Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by ray-dude

  1. I I certainly hope to do precisely [email protected] (I think you’ve been reading my proposal to Chris :)
  2. Part 1 - Introduction and Digital Audio Optimization Foundations (Link) Part 2 - Enter the Extreme Part 3 - First Impressions and Basic Configuration (Link) Part 4 - Tweaking Up the Extreme (Link) Part 5 - Extended Listening Impressions, Learnings, and Conclusions (Link) In Part 1 (LINK) of this series, I share my experiences with digital audio optimization, and how when the time came to move beyond my optimized DIY NUC-based two server setup, those learnings and experiences brought me to the Taiko Audio SGM Extreme. As I was researching approaches to push digital server performance to its limits, it quickly became clear that Emile Bok (owner and guru at Taiko Audio) had already done orders of magnitude more experimentation and discovery than I could ever hope to accomplish, and had already pushed a lot of the hypotheses and things on my “to explore” list to their limit, and had already optimized the big levers WAY beyond anything I was aspiring to. Here was an opportunity to learn a tremendous amount about the digital audio optimization road ahead from one of the most respected people and companies in digital audio. How can a self-respecting übergeek audiophile say no to that? Well, what I learned so impressed me that I ordered an Extreme unseen and unheard. After a Pandemic-induced delay, it finally arrived in California a month ago, and boy was it worth the wait. Enter the Extreme The Taiko Audio SGM Extreme is aptly named: an Asus WS C621E Sage motherboard with two Intel Xeon Silver 4210 CPUs (10 physical cores and 20 logical cores(!) each), 12 hand selected custom 4GB RAM modules (48GB total), dedicated Optane storage for OS and up to 24TB of optional M.2 PCIe storage for local music storage, custom power supply, custom precision machined casing for fanless operation and vibration control, generous use of panzerholz vibration absorbing material on key components, an obsessively minimized and tuned OS to dedicate as much CPU to playback as possible, all in a 100 pound beast of a precision machined and assembled package. The Extreme hit all the key attributes that I was looking for: massive number of Xeon cores to distribute and isolate workload (while only using a vanishingly small part of system capacity during music playback), dual CPU motherboard to better isolate music processes and threads from other OS processes and threads (the ultimate in network-bridged servers, if you will), a staggering amount of PCIe storage to keep all music content local and as logically “close” to the CPU as possible, clean and massively stable power, passive cooling (fanless) even with all that power, an obsessive level of attention to RF and mechanical vibration control, fast low latency memory with affinity to different CPUs (even more isolation), a highly tweakable base OS (Windows) that has been pre-tweaked to maximize audio performance, and a person and company behind it all with an obsession to build the best possible digital audio server for the best possible audio playback, and to continually evolve it going forward. Of course, the price tag was very scary, but it was within a reasonable factor of what I was already budgeting for a high-power system build. Given the opportunity to not churn through endless variations of motherboards and cards and cases and power supplies (coal and water and chalk for those that have read Part 1), I jumped at the opportunity to purchase an Extreme unseen and unheard. My sincere thanks to Emile for his patience and openness during my research process. Alas, during the six week lead time between when I ordered the unit and when I expected to receive it, the world changed and we all went into lock down. After an additional 4 week delay (some understandable supply chain challenges, and FedEx prioritizing medical supply shipments over audiophile equipment), I received my Extreme mid April, with an unexpected amount of home time to put it through its paces and enjoy the music. Out Of The Box When my Extreme arrived, my FedEx driver was (the say the least) VERY unhappy as he carried the box to my door. The string of creative expletives he let loose both coming to the door and walking back to his truck was as stark and impressive and inspired as an Ornette Coleman solo. The Extreme is a beast, weighing in at 100 pounds. Fairly intimidating, and not something you want to drop on your toes. The commotion was enough to draw my high school senior out of her room. After letting her know that my new music server had finally arrived (“It’s called ’The Extreme’ sweetie”), she quipped “That seems on brand”, dropped the mic, and went back to her room. After an appropriate time to let any FedEx-contributed CV19 nasties on the outside box fade away, it was time to slide it into my family room and unbox the beast. Inside the box is the Extreme and nothing else (BYO cables). The custom CNC machined case work is meticulous and stunning, and built like an armored vehicle designed by Dieter Rams: Overall each case takes multiple days to CNC mill, but the end result is simply spectacular. The case work is precision CNC machined aluminium (10mm thick top and bottom panels, and 15mm thick front and back panels), with a generous mix of copper and panzerholz (tank wood) materials. Taiko claims that the ~6,000 ventilation holes were designed for more than just cooling, and are also radiation waveguides (RF cleanliness is next to digital audio godliness). The cooling fins look identical, but are solid milled aluminum on the left side (for power supply cooling) and solid milled copper(!) on the right side (CPU cooling). The machining precision is stunning. Parts are perfectly flush, and (although not easily seen in the photos), the finish on the aluminium is perfectly even and quite beautiful. The fit and finish is well beyond anything I’ve seen in a small volume consumer product like this. Interesting details abound, like the use of vibration absorbing panzerholz wood for the footers (more on panzerholz later). Aside from the illuminated power switch on the front panel, all the goodies are accessible on the back: A USB 3.1 Gen 2 socket to your DAC, multiple USB 3.1 Gen 1 sockets and LAN ports for data, a StarTech PEX1000SFP2 PCIe optical Network card, a 280GB Intel Optane 900P storage drive for the operating system, a VGA(!) port, an ASUSTek Hyper M.2 x16 v2 PCIe storage card for your music library (it comes standard with 2TB, but I purchased 8TB of storage to fill a single card), the power supply, and a mysterious port for future options. Optionally, you can order the Extreme with S/PDIF or AES/EBU (single, dual, or quad wire) connectivity. You can also optionally order up to a total of 3 ASUSTek PCIe storage cards, for up to 24TB of music storage, in increments of 2TB. The Belly of the Beast The inside of the Extreme is even more stunning than the outside: The Extreme is powered by a custom linear power supply with a 400VA transformer (I couldn’t get in there to see details), custom Lundahl choke regulators, and a stunningly large 700,000µF capacitor bank with the highest audiophile grade Mundorf capacitors, and even a pair of Dueland(!) capacitors. This bank of capacitors is perhaps the most extreme part of the Extreme, a true extravagance of riches (I’ve read about Dueland capacitors but had never even seen one in person, let alone two). Emile has not spoken too much about his power supply strategy, but certainly the heart of the secret sauce of the Extreme starts and ends here with this remarkable power supply. I am still waiting (and waiting) on my custom Paul Hynes SR7 build, but my intuition is that it would be a very interesting horse race between Paul’s design approach (extreme attention to voltage regulation and output impedance), and the approach Emile is taking with the Extreme power supply (extreme focus on filtering and fast current reserves). Looking at this bank of capacitors and hearing what I’m hearing, I can’t help but think that Emile may have gotten it right. Although I’ve always found Paul’s designs to excel at speed and dynamics vs capacitor heavy designs, massive massive filtering does open the door to more brute force capacity on the power side of the supply. This is “If it’s worth doing, it’s worth overdoing” taken to the Nth degree. Will it be easier for Emile to evolve his design strategy to faster and faster supplies over time? Time will tell, but I suspect there is a lot more goodness to be found in Emile’s strategy. I can’t wait to hear how both power supply strategies evolve over the next couple years! Panzerholz again makes an appearance on the interior of the Extreme, adding vibration absorption and isolation to both the critical transformer assembly and the power supply coil. The panzerholz mounting platform for the transformer assembly is particularly impressive, and a very welcome thing to see on such a performance critical component. Everywhere you look inside the Extreme, you see this type of attention to detail on performance critical components (see the copper offsets? The intentional blend of materials to diffuse resonances?). The care and attention put into packaging and integrating the finest grade components available is stunning. The careful blending of materials to complement each other speaks to the care and thought and what must have been endless prototyping that went into the design. Shifting attention to the heart of the machine, the Extreme is built on an Asus WS C621E Sage motherboard with dual Intel Xeon Silver 4210 Scalable 10 core CPUs, each with 10 cores and support for 20 logical cores (40 logical cores total). The Extreme also comes standard with 12 sticks of custom hand selected industrial 4GB RDIMM RAM, for 48GB RAM total. This is a beast of a motherboard, stocked to the 9’s. (image and schematic credit to ASUS, extracted for convenience from their