- Part 1 - Introduction and Digital Audio Optimization Foundations (Link)
- Part 2 - Enter the Extreme (Link)
- Part 3 - First Impressions and Basic Configuration (Link)
- Part 4 - Tweaking Up the Extreme
- Part 5 - Extended Listening Impressions, Learnings, and Conclusions (Link)
In the first three parts of this review of the Taiko Audio SGM Extreme (linked above), I looked at the Extreme out of the box, and the remarkable performance it is capable of. In this Part 4 of my review, my focus shifts to if and how to make the Extreme better by tweaking factors outside the Extreme itself (networking, power, etc).
So Can We Get the Extreme from Ridiculous Speed to Ludicrous Speed to Plaid?
Emile has often said that the only attention you need to give your Extreme is good power, good network, and good mechanical isolation.
Taiko recommends that the Extreme be on an electrical circuit that can draw 2kW of power (in the US, this would mean the Extreme has a dedicated 15A or preferably 20A circuit that it is not sharing with other devices), with good quality outlets (rhodium, silver, or gold connectors) with a high quality power cord. Although the Extreme itself only requires a small to modest amount of power when running (as little as 60W), it greatly benefits from copious amounts of transient current (the faster the wall can respond to current demands from the Extreme, the better Extreme can deliver and regulate power to its internal systems). It will work great on a standard 15A shared circuit (that’s how I was running it out of the box), but it just does better if there is more transient power available to it when it wants it. Lastly, Taiko also recommends not having any powerline networking polluting your electrical system with added noise.
For networking, Taiko recommends a low RFI network connection (impact will vary based on your local setup and circumstances). The Extreme comes with a StarTech optical network card, and Taiko recommends that those looking for the ultimate performance explore whether an optical network connection works better than copper ethernet for them.
Lastly, the Extreme is a beast, and requires a stable platform. Taiko recommends that it be on a stable shelf or platform as close to the floor as possible to minimize adverse impacts from mechanical vibrations.
So if we assume that cpu/motherboard/memory, storage, internal power, clock, and OS/software have already been optimized to the Nth degree and can’t be improved on, my next step was to assess the impact of optimizations of:
- Power - Assuming the internal power supply can’t be easily improved upon, how much lift is to be had moving to a dedicated 30A circuit, isolation transformer, and with a premium power cord?
- Network - What optical network topology is optimal for streaming, and what kind of lift over stock copper ethernet?
- USB - How do different USB cables impact sound quality?
- Mechanical - How does sound improve as mechanical isolation improves?
For these tests, I stepped up the tweak ladder one OCD step at a time, then pulled it all together when I moved my Extreme to its final permanent home.
As a reminder, my tests so far of the Extreme have been with a generic power cord plugged into a generic power strip (with all my other AV equipment and various SMPS plugged into it) which was in turn plugged into a generic outlet on a shared 15A circuit behind my media console.
The performance of the Extreme with this level of (lack of) attention to power hygiene is remarkable, and a testament to the quality of the design of the Extreme power network and casing. That being said, can we make it better by providing the Extreme better cleaner power and more power?
Starting at the electrical panel, I evaluated the following upgrades:
- A dedicated 30A circuit (10-2 Romex) connected to an upgraded high conductivity outlet, dedicated only to the Extreme. The line length from my electrical panel to this outlet is ~15', so it is in extremely close proximity to my home electrical panel.
- A dedicated Topaz Ultra Isolation Transformer (model 91018-31T). This is a 1.8KVA model, with ultra low capacitance. I have my isolation transformer wired for balanced power output.
- A variety of upgraded power cords, ranging from a Pangea AC-14SE to an Elrod Power Systems EPS-3 Signature to a Sablon Audio Prince power cord with the beast of a Bocchino IEC connector on the Extreme end.
- (PENDING) A Synergistic Research Orange fuse (still trying to source one..will update when I do)
All other components (network, USB, mechanical) were left at their generic baseline. Playback was local content in Roon with HQPlayer (JPLAY ASIO driver) in a one box solution.
Impact of Dedicated Circuit
Going from a shared circuit to a dedicated 30A circuit, even with the generic power cord, the increase in speed and dynamics is absolutely stunning (wow!) but with perhaps a touch of edge or roughness.
When I first put in my dedicated circuits, it was dollar for dollar one of the best system upgrades I had implemented. That being said, the impact of a dedicated circuit on the Extreme is well beyond what I expected. Backgrounds are much blacker, and the physicality and control of dynamics are kicked WAY up. Details are more vivid and tangible, and depth resolution and spatial placement even more refined and extended. Everything that I was hearing before is simply better, faster, more controlled, and more coherent. Fantastic!
I have two dedicated circuits (one for my DAC and analog components, one for my digital chain) for a reason. Consolidating my entire digital chain into a single box with the Extreme alone on a dedicated circuit is an absolutely wonderful place to be. Does power at the wall matter for the Extreme? Most definitely, most emphatically yes!
Impact of Isolation Transformer
Next I plugged my Topaz Isolation Transformer into my dedicated circuit, and plugged the Extreme (still with generic power cord) into the output of the isolation transformer.
