A novel way to massively improve the SQ of computer audio streaming

[austinpop]

56 minutes ago, Superdad said:

 

I think all of you in this thread entered that realm a long time ago! 

 

No argument there!

 

But oh it sounds so goooood!!

[austinpop]

8 hours ago, Superdad said:

 

I think all of you in this thread entered that realm a long time ago! 

 

Tonight will be one of those surreal nights, then!

 

I borrowed @limniscate's LPS-1 for the night, so I have now got 3 LPS-1's in the house. Once the family is abed, I'll be doing some comparisons. I want to see if my Y-cable is holding back SQ. I am going to compare:

 

 

with:

 

 

It should be very interesting. For the sake of my wallet, I hope it is a wash, or close.

[austinpop]

On 5/27/2017 at 8:59 PM, austinpop said:

 

Tonight will be one of those surreal nights, then!

 

I borrowed @limniscate's LPS-1 for the night, so I have now got 3 LPS-1's in the house. Once the family is abed, I'll be doing some comparisons. I want to see if my Y-cable is holding back SQ. I am going to compare:

 

 

with:

 

 

It should be very interesting. For the sake of my wallet, I hope it is a wash, or close.

 

Well, after 2 nights of listening, I must say it is quite close. 

 

Having an LPS-1 dedicated to each of the switch, sMS, and tX does add a touch more authority and ease to the music, but you have to really listen hard to distinguish. Without that point of reference, my shared LPS-1 between the switch and the sMS sounds wonderful.

 

Obviously, with no budgetary restrictions, a separate LPS-1 is ideal, but I know now that my config is really close to that ideal.

 

Like an ultra cap bank, I will let my coffers recharge for now. Eventually, when @limniscate receives his Hynes SR7 far, far in the future, I want to see how that improves things over the LPS-1s.

 

And the dance continues...

[austinpop]

4 hours ago, Superdad said:

 

AFAIK SOtM have never published a phase-noise plot for their clock board.

 

The 25.0MHz Crystek CCHD-575 we use in the ISO REGEN typically has 10Hz offset phase-noise values of -108dB to -112dB (Crystek admitted to me that they need to update their published plots for the 575; current production is vastly better than what is shown on the web site; The random samples they sent us with individual plots are where the above numbers come from.)

That's quite amazing for a production XO that fits in a small space, does not take a lot of power, and does not cost a bundle to use. 

 

Yet all this talk of clocks leaves aside actual USB output signal integrity (see eye-pattern) and impedance match.  Something we know a thing or two about. 

 

Of course, in the end what matters is how it all sounds to you (and maybe if there is anything left in you wallet afterwards).  

 

 

I would so love to compare and contrast the ISO-Regen in my Ultra chain, but it does get surreal in terms of LPS-1s I'd need.

 

I know Alex would give me good terms in terms of return policy and so on.

 

Still thinking about it. Luckily my recent Austinpop-alooza spring sale was a sell out, so the balances are +ve again. ??

[austinpop]

3 hours ago, limniscate said:

I just placed my order for 12V dX-USB Ultra & sMS-200 Ultra both with external master clock option, switch mod, 4 dCBL-CAT7, 2 iSO-CAT6.

 

I think insiders would call this the "full Romaz!"

[austinpop]

7 hours ago, Narcissus said:

Rajiv, I have sourced a 5 port Zyxel switch GS105B V3 which I intend to buy but it seems it requires a 9v, which I believe is the same requirement of your switch "Zyxel GS108B V3".

 

However, the above mentioned Dlink switch requires 5V. So I'm kinda confused..

 

Hmm, my GS108b v3 requires 5V. See the picture here: https://www.zyxel.com/uk/en/products_services/8-Port-Desktop-Gigabit-Ethernet-Switch-GS-108B-v3/

 

I asked May, as part of replacing the stock regulators, to use linear regulators that would allow me to supply 7v. They (SOtM) asked me to limit the number of ports used to 2-3, which is exactly my use case.

 

My switch runs very cool, so ive had no issues with excessive heat at this voltage.

[austinpop]

19 hours ago, Johnseye said:

Here's some very good, simple info on master clocking.  It's from a recording studio perspective so keep in mind the AD references, but the same principles apply.  The only caveat I can think of is that with the increasing number of devices we are utilizing in these chains, that drift occurs and can't be recovered.  This should be an indicator to simplify, not add more devices.

