

I should clarify the sound via U18 via I2S gives a very good sound too, it's just that the internal renderer is even better - the differences are probably very subtle to a casual listener. They both sound great, but to me the LAN is satisfyingly better.

@pinwa Awesome - welcome to the club! Your observations of the character of differences are consistent with mine and other users of master clocks. I find that far from being fatiguing it is less so as notes are more accurately (and naturally) rendered, losing some artificially etched edges and glare.

What brand / model cables are you using?

@peareye Given your clarification that you predominantly listen to YT off your MacBook I agree that a USB based connection is what you'll need to use so considering DDCs to help optimise it makes sense. 

Reason being, as notwithstanding my preference for the sonics of the R26 LAN input/renderer, the R26 doesn't support any LAN streaming protocol I'm aware of that'd support YT audio. Happy to be corrected here if there's a workaround way to pipe it via UPnP but doubt it'd be easy or ideal if there's video lipsync considerations.

I also listen to and watch a bit of YT, in my case via my TV, using its optical out. I feed this to an Ifi s/pdif iPurifier 2 reclocker (a DDC) then into the R26 coax digital input. This sounds much better than going direct into the R26 optical input which doesn't do any reclocking of the jittery signal from the TV (so is thin and glare-y) and slightly better than going USB direct from my Mac Mini into my U18 DDC but not as good as the LAN for equivalent non-YT music.

That said having recently experienced a substantial improvement from inserting the Holo Audio Titanis USB regenerator/ galvanic isolator in between the Mac and the U18 I reckon that would likely now better the TV/Ifi chain for YT content. I must try it. 

Re clock cables the one you referred to looks to be an HDMI cable which could be used to carry i2S signal from a DDC to the R26, but for the external clock connections (ideally to both DDC and R26 so they're synchronized) you need a BNC-BNC terminated 50 ohm coaxial cable like the Gustard C2 or LMR400.

@soix Thanks for the recommendation of trying the shorter I2S/HDMI cable length - there was a subtle difference in favour of the shorter 1ft length vs the 2ft incumbent, though I guess if there’s such a thing as burn in for HDMI cables it might improve further. Incidentally the logic of shorter is better for I2S makes intuitive sense to me and may also help explain why the internal streamer of the R26 (and other high quality DAC/streamers) can sound very good indeed - the internal i2S connection will be extremely short - just the distance between the streamer and main DAC circuitboards essentially.

@peareye The HT DC III is an excellent cable - clearly the best one I’ve tried. HT >> LMR400 /> C2 >> Minicircuits 50ohm >> Pro audio 50 ohm > Hifi coax digital 75ohm.

Best performance/ price ratio is the LMR400 - which after a week’s burn in - and it really does take that long to sound good - it surpassed the C2 in terms of resolving power and dynamics, though the C2 was warmer and smoother so more forgiving. Downside of the LMR400 is it is really thick and rigid plastic tubing like a plastic hose so is cumbersome to use and can feel like it is putting a fair bit of lateral stress on BNC sockets.

@pinwa John Swenson certainly knows his stuff (it’s an interesting paper isn’t it) so all the best with that - should be a cost effective route to achieving a good standard for sine wave. Incidentally I’m a sine wave fan - worked best for me with the OCK 1 & 2 with the R26 with all the cables I’ve tried so far.

It's remarkable how much difference a very good clock makes isn't it. Well you'll be pleasantly surprised by the further benefits that are possible with better cables (or cable / filter combos). 

Source of the HT DC III: Audio Accessory HK - a trading/website name of Labkable an official distributor of HT cables:

https://www.audioaccessoryhk.com/pr...y-dc-iii50-ohm-single-crystal-copper-成品線-bnc/

You need to use Google webpage translate.

You order and pay for it then they email you afterwards to bill you separately for international shipping, total cost works out around USD150, so not cheap but a little cheaper than the C2 but with performance on another level. Its resolution and refinement is superb.

You may be able to source it in the States too but likely more expensive.

@peareye I’d say invest in the best clock you can afford, then upgrade other stuff as and when you can afford to. The OCK-1 is great, I was very happy with it, but after 6 months temptation got the better of me and now I’ve heard the OCK-2 I can’t go back. 

What’s your network chain to your MacBook btw? As if it involves an Ethernet connection I’d strongly recommend an Ifi LAN iSilencer. For USD90 it was an excellent bang for buck upgrade to my streaming based system and I already had a  very good LHY switch.

@peareye Nice one. Let us know how you get on once it arrives.

Understood re your preference for using USB > DDC > R26, I was asking about how your MacBook connects to the internet as I've found that network chain improvements upstream of my Mac Mini bring clearly audible improvements irrespective of which input I use on the R26. The ifi LAN iSilemcer is one of the simplest and most effective such tweaks I've tried.

@sls883 Both! Both is best for me, after quite a bit of experimentation. Holds true for both PCM NOS and OS modes and DSD equivalents, though effect of externally clocking the R26 is more pronounced when using PCM NOS OFF and DSD Direct OFF i.e. internal oversampling ON.