In my NUC reference system, I’ve found that the Topaz gives significantly darker backgrounds, and especially when run balanced, brings a sense of control and stillness between the notes that engages me immediately and consistently (that classic “midnight music listening” feeling, but any time of the day).
Listening to the Extreme with the Topaz, playback certainly felt smoother (aligned with what I’m used to) and that slight edge and harshness I was hearing was tempered, but that incredible speed and dynamics that I was hearing with the Extreme was significantly diminished.
As much as I love that “midnight music listening” feeling, I love the speed and dynamics of the Extreme even more. While the Topaz isolation transformer was certainly cleaning up some nasties, it was slowing things down too much, and simply was not able to keep up with the Extreme.
As with all things power, every home situation and neighborhood situation is different. The choice between a power conditioner or power regenerator or large battery or isolation transformer or any combination of these will heavily depend on your home electrical setup, and what is going on in the electrical grid in your neighborhood. Experiment and see what works for you, but keep in mind that the better and faster and more power you deliver to the Extreme, the happier it is (think of it as a very large monoblock amplifier, not a computer).
Impact of Upgraded Power Cords
Still going directly to a dedicated circuit outlet (no isolation transformer...my digital audio Topaz is now officially retired), next I auditioned a series of upgraded power cords that I had available to me. I just a couple days before this listening test had received one of the traveling demo Sablon Audio Prince power cords (well broken in and very well traveled exterior) as a loaner, so I was excited to hear what it could do in my system.
Going from the generic power cord to a Pangea AC-14SE, I immediately noticed a smoother presentation, with a slight(?) apparent loss in dynamics or speed (difficult to decouple some of the changes in my listening tests). There were significantly better dynamics and speed than I had heard with the Topaz isolation transformer. Certainly a very relaxed presentation, and a step forward from the generic power cord.
With the Elrod EPS-3 Signature, there was a significant increase in speed and dynamics, with even better smoothness. The refinement of presentation and elegance reminded me of what I’ve heard in the finest analog systems, but with all the holographic engagement and transparency that I’ve come to love in my digital system. This was starting to get very very interesting indeed!
The Sablon Prince is an interesting beast, with the HUGE Bocchino IEC connector on the Extreme end of the chain. Although it is labeled as “King” on my cable, the Prince has the Boccino connector on the IEC end and a normal connector on the plug end. The Sablon Audio King power cord has a Bocchino on both ends of the cable, but Sablon does not provide it for demos because of the impracticality of the monster connector on a wall outlet or power conditioner.
Going from the Elrod to the Sablon Prince, the increase in speed and dynamics was immediate and striking. While still buttery smooth and elegant, speed and coherence made a massive step forward. My first 10 second reaction listening to the Sablon Prince is very similar to when I first heard the 2020 Sablon USB cable in my system: things just clicked and it sounded perfectly neutral and transparent, with incredibly coherence and dynamics, invisible and just right – “Yup, we’re done testing here...let’s enjoy some music”.
Going back from the Prince to my generic power cord, the added edge with the generic cord is very clear, and the dynamics are quite a bit muted vs the Prince.
Comparing the Prince direct to the dedicated circuit vs the Prince with the Topaz Isolation Transformer, I’m not hearing any improvement in background levels or smoothness, but significantly diminished speed and dynamics with the isolation transformer. Whatever conditioning benefit I’m getting from the Topaz seems to not matter with the Sablon in the chain (the Prince is taking care of business), but the Sablon seems to be able to provide more of the current that the Extreme craves.
If the best cable is the cable that sounds like there is nothing there, the Sablon Prince is the closest I’m hearing to that with the Extreme and the cables I currently have available to me to test. Now the only decision is whether I purchase a Prince, or take a leap for the King without having had a chance to hear it!
(NOTE: The Sablon Prince power cable is a loaner unit kindly provided by Mark Coles of Sablon Audio. I have no affiliation or association with Sablon Audio or Mark other than as a very happy owner of Mark’s 2020 USB cable. I’m under no obligations or expectations associated with the loaner unit, except to give it a listen, and pass it on to the next audiophile in the queue. My thanks to Mark for making this gem of a cable available to me!)
Impact of Upgraded Fuse (PLACEHOLDER SECTION)
The Extreme uses a single Schurter 4A fuse. Given the positive impact that other power tweaks have had, is there an upside to using something like the Synergistic Research Orange fuse? As I am able to source a Synergistic Research Orange fuse to audition, I will update this review.
Again a reminder, my tests so far of the Extreme have been with a generic Cat 5 ethernet cable connected to a shared SMPS-powered generic switch in my media console, connected over a length of generic Cat 5 home ethernet to an Ubiquiti EdgeRouter 10X home router connected to an AT&T-provided optical network terminal and Pace gateway/router for my fiber internet connection.
As I noted above, in this configuration there was a notable gap between content played from local storage on the Extreme and content from on my home Mac Mini file server. There was also a much bigger step down to content streamed from TIDAL.
Can we make these gaps smaller or even go away by providing the Extreme with better copper networking, switching to optical networking, and/or a more direct path to my fiber internet connection?