 

http://www.soundonsound.com/techniques/does-your-studio-need-digital-master-clock

 

And to summarize key points.

 

"The issue here is that it is much easier to design a good clock circuit using a fixed crystal inside the unit than it is to design a circuit which accepts an external clock signal and synchronises to it well. External clocks are likely to acquire some element of jitter (largely because of the inherent effects of clock cabling), and the common slave‑locking circuits can introduce further timing variations of their own, both of which can be quite hard (and expensive) to remove. As a result, an A‑D will often perform less well when synchronised to an external clock than when it is running on its own internal crystal. The noise floor may rise, and there may be more low‑level distortion products and artifacts. That being the case, it makes sense to use a device which can handle external clocks well as the slave, and a device which works poorly on external clocks as the master. In that way, the maximum audio quality can be achieved for all devices."

 

I already posted the information on interface jitter from this article in a previous post, but here's some critical information to consider when the SOtM uses 75ohm BNC interfaces for the clock chain.  I would expect Lee to ensure this does not occur with their interfaces.

 

"It's also critically important to terminate word clock signals correctly. The word clock input terminal (normally a BNC connector) on some devices is deliberately unterminated (high impedance), while some are permanently terminated and others are switchable. If the line is unterminated, the clock signal will be reflected back down the cable, which could prevent some devices from detecting and locking onto the word clock information. The usual way to apply a termination in these cases (as shown by the diagram, left) is to fit a BNC T‑piece to the word clock input terminal, attach the clock cable to one side and a 75Ω termination plug to the other. Double terminations (where an extra termination is added inadvertently to one already on the line) will result in a low-level signal with the same potential problem."

Conclusions

Overall, it should be clear from these tests that employing an external master clock cannot and will not improve the sound quality of a digital audio system. It might change it, and subjectively that change might be preferred, but it won't change things for the better in any technical sense. A‑D conversion performance will not improve: the best that can be hoped for is that the A‑D conversion won't become significantly degraded. In most cases, the technical performance will actually become worse, albeit only marginally so.

Having said all this, the use of a master clock may well make a digital audio system much more convenient and stable to operate — and that may well be a very desirable benefit in its own right, easily outweighing any minor performance compromises of slaved A‑D converters. Or it may ensure the required synchronism between sample rates and video frames is achieved, which is absolutely vital in any video‑related work.

The obvious conclusion is that in simple digital audio setups a master clock is usually unnecessary, although it remains critical that multiple digital devices are clocked sensibly. In more elaborate digital audio systems, a master clock can make the task of slaving multiple units much easier and neater, and allow the system to operate more reliably. In systems where digital audio is being used in synchronism with video, an appropriate master clock is absolutely essential. But in any of these cases, the use of a master clock will not improve the audio quality achieved by the converters in any technical sense — and the most expensive clocks fare no better in this regard than the least expensive. The only relevant criteria for purchase is whether the clock provides the facilities, inputs and outputs required, and is designed sufficiently well to conform with AES11 Grade 1 standard.  

 

This is an interesting article, and reinforces what I had previously read about master clocks, and why I had never really dabbled in this space. It's only when Roy described his findings with the sCLK-EX that my interest was piqued. And now, having experienced the SQ first hand, I am sold on the benefits of a superior clock.

 

I should preface the next remarks by saying that I am not an expert, nor even very informed, in the realm of digital audio clocks.

 

The clock distribution that Roy, I, and others have achieved with the configuration like mine, shown below, is not a master clock topology. In a master clock topology, each device has its own independent master clock, that generates all the required frequencies within the device. In addition, it has the ability to slave to an external master clock, when one is connected to its master clock input. This input is at a standard frequency of 10MHz (from what I've seen). The internal master clock, now a slave, synchronizes to this master clock, but still generates the derived frequencies within the device itself.

 

 

In contrast, in our sCLK-EX topology, there is only one master clock across all three devices. The sCLK-EX board in the tX-USBultra is generating clocks for all the devices. As part of the mods, the default clocks in the switch and the sMS-200 are replaced by clock signals sent from the sCLK-EX board. Without the tX-USBultra, and hence the sCLK-EX board powered on and active, the switch and sMS won't even function. These clock signals are at the required frequencies for the specific devices. If I remember correctly, its 24 and 25MHz.