website and manual) Some interesting details to note from the technical documentation for the ASUS motherboard. Each Xeon processor has direct access to its own bank of 6 memory slots and its own PCIe x8 and x16 channels. By populating all the memory slots available on the motherboard, Taiko has built in the ability to segregate and isolate processes to different CPUs (and CPU cores), memory, and PCIe channels. They have also spread memory access across the entire bandwidth of all the available memory channels on the motherboard. The heavily tuned Windows 10 Enterprise installation of the Extreme takes advantage of this topology to deliver maximum sound quality on playback. Conceptually, this sort of dual CPU strategy appeals to me, delivering all the process isolation benefits of a dual server configuration (music server and music endpoint), but with an ultra performant and coherent CPU to CPU Ultra Path Interconnect (UPI) bus connecting the CPUs instead of ethernet. The dedicated DAC USB 3.1 Gen 2 port uses an advanced ASMedia 3142 USB controller, which acts as a PCIe bridge to USB. These ports are capable of 10Gb/sec data transfers. Although audio would never require these data rates, it speaks to the performance these USB controllers are capable of, and the general philosophy of the Extreme design to have the highest bandwidth and most powerful components possible, and to utilize as little of that bandwidth and power as possible. Given the criticality of the USB interface to digital audio, it is also interesting to note that the ASMedia 3142 has its own custom PHY layer and control layer. I wonder how much of the sound quality advantage of the Extreme comes back to this particular USB controller implementation (recognizing that the other side of the USB handshake in the DAC is almost certainly using a more generic USB chipset...at some point the DAC itself becomes the performance bottleneck in the chain). Later I’ll discuss my comparison of USB direct from the Extreme vs through a top of the line SOtM tX-USBultra special edition powered by a Paul Hynes SR4. The Extreme also uses a custom passive CPU cooling system, leveraging six 80W capable heat pipes and the incredibly large copper cooling fins on the right side of the Extreme (machined out of a solid 53 pound block of copper!). If you want to know where that 100 pound weight comes from, the copper cooling fin for the CPUs accounts for 22 pounds of it on its own. The investment in passive cooling is both extravagant and stunning. In operation, even with the crazy amount of computer power bundled into the Extreme, the unit is essentially silent, with the case running barely warm to the touch. As shipped, the Extreme has several open PCIe slots for future expansion and upgrades (for example, an audiophile-grade USB card or network card from JCAT, or additional PCIe storage). As shown in the block diagram above, by default each PCIe slot has an affinity for one or the other of the CPUs: CPU 1 - 1 (16x, blank), 2 (8x, SFP card), 6 (8x, blank), 7 (16x, Music storage card) CPU 2 - 3 (8x, OS storage card), 4 (8x, blank), 5 (16x, blank) This is another tuning opportunity to have I/O cards more directly linked to the CPUs where dependent processes are running. In the case of the Extreme, CPU 1 has direct affinity to all music I/O (network, USB, storage), with CPU 2 delegated to OS I/O. By delegating system and music processes to cores on the appropriate CPU, Emile is able to achieve an optimal balance for sound quality. The ASUSTek Hyper M.2 x16 v2 PCIe storage card supports up to 4 2TB M.2 storage drives (up to 8TB per card). Although its on-card fan is disabled (presumably to reduce system noise), the Extreme has no issues keeping things passively cool at music data rates. The machining of the interior components is as precise and clean as the external components, with a generous and thoughtful mix of aluminium, copper, and panzerholz throughout (when you consider the importance of vibration/RF/temperature control in a device like the Extreme, the elegance and brilliance of the way various materials were selected and blended for different properties really comes to the fore). Interestingly, there is an additional connector on the power supply copper panel, with +5V and +12V DC available. This is intended for future expansion (as Emile says, Extreme is a project, not a product). Along with the optional port on the back of the Extreme, I’m intrigued by what may be in works as the Extreme project evolves. It is also tangible proof of Emile’s commitment to make the Extreme as future proof as possible – as new motherboards and new connectivity standards emerge, the Extreme is designed to evolve with them (and as anyone that has tried to get the most of out USB and digital audio would attest, a new connectivity standard for digital audio would be a godsend). The Softer Side The Extreme comes pre configured with a highly tuned custom version of Windows 10 Enterprise LTSC 2019, as well as Roon and JPLAY