Starting at the back panel of the Extreme, I evaluated the following network upgrades:
- A variety of upgraded ethernet cables, from the surprisingly good Cable Matters SFTP Cat 8 cable and the even better Monoprice Cat 8 cables to the brand new Sablon Audio 2020 Panatela Reserva ethernet cable that arrived just in time for this review
- An Uptone Audio EtherREGEN network switch with Uptone Audio UltraCap LPS 1.2 power supply, with the Extreme on the recommended “B” side of the network isolation moat
- An EtherREGEN with a Planet Technology MGB-TLX SFP fiber transceiver on the SFP port on the “A” side of the EtherREGEN, connected to the StarTech PEX1000SFP2 PCIe optical network card in the Extreme (with a matching Planet Tech optical transceiver)
- A Sonore opticalModule fiber media converter (FMC) powered by an Uptone Audio LPS 1.2 power supply, with two Planet Technology transceivers, converting wired ethernet to optical ethernet connected to the Extreme’s StarTech optical network card
- A Sonore opticalModule connected all the way back to a Ubiquiti EdgeRouter 10X gateway/router (bypassing all the intermediate home ethernet wiring and switches) with a long optical fiber run to the Extreme
- Various power and topology optimizations for the AT&T provided Pace 5268ac router/gateway and Alcatel-Lucent G-010G-Al optical network terminal (ONT) for my optical-fiber-based Internet connection
For these tests, I left the Extreme plugged directly into my dedicated 30A circuit with the Sablon Audio Prince (I can’t give that up now that I’ve heard it), but with the original generic USB and mechanical baseline setup. Playback was local content, file server content on my Mac Mini, and TIDAL content, all played through Roon with HQPlayer (JPLAY ASIO driver) in a one box solution.
For the component level tests, I limited my testing to TIDAL streaming content only to force music to stream from the Internet through all the intermediate network connections and devices. When I got to an optimal configuration for streaming, I came back to compare streaming to file server content to local content with an optimized network.
Impact of Ethernet Cables
When I first received my Uptone Audio EtherREGEN, I was stunned by the sound quality impact network connectivity could have. As surprised as I was by the impact the EtherREGEN was having, I was shocked at how incredibly audible differences between ethernet cables were. I have no idea how and it is a difficult conversation to have even with dedicated audiophiles, but for me, with the EtherREGEN I could tell apart ethernet cables with just a couple seconds of listening. Incredible, and an experience that forced me to reconsider a lot of things that I had taken as givens.
In my experiments with the EtherREGEN, with better shielding and higher bandwidth cables (Cat 8), timing detail, coherence, and sense of space were all significantly improved. If you’ve stood next to a live musician as they are playing a bass or a low string instrument, you know the whole body feeling you get from the strings and soundboard. With the EtherREGEN and better (not more expensive, but electrically better) cables, that whole body feeling was very real, and added significantly to the sense of being in the presence of the musicians.
As an added benefit, on some of my best recordings, the low space resonances from the room where the recording was made all snapped into focus for me, giving me a very strong sense of the space where the recording was made. That feeling of being in the cathedral or the club or concert hall is awesome. The icing on the cake was that I was able to hear all this with the EtherREGEN and inexpensive but quality Cat 8 ethernet cables ($10-15 range).
Even without any attention to network or cables, I was hearing hints of that same whole body feeling with the Extreme (first time hearing it without the EtherREGEN). Using my generic media console network switch, as I went from generic Cat 5 ethernet cable to a Cable Matters and Monoprice Cat 8 cable, I heard a very slight improvement in bass detail going from generic Cat 5 to Cable Matters and Monoprice Cat 8.
However, even in this configuration, things significantly opened up and became more natural sounding with the Sablon ethernet cable. This was a surprise, given the cruft before the cable., but even in this configuration cables do have an audible impact with the Extreme (a very non-subtle and welcome impact with the Sablon 2020 cable). Bass control and sense of space definitely has moved to the positive with the Sablon.
Given how small the impact was in this configuration going from generic Cat 5 to the Cat 8 cables, I was quite surprised at the impact of the Sablon ethernet cable, even with all the other ethernet cables in the chain. Could this be a hint as to how to get the most out of the Extreme with copper networking?
(NOTE: The Sablon 2020 Ethernet cable is a loaner unit kindly provided by Mark Coles at Sablon Audio. I have no affiliation with Sablon Audio except as a customer, with no expectations or obligations for the loan of the cable except to give it an audition in my system.)
Impact of EtherREGEN for Copper Ethernet Regeneration
In this test, I introduced an Uptone Audio EtherREGEN to the chain. The EtherREGEN (see Rajiv’s comprehensive review here) is an audiophile network switch that provides both ethernet signal regeneration, and a signal isolation “moat” between it’s “A” side and a single RJ45 100mbps network port on its “B” side. For these tests, I powered the EtherREGEN with an Uptone Audio LPS 1.2 power supply set to 12V.
My experience with the EtherREGEN with my reference NUC setup was outstanding. Does it add anything to the Extreme, even though the network speed going to the “B” clean side is reduced to 100 megabit ethernet?