 

Clearly, by keeping these clock wires short - 20cm in my case - whatever jitter is added by the clock wires does not significantly detract from the SQ improvements.

 

 

Here's what we don't know. If, in this topology, the sCLK-EX board in the tX-USBultra now becomes the slave to an "even better master clock" like the Mutec Ref10, or the as-yet-unannounced SOtM master clock, will this actually improve the SQ? Or will the well-known concerns about master clocks, as described in the referenced article, come into play and render the master clock improvements moot?

 

Only time - and our ears - will tell.

 

[austinpop]

9 hours ago, limniscate said:

Obviously, I didn't like reading @Johnseye's post about master clocks since I just ordered the external master clock option for both the dX and the sMS.

 

Eric,

 

See my post immediately above. You're still locking in the sCLK-EX improvements.

 

The only thing you don't know is whether, in the future, if you buy an amazing master clock like the Ref 10, you will hear an additional improvement. That will determine whether the extra - what, $200? - you spent for the master clock inputs was worth it or not.

[austinpop]

15 minutes ago, sadekkhalifa said:

@austinpop does the Ref10 need a master clock with 12v connector or 6-9v connector?

 

No idea. I have not really looked into this. It's an upcoming device from Mutec.

[austinpop]

I was really dismayed to read reports about the fluctuating and increasing prices for the Ultra stack. However, based on what I can see, the price increase is not too great. I paid approx $1235 for what is listed here: https://sotm-usa.com/collections/ultra-series-mods/products/sms-200-mod-tx-usbultra?variant=40555633420 for $1400. But some of that difference is due to the use of the more modular clock cables with SMB connectors.

 

So yes - a price increase, but not as bad as I thought.

 

Ultimately, it's economics. A company like SOtM may certainly choose to test how much price increase the market can bear, but at some price point, the demand will sag, and the company will need to make some hard choices. We'll see.

 

IMO, the way I did it, where there is exactly one sCLK-EX board in the chain, and all 4 clock taps on the sCLK-EX are used in the chain, has the best SQ and is the most cost-effective.

 

Let's look at it this way. Say you already own an sMS-200, with an LPS-1. What are your options?

 

Option 1: The SOtM Ultra chain

• sMS-200 mod + tX-USBultra + switch:         $1400
• LPS-1 to power the tX-USBultra:                     $395
• Total:                                                               $1795

Option 2: The Uptone ISO-Regen with LPS-1:       $655         bundle price

 

These are a couple of the key options, among many, facing buyers in this space today. I think we are all eagerly awaiting anyone who can compare these two configurations and see whether the ISO-R can approach, or even exceed the Ultra chain, at a fraction of the cost.

 

I've said it before and I'll say again - if anyone is willing to lend me an ISO-R for a few days, I'll be happy to try this in my system and report back my impressions. Most likely, I can also enlist @limniscate into the effort. Since he has an mR, we should also be able to compare a third option: mR+ISO-R. 

 

Going back to what started this post, ultimately buyers (and economics) decide the option that best works for them.

[austinpop]

30 minutes ago, Bamber said:

It would have been a lot cheaper for me to purchase a new SMS-200 and pay the $900 to upgrade to the Ultra with master clock.

 

Their crazy pricing has put me off buying the other items I had planned to complete the Romaz chain.  I could have easily spent another £4K or £5K on their products including the MC when it is released. 

 

I suggest you contact them directly and negotiate what you think is fair. 

 

Just to be clear - no one - not even Roy - has reported on the Ultra chain with SOtM's, or anyone else's, master clock. We just don't know if, or how much better, this will further improve SQ. The $200 MC input option is a choice, for those who want to be ready for this upgrade. 

 

Speaking only for myself, my pockets aren't infinitely deep, and I've already over-spent on the streaming transport portion of my end-to-end chain. The next $$$ I spend should probably go towards other parts of my chain to maintain system balance.

 

I'm sitting out the whole MC upgrade path until others have reported on the improvement, and I can weigh the cost/benefit.

[austinpop]

2 minutes ago, Bamber said:

I could purchase a ISO-Regen with LPS-1 bundle and have it delivered to your address if Superdad allows that.

 

I could transfer enough funds to you to cover shipping to me in UK once you are finished testing.  That way you could keep it for 3 or 4 weeks.

 

You could... but will you?  

 

But seriously - that would be awesome! We can discuss over PM.