The Harmonic Technology Digital Copper III is an excellent cable, quite a step up from the Gustard C2 and the cheaper but excellent performing LMR400. Ramble Audio HK has the best pricing and shipping cost globally.

@boulder_bob  that’s the oft quoted theory, certainly. But it depends on fhe DAC design and even the particular setting selected. The majority of Gustard X26/A26/R26 users over at HF, including me as noted above, find clear benefits to clocking both even when using I2S. When using R26 internal OS modes (NOS OFF) the benefit of clocking the R26 is greater than for clocking the U18, with NOS ON the reverse is true, with the effect of clocking the DAC more subtle but still satisfying and additive in both cases. The magnitude of the effect is very cable dependent. Do a bit of experimentation and trust your ears is always my advice here. I use a DIY-tweaked OCK-2 but plenty of Gustard users at HF use AD clocks.

Hi there, sorry been very busy with work. Some responses:

- I found sine works better with R26

- HF refers Headfi

+ tweaks for OCK-2 were inverting it, removing cover and stacking a sandwich of tungsten cubes and fo.Q damping material on OCXO and toroidal PS, looks like a mad scientist experiment with a graphite foil replacement for the aluminum base with cutouts like a tube amp would have for the tubes. Crazy how sensitive OCXOs are to this, have theorized it may help both with thermal smoothing of the oven due to more mass and of course vibration damping. Others have found similar with Tungsten cubes. Also a lot of signal grounding of spare BNCs (switch 8nternal GND ISO jumpers to GND for max effect) and a Flux-50 style AliX AC RFI filter. Carbon fibre fabric (soft not cured) shroud on top. All these each have a cumulative  audible effect  that I did a lesser version of these tweaks with the LHY SW-8 which has very similar design. I would like to revert them to normal without the crazy and precarious stacks but they sound too good so each time I start to test this I stop there. Not for everyone but has given me a taste of what more expensive clocks and lower phase noise oscillators must perform like. For now this is what my budget permits.

@sls883 Cool, I hope it goes well for you. Noting leaving it on for the oscillator  stabilize for days or even weeks initially will improve the sound, cables matter too. 
Re coax yes you’ll still get a benefit, more if R26 has PCM NOS set to OFF I.e. it is doing internal over sampling. 

@sls883 Never noticed much if any difference with copper caps TBH for RCAs, haven’t tried for BNCs, maybe get both if cheap, see if you notice a difference. Other factors have far more effect IME..

sls883 Sorry I just realised you literally meant shorting the ground/shield and signal conductor, hmm not sure I can endorse that with a sensitive clock. I’d google or better get ask LHY before proceeding.

I was picturing a metal shielding cap that does not make contact with the inner conductor, just the outer shield/ground.

@sls883 Awesome! I think the R26 in particular benefits from a better quality oscillator, as its internal one is a little jittery in some tests I’ve seen. The differences you describe are typical of better temporal/time domain precision, soundstage and 3D/decay one might expect but the extent of the effect on bass is quite something isn’t it. Who’d have thought better timing would make bass more punchy and weighty. Expect longer decay tails and richer bass and midrange texture with better clock cables. Also a tweak to try in due course that has worked for me with every cable, cheap or more expensive, is to wrap the BNC plug/socket junction at both ends with about 1” copper foil tape with conductive adhesive - costs next to nothing but sounds like a solid cable upgrade.

@pindac Cheers for the link, gee that guy went deep into the weeds on controlling temperature variation. Interesting.

FYI I have a 1”/2.54cm cube tungsten block as the base of the stack on the OCXO. With tungsten’s high density - same as gold - it weighs 316 grams so has a lot of thermal mass for its size (at least 10x a largely air filled OCXO is my guess) so once the assembly is up to target temperature (which it will delay proportionate to the initial mass) must help with smoothing micro-variations of the oven temperature, presumably also smoothing the required activity of the oven’s heater and hence its current demands on the OCXO circuit. But there’s clearly vibration damping effects too - for which tungsten is known - adding what is % wise negligible mass of fo.q polymer sheets and much smaller cubes on top has quite an additional effect on the sound. Longer decays, better transient attack etc.

I bought an equivalent of the Pasternack from Mini-circuits, another specialist manufacturer of RF cables and other stuff for industry and lab use. Selected it based on very good specs for shielding and low levels of internal reflections, the best specs I could find for BNC terminations when measured at 10Mhz.

https://nz.mouser.com/ProductDetail/Mini-Circuits/141-24BM+?qs=xZ%2FP%252Ba9zWqY%252BuUDA9xTQ2Q%3D%3D&countryCode=NZ&currencyCode=NZD

It sounds good vs some other cheaper BNC cables I had but unfortunately not as good as an AliX LMR-400 clone which is in turn some way behind the Harmonic Tech Digital Copper III. It seems there’s more in play here sonically than just the fundamentals of shielding level, accurate impedance, good terminations and low reflections.