playback software. Optionally, you can ask Taiko Audio to install HQPlayer Desktop for you. Once I tracked down a VGA to HDMI adapter (wow, was that a dusty box), I took a peek at the BIOS, but (other than being a bit overwhelming with all the server-class BIOS settings) nothing stood out as interesting. I confirmed that secure boot was off, so at some point I may try to boot up the Extreme with Euphony on a USB stick. The system is intended to be run headless (no monitor, no keyboard, no mouse), and comes with VNC Viewer for connecting to it over your network. It also comes preinstalled with TeamViewer for remote support (Emile has been quite generous in hopping on the machine to check things out and tune things up, even with the 9 hour time zone difference between California and the Netherlands). By using this particular version of Windows, Emile has been able to strip down the installation to the bare minimum components needed for a music server. By significantly reducing the number of running processes, each process has less contention on each of the 40 logical cores. In addition, Emile uses tools like Process Lasso Pro to allocate processes and threads to where they run best, and contend the least. While this is fantastic for a turnkey optimized system, be aware that if you install any additional software on the Extreme, you may be unintentionally degrading sound quality by disrupting the optimized allocation of Roon and JPLAY processes to between the dual CPUs. Caution before you start experimenting with different software packages and system configs. Similarly, the LTSC version of Windows allows Taiko Audio to validate and bundle system updates before sending them out. As a best practice, you should resist the temptation to be updating OS and software packages on the Extreme, in case there are changes that disrupt the sound quality tuning in the system. As a very late breaking example (after I had finished all the listening tests and this review), there was a recent change that had a significant adverse impact on sound quality that required a retuning of the system. Emile kindly remoted into my system to make the adjustments (and things sound fantastic!). It is a reminder of how delicate the tuning is for the Extreme, and how even at this level of performance and system capacity, the delicate nature of the remarkable magic the Extreme can deliver. That being said, in practice you plug the Extreme into your home network, turn it on, connect to and configure the Roon Core running on the system, and Bob’s your uncle. When the spirit moves you, use your favorite VNC client (I happen to use Screens on my MacBook Pro) to connect to the box for things that you can’t do within Roon (such as add a driver for your DAC, copy over your music library, etc). I’m a Roon and HQPlayer user, so it was very easy for me to drop into the Extreme environment. I had not used JPLAY before, but it is pre configured and shows up in Roon as a USB driver, so it has been very easy to use. On my machine, Emile was kind enough to also install the Chord ASIO driver for me, giving me three different ways to connect to my Chord DAVE (WASAPI driver, JPLAY driver, and Chord ASIO driver). There are differences and tradeoffs between these drivers, which I will describe later. So what about other playback software other than Roon? I have not had a Windows box in my home for over a decade, but I’ve been intrigued by software like JRiver Media Center and foobar2000. For folks looking to tailor their system with their preferred music software, these packages can be installed and run on the Extreme, but appreciate that the system has been carefully tuned to have the maximum sound quality when used with Roon. At some point, I’m looking forward to being more comfortable with my understanding of how the process allocation optimizations on the Extreme are set up, so I would have confidence to do my own experiments and listening tests with alternative music software. Until then, I’m very glad that Emile has leaned into my favorite software (Roon) as the playback software of choice for the Extreme. Based on my interactions with Emile so far, I also have complete confidence that if a better package comes along, Emile will be able to retune the system to get the most out of whatever that package turns out to be. (As an aside, it is difficult to overstate how incredibly engaged and accessible Emile is for customer support and system support. As more and more people sign on for the Extreme project, I hope he’ll be able to occasionally get some sleep and time with the family, but his commitment to share his knowledge and expertise and take care of the community is a big part of why I had such confidence signing up for this journey.) So all this looks incredible. How does it work and sound in the real world? In Part 3, I’lll jump into first listening impressions, and configuring and optimizing the out-of-the-box performance of the Extreme. 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.