[austinpop]

18 minutes ago, AmusedToD said:

That's why I am not rushing to get the modded switch. 

 

Just to be clear - I use the modded switch in the direct connection (i.e. bridged server) path. So the two - modded switch and direct connection - are not mutually exclusive.

 

Mine looks like this:

 

Cable modem <> Router <> (PCIe Eth) Win 10 Server Bridged (USB Eth) <> modded switch <> modded sMS-200 <> tX-USBultra <> DAC.

 

Hope that's clear.

 

 

[austinpop]

2 minutes ago, ted_b said:

If you are going through a switch how is that "direct" in your mind?  I assumed when we talk direct we really mean nonstop, and assumed the sq advantages are akin to a nonstop flight, i.e. Faster, no changing planes and no risk of schedule conflicts

 

Hi Ted,

 

I'd refer you back to Roy's first posts describing the direct connection. The "direct" comes from bridging the server, so the endpoint is isolated from the router, rather than connected to the same switch.

 

Adding the modded switch between the server and the endpoint still maintains that isolation from the router. The modded switch in this path, due to its clock mods, linear regulators, clean LPS (I use the LPS-1) is really acting as an Ethernet reclocker.

 

 

[austinpop]

9 minutes ago, tboooe said:

Two questions Rajiv...

 

1.  I assume you are running some sort of dhcp server on your win10 pc?  As I understand it the sms200 does not have any settings to allow a fixed ip.

 

2.  Or does your switch have a dhcp server built in?

 

No Tommy - no DHCP server is necessary.

 

When you bridge two (or more) adapters in Windows 10, 2 things happen:

1. a new "Network Bridge"adapter gets created. Within it's properties, you then assign it a static IP (although dynamic IP also works)
2. the original adapter objects show as "bridged," and no longer allow you to set IP properties within them.

Under the covers, WIndows is providing a virtual switch (layer 2) between the two adapters. The sMS-200 continues to be in the same subnet as the server, and everything upstream of it in the network.

[austinpop]

Just now, austinpop said:

 

No Tommy - no DHCP server is necessary.

 

When you bridge two (or more) adapters in Windows 10, 2 things happen:

1. a new "Network Bridge"adapter gets created. Within it's properties, you then assign it a static IP (although dynamic IP also works)
2. the original adapter objects show as "bridged," and no longer allow you to set IP properties within them.

Under the covers, WIndows is providing a virtual switch (layer 2) between the two adapters. The sMS-200 continues to be in the same subnet as the server, and everything upstream of it in the network.

 

I should add:

• the sMS-200 still gets its DHCP address from the DHCP server on the subnet - typically running in your router
• Like many others, to get the network bridging to work successfully in W10, I had to bridge 3 adapters: 
  • PCIe Ethernet
  • USB Ethernet
  • PCIe Wifi, enabled, but not connected to wifi
• For some reason, the presence of the 3rd (wifi) adapter in the bridge was necessary to get it all to work. Others - many of them - found the same to be necessary.

[austinpop]

14 minutes ago, AmusedToD said:

 

Yes, I understand that, but I would call this connection just bridged (over a switch) as opposed to direct connection with nothing in between the PC and the streamer.

 

Apart from modding a switch, I would need to mod my sms200Ultra and get a separate linear PSU for the switch. So that's probably in the $700-800 zone, and I am not prepared to go that far.

 

Instead I decided to invest in a proper DSP software (Acourate) and a calibration mic, and get the most out of my Roon based system within the constraints of my listening room. 

 

Fair enough!

 

Make your own choices - after all, only you have to live with them.

[austinpop]

Here is @romaz's first report on what he called the "triple combo," and what I've been calling the Ultra trifecta.

 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-the-sms-200-and-microrendu/?do=findComment&comment=633020

 

With regard to placement of the switch, note he says (emphasis mine):

 

Quote

With my NAA in the standard configuration (connected directly to a router), this reclocking switch made a small difference but barely worth the $100 I spent for it even when powered by an LPS-1. With either NAA directly connected to my Mac Mini and with this reclocking switch in the "direct" path, however, I was quite surprised by the significant improvement in terms of soundstage and clarity.

 

[austinpop]

6 hours ago, tedwoods said:

First off, major props to @romaz , @austinpop and the many contributors here is in order.

Thank you guys for your relentless efforts: the direct connection, especially with switch re-clocking is indisputably better sounding.