To test, I compared the Extreme connected to the “B” clean side of the EtherREGEN (with a premium ethernet cable) and the EtherREGEN connected to a generic switch from the “A” dirty side with a generic Cat 5 cable, with the Extreme connected to the generic switch directly with a premium ethernet cable.
The answer is a most definite yes. All the attributes that I heard before with the EtherREGEN are back - increased speed and control (especially for bass), more subtlety and detail, awesome openness and sense of space.
Repeating my ethernet cable tests, but this time with the EtherREGEN in the chain, I was able to hear how really special the Sablon 2020 ethernet cable is. Just lovely - neutral, transparent, coherent, fast, and relaxed presentation. Quite a stunning improvement over the Monoprice Cat 8 cable, which was my prefered cable with the EtherREGEN in my NUC-based chain. The music and the room around the musicians just draws you in. Like my “first 10 seconds” reaction when hearing the Sablon 2020 USB cable and Sablon Prince power cord, the Sablon 2020 ethernet cable comes across as neutral, transparent, coherent, invisible and just right – “Yup, we’re done testing here...let’s enjoy some music”. Mark is really on a roll, and I’m loving everything I’m hearing from him.
So what is going on with the Sablon ethernet cable that is resulting in this striking level of impact? The Sablon cable is quite impressive, with Cat 8.1 Telegärtner connectors, robust shielding, and Mark’s favorite conductors. These connectors are distinguished by extremely high bandwidth (bandwidth seems to be emerging as a key measure for ethernet cables and audio). Mark uses his favorite conductors for the body of the cable itself, with 5 layers of shielding. Is the extreme bandwidth + shielding + conductor quality what is creating this synergy with the Extreme? Clearly, the Sablon ethernet cable is keeping up with the Extreme much better than the other Cat 8 cables I have on hand. Even for cabling, speed seems to make a big impact with the Extreme.
Hearing what the EtherREGEN is able to do with the Extreme, I’m reminded again of what an outstanding value the EtherREGEN is for wired ethernet (just lovely...I had to pull myself away from the combination of Extreme + EtherREGEN + Sablon 2020 ethernet cable to continue with testing). How will moving to optical ethernet change things?
Impact of Using EtherREGEN as a Fiber Media Converter
Note that Emile did help to configure my optical network interface with some experimental tweaked parameters and driver configurations. The tweaks included configuring the optical network connection for Microsoft Client Network, File and Print Sharing, and IPv4 only. Also, disabling most of the advanced features in the Realtek PCIe GBE driver, and tweaking some buffer sizes to what Emile believes sounds better with “thin” sounding networks. All my listening tests with optical networking were from this baseline, and I did not experiment with optimizing any of these parameters myself (even my audio nervosa has its limits). I made no changes to the copper networking interfaces, which were already tuned from the factory.
The Extreme is my first exposure to optical ethernet. I have held off doing experiments with optical ethernet as I’ve been waiting for the Extreme, knowing that the optical network interface is generally the preferred interface on the Extreme. For these tests, I picked up a pair of the Planet Technology MGB-TLX optical transceivers that Emile had recommended from his tests. I also used the Sablon 2020 ethernet cable from my generic media console switch to the EtherREGEN.
The EtherREGEN does have an SFP port for an optical transceiver on the “A” “dirty” side. This allows optical in on the “A” side, and copper out of the “A” side or “B” “clean” side to the Extreme. It also allows copper into the “A” or “B” side of the EtherREGEN, and optical out to the Extreme from the “A” side (the EtherREGEN is symmetric, and can be operated in either direction). The later configurations are what I evaluated, basically using the EtherREGEN as a Fiber Media Converter (FMC), but with a moat (with the 100mbps ethernet limit from going across the moat) and without a moat (with the full 1gbps ethernet on the “A” side).
For the first test, I wanted to compare wired ethernet to optical ethernet on the Extreme. For this test, I used a generic ethernet cable to connect the “A” side of the EtherREGEN to my generic media console network switch. I then I compared wired ethernet from the “B” side of the EtherREGEN to the Extreme with the Sablon 2020 ethernet cable (the optimum configuration from the previous section), to optical ethernet from “A” side of the EtherREGEN to the optical interface of the Extreme (no moat but fast ethernet for optical, vs moat but slower ethernet for wired).
With the optical input to the Extreme, detail and control and dynamics all stepped up significantly. The sense of space and tangibility was improved as well. No question, even without the isolation moat of the EtherREGEN, the optical network path with the EtherREGEN is a huge step up from wired ethernet.
But how much of that is due to the slower ethernet speeds when you step across the EtherREGEN moat, and how much to the optical path directly? To test that, I rewired the network to use the Sablon 2020 ethernet coming from my generic media console switch to the EtherREGEN, with optical going from the “A” side of the EtherREGEN to the Extreme.
By connecting the Sablon ethernet cable to the “B” side of EtherREGEN, I would get the advantage of the moat but with slower ethernet speeds. By connecting the Sablon ethernet cable to the “A” side of the EtherREGEN, I would get fast ethernet speeds but no moat. In both cases, the final leg would be optical to the Extreme (moat but slower ethernet for optical, vs no moat but faster ethernet for optical).
