[WindowsX]

These is always a matter of tastes. You should consider using an ocxo instead of a rubidium clock, you would get lower jitter.

Rubidium clock has lower phase noise which cause jitter and better frequency stability. The advantage of OCXO is having way faster setting time while Rubidium needs 3 days to settle down. If you leave Rubidium permanently on, that'll way better since high precision applications use Rubidium and Cesium for good reason.

But as you say, some people also prefer OCXO over Rubidium, along with artifacts and less precision details that can sound more soothing to their ears in their systems. That is fine and all, not everyone can afford $100-300k systems to make proper use of Rubidium clock better.

 

[FredA]

Rubidium clock has lower phase noise which cause jitter and better frequency stability. The advantage of OCXO is having way faster setting time while Rubidium needs 3 days to settle down. If you leave Rubidium permanently on, that'll way better since high precision applications use Rubidium and Cesium for good reason.

But as you say, some people also prefer OCXO over Rubidium, along with artifacts and less precision details that can sound more soothing to their ears in their systems. That is fine and all, not everyone can afford $100-300k systems to make proper use of Rubidium clock better.

I thought rubidium had the edge on frequency accuracy. Long term. Not on jitter. The vast majority of clock makers use ocxos. Esoteric is one exception. The best regarded clocks seem to be Cybershaft at the moment. Not saying Esoteric is not good. Too expensive for me anyway. For most anyway.

 

[WindowsX]

I thought rubidium had the edge on frequency accuracy. Long term. Not on jitter. The vast majority of clock makers use ocxos. Esoteric is one exception. The best regarded clocks seem to be Cybershaft at the moment. Not saying Esoteric is not good. Too expensive for me anyway. For most anyway.

I don't know where you googled that from but most highend studios use Rubidium clock for 10MHz input having least frequency drift (which is the direct cause of jitter). How can Rubidium clock has more jitter if it can send 10MHz frequency over 1000 times more precise than OCXO?

Jitter in oscillator happened from frequency sending samples not on exact time. Maybe few ppm or ppb dependng on how accuracy that oscillator is. The reason why Esoteric is one of few DACs that use Rubidium clock in highend market was because Esoteric's higher end models use narrower PLL to lock clock signal. This will make clock module with higher precision to perform better.

I used to own Esoteric D-03 and word clock needs frequency precision to be under ±5 ppm. I used Antelope OCXO clock before and it couldn't keep Esoteric D-03 working with clock signal locked while other highend DACs I had at that time worked fine. That's probably one of reason why more affordable clock is more popular among more affordable DACs.

So the notion about OCXO being more preferable as general consensus was from misconceptions among midfi computer audiophiles. I have never seen any Esoteric owners selling his Esoteric G1X Grandioso 10MHz clock for something like Cybershaft. Besides Esoteric, CH Precision also implement rubidium clock too.

I hope this will help you realized that there's no universally better clock upgrade for any DAC. You need to pick the right clock implementation for the right DAC design. Cybershaft may work better for devices like $3-5k DACs that accept 10MHz clock input but some $20-30k DACs may prefer others. Or maybe listeners themselves will say so. Even Esoteric clocks need to be used with device from around the same generation to match sound signature and clock synchronization to work at optimal performance. There's also 10MHz/MCLK/WCLK to choose for different approach too.

I once have friends bringing 10MHz with OCXO clock and said it sounds better than 10MHz Rubidium clock in my room but to me it feels like he was appreciating 'perceived' improvements because he never experience true highend before. And it's not my job top judge him that his opinion is right or wrong. He's free to pick what he prefers more with his own money. And that makes audiophiles different from scientists because they can be right on their own hobbies and don't need to prove everyone to agree with them.

Regards,
Keetakawee

 

[FredA]

Short-term stability (what causes jitter when missing) and long-term accuracy (or frequency accuracy) are two different things. That said, the listening test is the ultimate. No arguing there.

 

[gimmeheadroom]

And I heard crapton of several DACs that cost 10 times more than Kingwa's. And it's not just money or marketing. He may build better performing DACs than some other $4k DACs but it's no where close to most $10k well known DACs in market today. For example, Rockna Wavedream DAC will clearly outperform all DACs under $5k in market today, period.

Kingwa is not selling $10K DACs and as far as we know, you have heard exactly one model. So I found your post totally irrelevant, straw-man style.