To my mind, only one thing remains, that desperately needs to be addressed by audio manufacturers:

The introduction of a stand alone audio ethernet re-clocker with built in filtering!

 

1 hour ago, greenleo said:

JCAT mentioned that they're going to introduce one.  But no sign yet.

 

2 minutes ago, lmitche said:

Isn't the reclocked switch from SOTM one of these?  Kinda like a REGEN for Ethernet?

 

Having personally experienced the improvement wrought by this Ethernet regeneration, the next question is, indeed, where do manufacturers go from here?

 

@JohnSwenson, who is consistently honest and candid in these forums, has stated that this is an area he is interested in, and the science of what we're hearing is not well understood. I hope I paraphrased his views correctly.

 

Even if the science behind these improvements on the Ethernet is not yet baked, several questions arise going forward:

1. what makes the SOtM sCLK-EX so special, as to surpass previous reclocked TCXO/OCXO switch offerings like the Pang, the AQvox, etc, that Roy had heard?
2. how will SOtM modularize their offering? Right now, they have a single clock board - sCLK-EX - with 4 taps that they distribute both internally and externally. This approach is inherently limited by the length of the clock wire. What would make sense is to figure out a way to embed a fit-for-purpose sCLK-EX in each component. The current Ultra approach is wasteful and expensive - i.e. embed a full-size sCLK-EX board in each Ultra component.
3. how will other manufacturers respond? And will we see new designs that compete with, and even exceed the SOtM solution, at equal or lower prices? As consumers, I sure hope so!

As I often like to say - it's an exciting time to be a computer audiophile!

[austinpop]

15 minutes ago, lmitche said:

Isn't it the fact that the four clocks on the SOTM clock board share the same reference?

 

Yes they do.

 

[austinpop]

23 hours ago, Bamber said:

I could purchase a ISO-Regen with LPS-1 bundle and have it delivered to your address if Superdad allows that.

 

I could transfer enough funds to you to cover shipping to me in UK once you are finished testing.  That way you could keep it for 3 or 4 weeks.

 

I gather Alex has blessed this enterprise, and an order has been suitably placed in the system. Since the ISO-R's are sold out until July, now we hurry up and wait.

 

But all that aside, many thanks to @Bamber for his generous loan. I hope to make the best of it and am eager to report my findings here.

[austinpop]

1 minute ago, limniscate said:

@austinpop

 

Does this mean I'll get to hear the ISO-Regen in my system as well?

 

I'm counting on it! That will give us a perspective on two completely different systems. Hopefully by then your Ultra trifecta will have arrived too.

[austinpop]

1 hour ago, JohnSwenson said:

I have not measured any of these clocks so I can't comment on their relative performance. But I can give some general observations.

 

First a clock in a real system can be very different than a clock being measured by itself. The phase noise of ALL clocks will change depending on noise on the power and ground network feeding the clock. So what is happening with other circuitry around a clock can significantly degrade its output from what you see in a spec sheet.

 

The lowest phase noise clocks are OCXOs, but not all OCXOs have very low phase noise. There are quite a few inexpensive OCXOs (less than say $150) that have worse phase noise than a 575.

 

The result is that just because two devices have OCXOs, does not mean they have identical jitter performance.

 

John S.

 

13 minutes ago, JohnSwenson said:

OCXO stands for Oven Controlled Crystal Oscillator.

 

The hierarchy of crystal Oscillators:

 

XO  Crystal Oscilator

TCXO Temperature Compensated Crystal Oscilator

OCXO Oven Controlled Crystal Oscillator

 

What all this has to do with is how the frequency of an oscillator behaves as the temperature of the crystal varies. All crystals will change their frequency due to temperature changes. If you map these changes you get a curve that at low temperatures has a large change per small temperature delta, and at high temperatures a large change per small temperature delta, but at some intermediate temperature the "temperature coefficient" (TEMPCO) is zero.

 

If the operating temperature inside the equipment is not at this point, changes in temperature will change the frequency. The TCXO and OCXO are ways to make this better than a normal "uncompensated" crystal.

 

Note that these frequency changes are long term changes, sometimes referred to as drift. They are usually NOT the short term variations we call jitter. Thus optimizing for TEMPCO is not necessarily a good thing for audio.