In these configurations, optical with the moat is better than wired, but it is still a definite step down from what I’m hearing with both optical and wired on the same side of the EtherREGEN (no moat).
For me, optical ethernet is clearly the winner, but so is fast ethernet. Whatever the EtherREGEN isolation moat is bringing to table isn’t making up for what is being lost with the step down in ethernet speeds. To my ear, the EtherREGEN moat just can’t keep up with the Extreme. Once again, speed seems to really matter with the Extreme.
Impact of Using a Sonore opticalModule as a Fiber Media Converter
During my tests with the EtherREGEN and optical, my unit was running quite hot (presumably from the extra power draw of the optical transceiver?). While Uptone notes that the unit is designed to run hot, it did get me thinking that by not using the signature moat feature of the EtherREGEN, I was basically using the EtherREGEN as a high quality John Swenson designed audiophile fiber media converter.
The other John Swenson designed FMC on the market is the Sonore opticalModule – wired ethernet in on one side, optical transceiver and optical output on the other. So how does the EtherREGEN with optical and wired ethernet on the “A” side compare with the opticalModule as a FMC? I compared them with each powered by an Uptone Audio LPS 1.2 power supply (12V for the EtherREGEN, 5V for the opticalModule), with the Sablon Audio 2020 ethernet cable going to my generic media console network switch.
While differences were less obvious on initial listen, the opticalModule is clearly and consistently better sound quality to my ear. The longer I listened to each configuration, the more clear and compelling the differences became. With the opticalModule, there are definitely more dynamics and control, better and more subtle sense of detail and space, and most importantly for me, a more relaxed and engaging presentation.
Reviewing my notes from when I first received my EtherREGEN and went through power supply and ethernet cable tests, the character of the differences I hear between the EtherREGEN and the opticalModule are very similar to what I experienced with the EtherREGEN as I moved from the stock EtherREGEN power supply to a Uptone Audio LPS 1.2 to a Paul Hynes SR4. As I improved power to the EtherREGEN, the EtherREGEN just opened up more and there was more goodness there. Going from the EtherREGEN to opticalModule is just more goodness in that same direction and of that same type.
My suspicion is that the opticalModule may be straining the LPS 1.2 less than the EtherREGEN is straining the LPS 1.2 (in the past, I’ve found that sound quality can be adversely impacted as the LPS 1.2 is driven closer to its current limits...it’s a wonder supply for low power devices, but has its limits as pushed). With the opticalModule running cool and effortlessly and beautifully I did not chase this further, but EtherREGEN owners may be well served to seek out high quality higher current capability power supplies if they are running the optical networking with the EtherREGEN.
Impact of Network Optimizations All the Way Back to My ISP Connection
The Extreme definitely rewards the listener who pays careful attention to network optimization – give the Extreme the cleanest fastest network bits you can, and it just opens up more of what the Extreme is capable of.
All my tests to this point have had a lot of other cruft between my fiber link to the street and the Extreme. What happens if you eliminate that cruft and take network optimization all the way back to your connection to your internet service provider?
As background I have AT&T fiber optic internet service to my home (1gb up and down). I have fiber optic coming in from the street to an AT&T provided Alcatel G-010G-A optical network terminal (or ONT, basically a fancy fiber media converter that also handles video signals and voice signals). That ONT is connected by ethernet to an AT&T provided Pace 5268ac router/gateway (WiFi, VoIP, TV, Ethernet). I only have data to my home (no voice or traditional cable), so the only thing I connect to the AT&T Pace gateway/router is my Ubiquiti EdgeRouter 10X. On my EdgeRouter, I have multiple VLANs (guest, IoT, home, etc), my WiFi access points, and all my wired ethernet for my home.
So what happens if I move my opticalModule to my closet and connect it directly to my EdgeRouter? The Planet Technology optical transceivers I’m using are rated to 12 miles(!). The long fiber run from my listening room to my network closet is considerably less than that.
With the intermediate Cat 5 home network and generic media console switch out of the path, sound quality took a significant step up – more openness, more control and detail, just more of the network goodness I’ve been hearing each step of the way.
How about getting rid of the AT&T router/gateway and connecting directly to the ONT? Alas, AT&T requires that you use their gateway and does not provide any bypass mode (yes that really sucks). Thankfully, the authentication mechanism AT&T uses for their router/gateway isn’t terribly sophisticated.
Subscription is managed at the ONT (you can’t easily steal service), but if you clone the MAC address of the AT&T provided gateway to your own router, you can swap in your own router after the AT&T router/gateway has authenticated against the AT&T network. For the more adventurous amongst us, it is also possible to extract the security certs from the AT&T router/gateway and install them on your own router with a WPA supplicant. I would never endorse such a thing (ahem), but it is an opportunity to see if eliminating another block in the network chain has any impact.
Hypothetically speaking, if one were to eliminate the AT&T provided gateway, yup another notable step up in sound quality, with more of the same goodness. What I’m (hypothetically) hearing is stunning, by far the most relaxed and natural and extended soundstage and dynamics I’ve ever heard.