 

[WindowsX]

Short-term stability (what causes jitter when missing) and long-term accuracy (or frequency accuracy) are two different things. That said, the listening test is the ultimate. No arguing there.

Yes stability and accuracy are two different things. Some rubidium clocks have poor short term stability but what Esoteric and what I use has good enough value (1000 times less than typical rubidium clock in GPS). I don't simply use Rubidium clock because it has better accuracy but also consider stability and read datasheet to confirm.

 

[WindowsX]

Kingwa is not selling $10K DACs and as far as we know, you have heard exactly one model. So I found your post totally irrelevant, straw-man style.

Straw man what? I heard all his DACs since Compass era and stopped after a few PCM1704K models released. One of my friend used to bring Kingwa's older DAC design to my place and I gave him audition to compare between Kingwa's flagship DAC at that time against studio DAC from Mytek at half the price. He quickly sold Kingwa DAC and bought Esoteric P-05/D-05 instead.

And I don't need to share all my experience nor convince you anything. I find his current design improved a lot enough for me to buy his DAC again. No need to get so worked up because your favorite product is not the best in the world.

 

[Articnoise]

Not so long ago I tried to get Kingwa to build me an R-27HE, but with the same DAC as in R7HE mk2 instead of R8HE. Kingwa said no to this so no R-27HE for me.

The reason I didn't want an R8HE DAC is because I am very skeptical of using op-amps and switching regulators in top hifi devices.

BTW Kingwa almost always says that there is only a small difference between different versions of his gear, if you ask.

 

[Articnoise]

Short-term stability (what causes jitter when missing) and long-term accuracy (or frequency accuracy) are two different things. That said, the listening test is the ultimate. No arguing there.

Measure correlated with Phase Noise is much more important than jitter.

 

[WindowsX]

Not so long ago I tried to get Kingwa to build me an R-27HE, but with the same DAC as in R7HE mk2 instead of R8HE. Kingwa said no to this so no R-27HE for me.

The reason I didn't want an R8HE DAC is because I am very skeptical of using op-amps and switching regulators in top hifi devices.

BTW Kingwa almost always says that there is only a small difference between different versions of his gear, if you ask.

R8 is more affordable version from his flagship R7 design. Asking him like that is like "please build me DAC with R7HE board but make it cheaper on power supply and others". What makes RUN more expensive than R8 is circuit board on DAC itself so he won't build that for sure. It's going to be bad product.

Measure correlated with Phase Noise is much more important than jitter.

Both are important and yeah lower phase noise makes better PLL implementation.

 

[FredA]

Not so long ago I tried to get Kingwa to build me an R-27HE, but with the same DAC as in R7HE mk2 instead of R8HE. Kingwa said no to this so no R-27HE for me.

The reason I didn't want an R8HE DAC is because I am very skeptical of using op-amps and switching regulators in top hifi devices.

BTW Kingwa almost always says that there is only a small difference between different versions of his gear, if you ask.

Switching regulators? Depends on how the transistors are controled. There is a world of difference between transistor that are switched on/off completly and modulating the current to create a servo-control. And these servo-controlled psu are used only on digital parts.

The r7 is no different with that matter. It also uses servo-controlled regulators. The clocks receives class-a regulation however.

 

[FredA]

Measure correlated with Phase Noise is much more important than jitter.

Of course. Jitter is just an overall measure of phase noise within a given banwidth (integrated quantity). As you imply, t is important to consider where the timing errors occur in the frequency spectrum. Those two measuremente don't use the same unit. Phase noise can be translated to seconds. Jitter cannot be translated to an angular quantity.

 

[Articnoise]

R8 is more affordable version from his flagship R7 design. Asking him like that is like "please build me DAC with R7HE board but make it cheaper on power supply and others". What makes RUN more expensive than R8 is circuit board on DAC itself so he won't build that for sure. It's going to be bad product.



Both are important and yeah lower phase noise makes better PLL implementation.

I would of cours pay extra I asked for a custom version that is built on the R7HE board.

 

[FredA]

I would of cours pay extra I asked for a custom version that is built on the R7HE board.

I don't think it would fit there.

 

[WindowsX]

I would of cours pay extra I asked for a custom version that is built on the R7HE board.

I know you're willing to pay for extra on top of R27 but how much will you pay? Wouldn't it be better for you to buy R7HE or custom version on top of that?