[Thetaburn]

Thanks a lot for mentioning that, I just ordered a decent quality 75 ohm BNC from amazon that people were using for audio clocking purposes, I'm gonna keep them the same 1m length as the HTD Copper III cable I'm using right now and just use the DIP switches to change one output to 75 ohms.

Even after 50 hours of burn in on my Ock 2 I'm still not too happy with it hooked up to the U18, the clarity I was hearing without it has been pushed further back and the soundstage too closed in. Everything sounds tighter but I lose details in the mids and lose some decay. I'm really hoping that the method you described will change that for me.

My purple fuse will also arrive today for the Ock 2 and I will give first impressions before burn in.

My Ock2 has been on for well over 200 hours.
A/B ing the ock2 today does reveal a difference. I am pretty certain it is not placebo.

With the clock on, the vocalist is more isolated. All sounds are a bit sharper and tighter.

Ultimately, I prefer Sine wave (especially on high volume).
With high volume, Square wave is too bright and sharp ( akin to a silver cable's sound signature ).
Sine wave with loud volume is fine.

I have A/B one cable to the U18 vs one cable to u18 and another to X26pro.
Preference is single cable to u18.
Possible reason is that I only have one quality HT DC3 cable. Other cable is an average Lmr 400 cable ( a bit harsh vs the HT dc3 ).

 

[MartinWT]

I can only repeat that my AfterDark clock connected to the U18 works extremely well, with the clock being fed through I2S into the DAC.

No clock splitting, no connecting to both U18 and DAC.

 

[Jake2]

My Ock2 has been on for well over 200 hours.
A/B ing the ock2 today does reveal a difference. I am pretty certain it is not placebo.

With the clock on, the vocalist is more isolated. All sounds are a bit sharper and tighter.

Ultimately, I prefer Sine wave (especially on high volume).
With high volume, Square wave is too bright and sharp ( akin to a silver cable's sound signature ).
Sine wave with loud volume is fine.

I have A/B one cable to the U18 vs one cable to u18 and another to X26pro.
Preference is single cable to u18.
Possible reason is that I only have one quality HT DC3 cable. Other cable is an average Lmr 400 cable ( a bit harsh vs the HT dc3 ).

I am glad to hear you have are enjoying a difference. I have the OCK-2, the same HT cable (after trying many), and the U18 but with the R26 instead of the X26 pro. I follow the various Gustard threads with interestk particularly because the three Gustard DACs all have the K2 clock synthesiser in common, so notwithstanding their differences, there may be observations of common applicability or benefit.

To this end for what it’s worth I prefer sine with my R26 and the OCK-2 and the OCK-1 before it, irrespective of cables used. I was going to add I have found the OCK-2, along with the OCK-1, and as of yesterday the Leo Bodnar, responds very noticeably to purer power. I was using it for a while with a stock mains cable into my PSM156 power conditioner. There was a further clear improvements, exceeding my low expectations, when I added an AliX Flux-50 style filter then later a better power cable. Recommend you explore this if you’ve not already. Also FWIW with the U18 in play my system sounds better with the external clock slaving both the U18 and R26.

 

[kennyb123]

Also FWIW with the U18 in play my system sounds better with the external clock slaving both the U18 and R26.

What do you mean exactly? It’s the word clock that does the slaving in this case, and these OCK clocks aren’t acting as word clocks. The word clock is passed to the DAC via I2S and the DAC becomes slaved to it. The word clock operates at only multiples of 44.1 and 48k so it’s critical that they be in perfect sync otherwise jitter will be introduced. A reference clock remains fixed at 10 MHz and helps to keep the timing precise for the interfaces that operate at set frequencies. It’s not a master/slave relationship because it can’t be because interfaces that lock onto it are running at frequencies other than 10 mhz.

 

[Jake2]

What do you mean exactly? It’s the word clock that does the slaving in this case, and these OCK clocks aren’t acting as word clocks. The word clock is passed to the DAC via I2S and the DAC becomes slaved to it. The word clock operates at only multiples of 44.1 and 48k so it’s critical that they be in perfect sync otherwise jitter will be introduced. A reference clock remains fixed at 10 MHz and helps to keep the timing precise for the interfaces that operate at set frequencies. It’s not a master/slave relationship because it can’t be because interfaces that lock onto it are running at frequencies other than 10 mhz.

Perhaps a misnomer, I should have said providing a reference clock signal to the K2 clock synthesisers of the U18 and R26, which the R26 K2 definitely seems to be using in some manner when Ext Clock is selected even with synchronous inputs. Yes I appreciate the theory that the only relevant clock for a synchronous signal like I2S is the embedded word clock from the U18, assuming no active rebuffering and reclocking on receipt, which Gustard doesn’t claim to do. However the audible differences are clear when toggling the R26 ext clock on whilst I2S from the U18 is selected an improvement of the same character and similar if not greater magnitude to selecting external clock on the U18. It very much suggests the K2 is benefitting from the more accurate reference clock to perform its operations in some manner, say when over sampling, though I’ve checked and there are audible benefits in NOS mode too.