Swapping out the SMPS on the EdgeRouter for a LDPWR DXP-1A5DSC dual stage power supply at 12V, energized by a PowerAdd Pilot Pro2 battery? Wow, quite a bit better still. Swapping out the generic ethernet cable between the ONT and the EdgeRouter with the Monoprice Cat8? Yup, even better. Swapping out the SMPS on the ONT for a LDPWR DXP-1A5S single stage power supply at 12V, energized by a PowerAdd Pilot Pro2 battery? Another material step up (this one really surprised me...time to hack into the AT&T fiber concentrator down the street and finagle a power supply upgrade?).
As a final bonus tweak, how about the router noise filtering dongle thingie that Mark Coles sent along with the Sablon 2020 Ethernet cable to demo? Yup, even better. Dang, this rabbit hole seems to have no end to it.
I should note that the EdgeRouter is a fine gateway/router (the price/performance/feature set is pretty amazing actually) but the real bonus is that unlike the AT&T provided Pace gateway/router, it can be easily powered by a modest high quality power supply like the LDPWR’s or a LPS 1.2. Clean power to the AT&T provided ONT and the EdgeRouter 10X made a startling difference in sound quality. I was anticipating a small change for the better, but it was a heck of a lot more than that.
Aside from telling me I’m going to need two Sablon 2020 ethernet cables (ONT to EdgeRouter, and EdgeRouter to opticalModule), all this is telling me that there is probably even a win to be had by getting an additional router/switch, and optically isolating my EdgeRouter 10X from the rest of my home network. In this configuration, I’d have a LPS powered ONT to LPS powered EdgeRouter 1, which is then optically connected to the Extreme and optically connected to a SMPS powered EdgeRouter 2. The EdgeRouter 2 would then be connected to my WiFi and the rest of my home network.
That would give me a minimal network path from the street to the Extreme, with fiber, LPS, and pretty good galvanic isolation from the rest of my home network. For now, that’s a journey for another day, and a different corner of this particular rabbit hole waiting to be explored.
So After All That, How Does Local Storage vs File Serve vs TIDAL Compare?
With all the same caveats before of not knowing the masterings on TIDAL, how do all these music sources compare after all these network optimizations?
Sound quality is danger close now, enough so that I can much more clearly identify TIDAL songs that are different mastering than what I have locally on my Extreme and on my Mac Mini file server. I still give the SLIGHTEST of nods for dynamics to local content, but we’re into a level of nuance that is almost disrespectful to the astonishing sound quality and musical performances I am hearing.
Closing Thoughts on Network Optimizations
Taking a step back, not only am I astonished at what I’m hearing with networking changes, but I’m even more astonished that I am able to hear it at all. For all my hopes that jumping to an end game digital server would put an end to the system optimizations, the Extreme is so astonishingly revealing and performant that everything you feed it (power and network) is holding it back in some way.
The only other “True Reference” component I’ve experienced like this is an mScaled Chord DAVE – Both DAVE and Extremes represent (to me) an absolute reference that you can only take away from. Put the effort in to feed the Extreme better inputs (power and network), and you will be rewarded handsomely way beyond your wildest expectations.
So how about focusing some care and attention to the output side of the Extreme?
USB Optimizations (limited)
Believe it or not, all my testing so far has been done with a generic printer-class USB cable between the Extreme and my DAC. Unfortunately, I have recently gone through a major investigation and upgrade cycle for USB with my NUC reference system, so I have limited cables and USB tweaks to be able to test with Extreme. Fortunately, these are the best I heard on my NUCs, so let’s dive in.
What is there to be gained by optimizing the critical USB connection between the Extreme and your DAC?
Starting at the back panel of the Extreme, I evaluated the following USB upgrades:
- A variety of upgraded USB cables, from my reference and favorite Sablon Audio 2020 USB cable, to the Uptone Audio USPCB and Monoprice SlimRun optical USB extension cable
- A maxed out SOtM tX-USBultra Special Edition powered by a Paul Hynes SR4 power supply
For these tests, I left the Extreme powered and networked in my optimal configuration described in previous sections. I also limited testing to content stored locally on the Extreme.
Impact of USB cables
My first test was to compare the generic USB cable with my reference Sablon 2020 cable. In previous testing, I had iterated through many USB cables (including DIY shielded cables) before settling on the Phasure Lush^2 cables (with DIY layered on top of those as well). I was amazed at how different materials and shield configurations had such a profound impact with USB cables.
When Sablon Audio released their 2020 update to their USB cable, it received rave reviews from many ears that I trust so I ordered one unheard. As I have mentioned earlier in this review, 10 seconds into my first listen with the Sablon, I was done experimenting with USB cables and shielding and put all that stuff away. The naturalness and transparency and relaxed presentation of the Sablon 2020 USB cable was perfect (no compromise) for me.