 

The TCXO is an electronic circuit added to the oscillator. It turns out that you can change the frequency of an XO by changing the capacitance across it. The TCXO contains some form of temperature sensor that feeds a device that changes the capacitance across the crystal, the change in capacitance cancels out the change in temperature thus significantly lowering the TEMPCO, . But in doing so it increases the jitter, that temperature sensor and cpacitance changing circuit generates noise. This increases the jitter of the crystal. Thus TCXOs are usually not a good thing in digital audio, they are optimizing the wrong thing.

 

The operation of the OCXO is usually described as maintaining the temperature of the crystal so that the frequency doesn't depend on what is happening outside, the oven keeps the temperature stable. But this is not the whole story. The REAL purpose of the OCXO is to maintain the crystal at that temperature where the TEMPCO is zero, not just any old temperature. If the crystal is at the zero TEMPCO point then the temperature stability doesn't matter that much.

 

There is a big problem with this. The normal way a chunk of quartz is cut is called the AT cut, this has a zero TEMPCO temperature right around room temperature. This is fortuitous for non compensated crystals, but a real pain in the neck for OCXOs. If the crystal temperature is too high the system has to COOL the crystal, if too low heat it. This is quite complex and very expensive to do well.

 

The solution is to use a crystal cut called the SC cut. This has a zero TEMPCO point much higher than room temperature, so all you need is a heater, which is really easy to do (current through a resistor). This is why it is called an oven, it is always heating.

 

A side effect of the SC cut is that IF it is done right you get lower phase noise. But the inexpensive OCXOs put all the money into making the oven and skimp on the crystal quality, so they don't get particularly great phase noise. The ones that DO put money into the crystal can have the lowest phase noise of any oscillator, period.

 

So why doesn't anybody use an SC cut crystal without the oven? At room temperature the TEMPCO is SO far from the zero point that it would be almost useless.

 

I hope this gives a little insight into the world of crystal oscillators.

 

John S.

 

 

 

 

 

Thanks for these insights!

 

I cannot tell from the sCLK-EX Press release or its product page any mention of TCXO or OCXO, so one has to presume the claims come from the overall board design, not the actual crystal oscillator.

 

@JohnSwenson - where does the oscillator used in the ISO-Regen (the Crystek 575?) fit into the XO/TCXO/OCXO taxonomy?

[austinpop]

3 hours ago, AmusedToD said:

Just a quick update on my side.

 

I emailed May from SOTM and asked her whether it's possible for them to modify the clock of an Intel NUC LAN controller so that it can use the superior clock inside the sms200Ultra and they said yes!

 

The idea is to use the Intel NUC with the modded LAN clock as Roon server (probably with the Roon OS), and then connect it directly via Ethernet to my sms200ultra. This way a switch in between can be avoided.

 

 

This is a very interesting approach, and I look forward to your findings. Let me make some comments and ask you for some clarifications.

1. How long do you expect you'll need the clock cable from the sMS-200ultra to the LAN controller? You should confirm this with May, but the SQ degrades the longer this cable is. My clock cables (hardwired, before SMB was offered) are of the order of 30cm. So if you're looking at much longer lengths, keep this in mind.
2. The approach Roy advocated, and I followed and verified, is to use all 4 clock taps on the sCLK-EX board. The sMS consumes 2 clocks, and you've talked about using another for the LAN controller. What are your plans for the 4th clock tap?
3. If that 4th tap is free, I do recommend you consider using it for one more reclocking. Candidates would be either a modified switch between the NUC and the sMS, or modify a tX-USB-HD on the USB chain between the sMS and the DAC.

I do understand that exploiting each additional clock tap with a modified component does add cost.

[austinpop]

2 hours ago, Johnseye said:

 

That's what Roy is doing, although he's taking it a little further than that.  We're all waiting to see if it makes a substantial improvement.

 

I know several folks here are in the process of getting an Ultra stack of various flavors, so it is going to be really exciting to hear their impressions. 

 

I do want to distinguish between 2 categories of stacks:

1. a stack that utilizes all 4 clock taps from an sCLK-EX board, whether housed in an sMS-200ultra, a tX-USBultra, or a dX-USB HD Ultra
2. a stack that includes using clock taps from more than one sCLK-EX boards, specially upstream of the sMS or the switch preceding it.

Last I heard, Roy was pursuing #2 above, which gets into mods within the music server's internal clock, or even further upstream in network gear like routers and modems.

 

It's an exciting time.... ah, never mind!