Over on the R26 thread we have tried via correspondence to seek clarifications from Gustard on a number of matters including in which input/R26 setting combinations the ’R26 benefits from the external clock’ (so as not to frame the question too narrowly). However the language barrier is a big one, with ambiguity and inconsistencies in their responses between what the online product material says and even the English manual which turned out to have an incorrect translation in one key respect, though unrelated to the role of external clocks.

I‘ll see if I can track down their response and share it, you’ll see what I mean by frustrating ambiguities.

 

[kennyb123]

I should have said slaving the U18 and providing a reference clock signal to the K2 clock synthesiser, which the K2 definitely seems to be using in some manner when Ext Clock is selected even with synchronous inputs.

I probably should have sought a clarification more directly. Are you running the reference clock into both the U18 and your DAC or to just into the U18? There’s a K2 in both devices so your statement is ambiguous in that regard. Same for the statement that they are both slaved. Both your statements could be interpreted to mean either way of operation. From the rest of what you wrote, I think you are saying into both, but just wanted to confirm it.

i actually initially expected that there’d be benefits into both, but @MartinWT had corrected me on that.

 

[Jake2]

I probably should have sought a clarification more directly. Are you running the reference clock into both the U18 and your DAC or to just into the U18? There’s a K2 in both devices so your statement is ambiguous in that regard. Same for the statement that they are both slaved. Both your statements could be interpreted to mean either way of operation. From the rest of what you wrote, I think you are saying into both, but just wanted to confirm it.

i actually initially expected that there’d be benefits into both, but @MartinWT had corrected me on that.

Yes I realised the ambiguities and tried to correct them in an edit above, but it froze halfway part done (hopefully clearer now). Yes, using a clock with both.

Also I should probably say at this point I am not seeking to contradict anything that Martin says about his experience, and I suspect he may be weary of discussing it, but there were a number of other X26 pro & A26 & R26 external clock users on the master clock clock thread who had a similar experience to me and even after some discussion we were unable to find an explanation for our different experience. Which is fine, it is what it is.

So that being said, I’ll see if I can track down Gustard’s email…. One mo

 

[Highfive99]

i actually initially expected that there’d be benefits into both, but @MartinWT had corrected me on that.

I perceive a clear and noticeable benefit into both. Others do too. Some systems benefit, some don't. There is nothing to be corrected on as there is no absolute.
FWIW, when I contacted Gustard they indicated to feed both.

 

[kennyb123]

Yes I realised the ambiguities and tried to correct them in an edit above, but it froze halfway part done (hopefully clearer now). Yes, using a clock with both.

Thank you. I may have to correct what I wrote earlier about the need to connect it to both devices in order to get the OCK clocks to be a net positive.

I had done some reading earlier about my REF10. I came away thinking it should provide benefit to both the DDC and DAC as it’s just providing a much better reference. How Gustard implemented it matters in this interpretation because the question arises about which interfaces they allow to benefit from this.

 

[Jake2]

Thank you. I may have to correct what I wrote earlier about the need to connect it to both devices in order to get the OCK clocks to be a net positive.

I had done some reading earlier about my REF10. I came away thinking it should provide benefit to both the DDC and DAC as it’s just providing a much better reference. How Gustard implemented it matters in this interpretation because the question arises about which interfaces they allow to benefit from this.

Totally, the devil is in the detail of implementation with this stuff, now if Headfi’s keyword search function would actually work… I’ll find and share my earlier post with their email. Grrr…

 

[kennyb123]

There is nothing to be corrected on as there is no absolute.

I disagree as one method should be more technically correct than the other. That doesn’t mean that the more technically correct will be the preferred choice as there are two many variables. Often with these things what some prefer is an increase in one kind of noise that makes another kind of noise less annoying. I think if we had more insight into Gustard’s thinking, we can free ourselves from balancing out suboptimal choices if we do choose.

 

[Jake2]

Found it!

Copied across a previous covering comment of mine referring to mmwwmm’s post with Gustard’s response. And like I said above, irrespective of how you interpret the below, I am confident that when I carefully checked this last year with the OCK-1 I found an audible improvement with an R26 connected external clock with I2S in NOS and DSD Direct modes as they are my primary use case with HQP. I should probably retest this with the OCK-2 I got a few months ago.