It’s taken several weeks of listening tests with the Extreme to finally get here, but I was very eager to finally hear how the Sablon 2020 USB cable sounded with the Extreme. Dropping it in, I had that same “first 10 seconds” reaction all over again. Just a wonderfully engaging and welcoming natural presentation, incredibly detailed, but effortless and without any stress. This is the most invisible and transparent USB cable I’ve ever heard, and I’m hearing it all over again. Such a delight, and in the world of audio nervosa, such a gift to completely trust a component like this and know it is just right (I’d missed this cable more than I knew).
Alas, I do not have access to other well regarded USB cables (like the Shunyata Sigma USB cable) to test, but if someone is willing to share, I’d be delighted to give them a listen and update this review.
The second USB cable discovery that allowed me to end my USB experimentation came from a fellow audiophile who’s ear I not only completely trust, but whose bias toward transparency and presence aligns extremely well with my own. He suggested that I audition the Monoprice SlimRun optical USB extension cable, which turned out to be gold. The Monoprice SlimRun converts the USB signal to an optical signal, and at the other end has a female USB connector that converts it back to regular USB. It is available in lengths up to 160 feet, so it gives tremendous flexibility in where you place your music server (and creates some distance from your DAC for any RF generated by your server).
The Monoprice SlimRun is not quite galvanically isolated, since there is a power line for the end connector and USB Vbus that runs the length of the cable, but there is an option to externally power the cable with a micro USB connector on the computer end of the cable. Powering the Monoprice USB SlimRun with an Uptone Audio LPS 1.2, you get all the essential benefit of galvanic isolation, but with USB data rates (no compromise), and cable lengths as long as you practically could hope for, with no apparent sound quality changes with longer lengths.
The Monoprice SlimRun is a USB extender, so It does require a USB cable to connect to the DAC at the end of the extension run. I variously use the Uptone Audio USPCB (an astonishingly good value device, made incredibly practical with the Monoprice SlimRun) and the Sablon Audio 2020 USB cable.
Comparing the Sablon Audio 2020 USB cable direct from Extreme to my Chord DAVE vs Sablon USB cable with the Monoprice SlimRun, the SlimRun (powered from Extreme) does add a touch of harshness to playback. Switching power on the SlimRun to a PowerAdd Pilot Pro2 battery cleaned that right up, giving the same wonderful presentation I’m used to. When I cheated and reclaimed a LPS 1.2 from my network router, things got much cleaner and got me to my end state.
Comparing the Monoprice SlimRun with Sablon USB cable vs the Uptone USPCB, I give the edge to the Sablon for all the reasons I cited earlier. Simply delightful.
Impact of USB Regeneration
With my NUC music servers, I heard tremendous benefit with USB regenerators like the Uptone Audio ISO Regen. Prior to getting the Extreme, I found my end game USB regenerator with the SOtM tX-USBultra Special Edition. The impact this had with my low power NUCs was magical, and it scaled wonderfully with the best power supply I had available (a Paul Hynes Limited SR4 @ 12V).
Placing it in the chain with the Extreme seemed to actually slow down the remarkable speed and dynamics of the Extreme. As good as the tX-USBultra SE is (and it is one of the crown jewels of audiodom), amazingly it seems like it is holding back what the Extreme can do natively.
To get a meaningful USB upgrade for the Extreme, it may be necessary to jump to something like the JCAT USB Card XE or JCAT USB Card Femto, paired with some high quality power supply or perhaps the spare +5V supply in the Extreme. I look forward to auditioning these at some point in the future.
I have become a big believer in mechanical and vibration isolation for any power supply and any piece of digital equipment (particularly DACs). With the astonishingly small signal levels that we’re impacting and hearing, even small vibrations will induce noise and reference power instabilities that are surprisingly audible.
Particularly with the Chord DAVE, that remarkable sense of depth and holographic space that I treasure so much is the most susceptible to vibrations. Rob Watts, the talented and always insightful designer of the DAVE attributes this sense of depth with the DAVE to extreme fine signal level noise reduction. I’ve learned to key off that sense of depth and space to identify when things are not what they should be.
The design of the Extreme reflects a significant and extremely thoughtful focus on mechanical stability and vibration isolation, particularly for key components in the power supply. When it comes to power supply stability and reference voltage stability, I’ve found that even seemingly small improvements in vibration isolation can have an outsized impact (as Emile has so succinctly explained, a voltage applied to a capacitor causes it to vibrate, and vibration applied to a capacitor causes it to generate a voltage).
With that background, out of the box I intentionally gave little attention to vibration isolation with the Extreme. As I noted earlier, Emile has already gone to tremendous lengths with the case work and internal vibration isolation around key components in the power chain (I’m blown away actually...that level of attention to detail in a commercial offering is way beyond anything I’ve seen before). For my testing to this point, I’ve had my Extreme inches away from a speaker on a generic media console with no effort put into vibration isolation, with nothing but a Pottery Barn towel under the Extreme.
So starting from here, I wanted to test the following:
- Take advantage of the flexibility of the Monoprice SlimRun USB extender to move the Extreme to the corner of the room, placed away from speakers and my DAC, directly on a wood floor
- Add a double stack of Taiko Audio Daiza vibration isolation platforms under the Extreme
For these final tweaks, I kept the Extreme in its optimal power and network configurations, and used the Monoprice SlimRun USB Extender and Sablon 2020 USB cable to connect the Extreme to my Chord DAVE DAC from a distance.