If you're using I2S, s/pdif or AES to connect the U18 to the R26 then the U18 will pass the wordclock to the R26 via that connection which the R26 will use (i.e. not reclock) and, depending on how you read the punctuation in Gustard's email clarification they provided to @mmwwmm (see below), seems to mean that the R26 only gets an improvement from an external clock only when oversampling is turned ON i.e. PCM NOS & DSD Direct OFF. That said I've found clear audible benefits from an external clock connected to the R26 (toggling EXT Clock ON/OFF) when it is fed I2S from the U18 when the R26 is set to DSD Direct and/or PCM NOS ON.

This is an answer from Gustard regarding the use of a 10Mhz external clock with the R26 and when its clock signal is used to reference the internal K2 synthesizer and when is not used.


When playing PCM

1. PCM NOS: OFF
The improvement provided by the external clock applies to all R26 inputs.

2. PCM NOS: On
The improvement provided by external clock applies to USB and LAN inputs, IIS AES coaxial optical Bluetooth does not apply.

When playing DSD

1.DSD DIRECT: OFF
The improvement provided by the external clock applies to all R26 inputs.

2.DSD DIRECT: ON
The improvements provided by the external clock apply to USB and LAN inputs, IIS AES coaxial optical Bluetooth does not apply.

 

[kennyb123]

seems to mean that the R26 only gets an improvement from an external clock only when oversampling is turned ON i.e. PCM NOS & DSD Direct OFF.

Thanks for posting that. My interpretation:

If using a U18 and therefore I2S, a reference clock connected directly to the R26, should provide a benefit when PCM NOS is off or when DSD direct is off. Their processor that does scaling seems to be leveraging the reference clock.

I also interpret from this that this benefit won’t be achieved when the reference clock is only connected to the U18. The U18 only passes on a word clock plus the music. The scaling will have to fall back to the internal K2 reference in this case.

If one is using a U18 and therefore I2S, it sounds like a reference clock should provide no benefit if the internal scaling is bypassed. That you still hear a benefit in this case suggests that maybe there’s more going on here.

I actually have no skin in this game as I have no intention of buying a U18. I’d just like to understand it better.

 

[Jake2]

Thanks for posting that. My interpretation:

If using a U18 and therefore I2S, a reference clock connected directly to the R26, should provide a benefit when PCM NOS is off or when DSD direct is off. Their processor that does scaling seems to be leveraging the reference clock.

I also interpret from this that this benefit won’t be achieved when the reference clock is only connected to the U18. The U18 only passes on a word clock plus the music. The scaling will have to fall back to the internal K2 reference in this case.

If one is using a U18 and therefore I2S, it sounds like a reference clock should provide no benefit if the internal scaling is bypassed. That you still hear a benefit in this case suggests that maybe there’s more going on here.

I actually have no skin in this game as I have no intention of buying a U18. I’d just like to understand it better.

That’s a fair take, but I’d suggest there may still be a middle ground of some, if lesser, benefit of providing the DAC with a lower phase noise word clock, even if the DAC’’ K2 oversampling scaling effectively does a degree of reclocking. Pretty sure that was a consistent theme of my experience with both s/pdif (via an iFi iPurifier s/pdif) and i2s via the U18.

I don’t actively use the U18 at present either, I prefer the R26 internal HQP NAA, but keen to see if a better clock and better power/grounding in my system now etc will mean the U18 can trump the R26 NAA. Also have an Afterdark Rosanna Diretta streamer I’m still setting up - will try via the U18 and direct USB into the R26.

 

[Thetaburn]

Found it!

Copied across a previous covering comment of mine referring to mmwwmm’s post with Gustard’s response. And like I said above, irrespective of how you interpret the below, I am confident that when I carefully checked this last year with the OCK-1 I found an audible improvement with an R26 connected external clock with I2S in NOS and DSD Direct modes as they are my primary use case with HQP. I should probably retest this with the OCK-2 I got a few months ago.

If you're using I2S, s/pdif or AES to connect the U18 to the R26 then the U18 will pass the wordclock to the R26 via that connection which the R26 will use (i.e. not reclock) and, depending on how you read the punctuation in Gustard's email clarification they provided to @mmwwmm (see below), seems to mean that the R26 only gets an improvement from an external clock only when oversampling is turned ON i.e. PCM NOS & DSD Direct OFF. That said I've found clear audible benefits from an external clock connected to the R26 (toggling EXT Clock ON/OFF) when it is fed I2S from the U18 when the R26 is set to DSD Direct and/or PCM NOS ON.

Thanks for the information.
Will A/B again later using NOS mode off. I have 2 lmr 400 cables and 1 HT dc3 cable on hand to try.

I do prefer the sound signature of NOS on and I am pretty certain there are benefits even with NOS on just using one cable connected to the U18.

There is no DSD Direct mode for the x26pro. So I assume that's a moot point.