Impact of Moving Extreme to Edge of Room and Onto Daiza Vibration Control Platforms
First order of business was to cajole my daughter and partner out of their rooms with even more chocolate, and impose on them to help me lower the Extreme to the floor. Once there (and toes safe), I used some furniture movers and lots of huffing and puffing in very tight quarters to get the Extreme happily in its new home:
The flexibility one gets with optical fiber networking and the Monoprice SlimRun optical USB extender can not be overstated. I could have put the Extreme virtually anywhere in my listening room, or even in any room in my house (the only restriction I had was proximity to my dedicated circuit power outlet). In the end, it puts a smile on my face to see a 100 pound music server capable of such astonishing music reproduction nestled under a century old Steinway (and until post Pandemic when we have house guests again, I can get away with having it there.
I placed my Extreme on a double stack of Taiko Audio Daiza vibration damping and control platforms. The platforms are beasts of precision milled 25 pound slabs of panzerholz, with an intriguing crop circle spiral pattern milled on the underside and substantial copper and panzerholz footers.
(Photos used with permissions from Taiko Audio website)
Getting the Extremes stacked on the Daiza’s in this tight space was (to say the least) EXTREMELY challenging. Thankfully no fingers were crushed, no disks were ruptured, and no scratches were scratched. Alas, given the pain of getting things in place (and the laughter of the folks eating chocolates and watching me suffer), my willingness and ability to test different configurations (one vs two Daizas, etc) and do proper A/B listening tests is basically zero. Post Pandemic, I’ll be delighted to host a listening session where someone (else) is changing the configuration while I do critical listening for audible changes.
All that being said, this panzerholz is pretty remarkable material (this is my first in person exposure to it). Panzerholz was first developed during World War II as a modest-weight armor material for tanks and other vehicles. By taking multiple layers of laminated wood, and compressing it under tremendous heat and pressure and steam to ~½ its size, the material becomes an ultra dense cellulose laminate which is actually bulletproof(!)
For audiophiles, panzerholz is compelling for its vibration dampening properties. By combining very high density with a low transmissibility and high internal damping, panzerholz is used for everything from piano soundboards to turntable bases and tone arms and cartridges to case work to vibration isolation.
For the Daiza, Taiko has taken the panzerholz that they strategically use internally to the Extreme, and built a vibration control platform that complements the design of the Extreme case itself. The crop circle spiral pattern on the underside is intended to tune the natural modes of the platform, with additional damping with the foam filled foundations for the footers and in the foam filled copper and panzerholz footers. The mechanical properties of the Daiza are engineered to be a complement to the aluminum case work of the Extreme, dampening the high frequency resonances of the casework while minimizing any self resonance (dampening rather than ringing). Emile and others have also noted the compound benefit of stacking Daizas, offering better damping and vibration absorption.
As I learned more about the unique material properties of panzerholz (and how it side steps a lot of the unintended challenges with other materials and vibration isolation strategies) and learned about the care that had been put into designing the Daiza as a mechanical complement to the Extreme, I took the leap and ordered a stacked pair for my Extreme, and a smaller Daiza for my Chord DAVE.
When I first got the Daizas, I was struck by how they felt physically, and how evocative they were. They had a wonderful warmth of wood, but with a density and stillness that was striking (almost like holding a dense deadened metal block, but with warmth and tactile inviting feeling of wood). Placing a smaller Daiza under my Chord DAVE, the impact was very immediate and tangible: much more imaging clarity and detail (classic benefits from vibration isolation for my DAC), but also a calm and stillness that I found to be very compelling and engaging, and even comforting. This additional sense of calm was a new experience for me, and not something I had heard with other vibration isolation materials and schemes.
Assessing the impact of the Daizas with the Extreme was more difficult. Given the newness of my experience with the Extreme and the enormous challenge of placing a 50 pound stack of panzerholz under a 100 pound beast of a server in a very tight space, it is very difficult for me to speak with specificity to the differences I am hearing with the Extreme on 0, 1, or 2 Daizas (for practical reasons, I jumped right to two, and I’m not touching the stack again until I hit the gym for a couple months in preparation). However, the end result of all the tweaking is a calm, ultra fast yet ultra controlled presentation that feels simultaneously effortless and incredibly powerful.
I’m grasping for analogies for what I am hearing, but I am reminded of the finest dancers, who have an incredible quiet grace and control that seems both absolute and absolutely effortless. They also evoke a quiet power that can explode at any moment into spectacular movements. As I reflect on and relish what I am hearing and experiencing, the Extreme on Daizas evokes that beautiful sense of being in the presence and in the moment with that remarkable dancer.
So with the Extreme all tweaked out and optimized, how does it sound? In Part 5 (LINK) I share notes from extended listening sessions, and bring it all together to close out this review.
Community Star Ratings and Reviews
I encourage those who have experience with the Taiko Audio SGM Extreme to leave a star rating and quick review on our new Polestar platform.