  3. If you're OK waiting until tomorrow, I go into great detail on storage strategies (and the why's) in Part 2. Basically storage is direct PCIe connected storage (for OS and Music), which is consistent with a lot of findings here and elsewhere that this sort of storage results in better sound quality (presumably by giving CPUs more direct access to storage)
  4. Terribly sorry about this guys! I love it when folks like Rajiv post their review playlists, and I realized as I was writing this that I couldn't do the same. It may be time to make a "Streaming Demo" playlist for future reviews and give up these old favorites. (And thank you @DuckToller for putting this together!)
  5. I figure I’ll owe anyone who makes it through all 5 parts a beer (or three
  6. Editor's Note: This week Audiophile Style is going to extremes. In fact, we are calling it Extreme Week here on AS. Three months ago when I first talked to Ray about publishing a review of the Taiko Audio SGM Extreme, I had no idea he would put an extreme amount of work into the effort and put together a treatise on his experience. When I read through it the first time, I knew we had to do something different. Thus, we have dedicated an entire week to publishing Ray's review of the SGM Extreme. One part each day, Monday through Friday. I hope the AS Community enjoys reading this review series as much as I have. - CC Reality Quest: Going to Extremes with the Taiko Audio SGM Extreme (Part 1 of 5) A back (and bank) breaking 100 pound $26,000 digital music server with 40 Xeon cores and 48GB of custom memory and 8TB of M.2 PCIe storage, unironically named “Extreme," with a reputation for being the ultimate mic-drop Summit-Fi for digital audio? Wow... So what is audiophile life like at these rarified heights? How does the Taiko Audio SGM Extreme stack up against an already OCD-tuned and optimized NUC digital music server/endpoint setup? Can one possibly justify this sort of expense and extravagance for a music server? If one were crazy enough to take this leap, are there ways to improve upon Extreme and take its performance higher still? Earlier this year, semi-rationalized irrationality got the best of me, and I decided I needed the answers to these questions. Many emails and discussions with audiophile friends and Taiko Audio guru Emile Bok later, I took the leap and ordered an Extreme. Alas, a month later the world changed, and after a brief Pandemic-induced supply chain and shipping delay, I received my Extreme and have been putting it through its paces ever since. TL;DR – It’s pretty damn awesome! (TL;DR)2 – Here is where I ended up after all the tweaking. In many ways, the Taiko Audio SGM Extreme represents the culmination of many years of community learnings about computer audio and computer audio optimization. Each aspect of the Extreme (physical, electrical, mechanical, compute, I/O, software, etc.) is taken to the (no pun intended) absolute extreme, all in the carefully designed and considered service of digitally reproducing the best sound possible. The Extreme is Emile Bok’s opus, a seminal product that is the distillation and embodiment of the state of the art in digital audio. Even for those that have not been able to hear it first hand, the Extreme offers much to appreciate, and especially for system designers and DIY types, much to learn from. To give each aspect of the Extreme the attention it deserves, this review is broken up into 5 parts: Part 1 - Introduction and Digital Audio Optimization Foundations In this part, I share my personal perspective and journey with digital music servers, and how those experiences and biases inform and color my experience of the Extreme. I also share details on equipment under test and the music I typically use when evaluating new components and systems. Those well-steeped in the practice and lore of digital music server optimization may wish to jump straight to Part 2. Part 2 - Enter the Extreme With the Extreme, the component selection, casework, power and mechanical design, system configuration, and numerous design details all reflect a deep understanding of system design and a thoughtful care for how all the pieces come together into a whole that is much greater than the sum of the parts. In Part 2, I dive into Emile’s masterwork, and break down these pieces and explore how they work together. Part 3 - First Impressions and Basic Configuration In Part 3, time to listen to some music, and see what the Extreme is able to do straight out of the box. Part 4 - Tweaking Up the Extreme In Part 4, it is time to take the Extreme from Ridiculous Speed to Ludicrous Speed to Plaid! How does careful attention to power, networking, USB, and mechanical factors impact the performance of the Extreme? It turns out, quite a lot! Part 5 - Extended Listening Impressions, Learnings, and Conclusions And finally in Part 5, time to put the A/B tests and audio nervosa tweaking twitches away, really listen to some music, and reflect on how all these technologies come together into what is truly a remarkable experience of music. Let’s dive into Extreme week! Digital Audio Optimization Foundations Digital optimizations and hygiene in digital audio can be a contentious topic (bits are bits, right?). Rather than debate mechanics, I am sharing my experiences in my home with my equipment with my ears and my biases and my aspirations for connecting emotionally to remarkable musical performances. Almost certainly yours will vary. Hopefully my experiences are helpful to you, but if not, that’s OK too. What matters for me is how the notes come together to something larger, a resonance that is greater than the sum of its parts. As an imperfect analogy of that “something larger” experience, consider how mixing coal + chalk + water + iron filings in ever more precise proportions using ever more esoterically sourced materials won’t result in a child that you will cherish and adore and make the center of your life. Both your child and the coal + chalk + water + iron filings are made of the exact same things. Arguing about the purity of the water or where the coal was sourced or the magnetic properties of the iron for me changes nothing. The real debate is how they come together, and how it comes alive and becomes meaningful. A few years ago, I was introduced to the Chord Mojo and then the Chord DAVE, and they have completely changed my experience of music. For the first time I've gone from “listening to music” to “participating in a performance”. That feeling of intoxication or euphoria from the performance has been driving me to optimize my analog chain, and now to go deep into the digital side of my chain. When a recorded performance truly comes alive, it draws me in the same way I am drawn to an amateur performance in a coffee shop or a subway. This sense of reality is difficult to describe, but it is only loosely correlated to what I think of as “typical” measures of distortion or tonal balance. When I am walking by a coffee shop, I can tell if there is live music being played inside (vs recorded music). Even with all the traffic noise and going through walls and glass, I can somehow tell it is real people with real instruments even from the street, and I am attracted to it and want to listen. That has nothing to do with our normal measures of tonal balance and noise floor or sound stage or imaging clarity or being in the sweet spot. The music coming from inside the coffee shop sounds like crap (by any normal audiophile measures), but it still sounds (and is) real. Somehow, our brains know how to deal with the natural distortion of real noises that are going through walls and barriers and overcome background noises, and to focus on the "real". When that “real” clicks in, for me it is a fairly abrupt transition, and my experience of music steps up to a completely new (and intoxicating) level. On the digital side, Chord DACs (and especially the Chord DAVE) are by far the best I’ve heard at creating this sense of reality. After I heard the Chord Mojo, I ordered a Chord DAVE unheard, and have since rebuilt my entire system around the magic that the DAVE is able to deliver. On the analog side, I’ve discovered that the more I got rid of, the more transparency I discovered and the more reality I was able to experience from the DAVE. I’ve replaced my beloved B&W 802d3’s (and associated amps, etc.) with high efficiency single driver Voxativ 9.87’s with 104db sensitivity Voxativ 4D drivers. My DAVE, with its lowly 2W output, can drive these speakers to ear splitting levels. More importantly, I’ve been able to eliminate amps and crossovers, having DAVE wired directly to a single ultra fast and ultra sensitive point source driver. The end result is an experience that is holographic, astonishingly quick and controlled, and extraordinarily immersive. The level of transparency and reality is intoxicating, and remarkably revealing of everything in the chain: any weakness or system stumble takes away from that sense of reality. How to preserve as much of that sense of “real” as possible? How to get more? With my analog and digital-to-analog system anchors in place, my attention shifted to the digital side. My digital audio journey has been driven by the core principle that any digital optimization must start and end with the DAC. Obviously, if one is starting with bit-perfect sources (and if you’re not, fix that), these bit-perfect bits are not affected at all by any digital optimizations (if bits were being changed, you would hear very audible pops clicks and drop outs). Any sound quality change you hear is due to something in your digital chain somehow and in some way impacting your DAC (even if you have no idea how or why). For me, the core question for digital hygiene and optimization is what can be done to the digital chain to minimize anything being injected into or induced in the DAC that impacts the core digital references in the DAC: reference clock, reference ground, and reference power. The impact may be direct, or it may be indirect through several intermediate components, over wires, or through the air. If you hear a difference with bit perfect audio streams, something is being transmitted to or induced in your DAC that is impacting the core digital references in the DAC (clock, ground, reference voltage), which in turn becomes audible when the signal gets converted to analog. As my digital chain has evolved from laptop to low power Raspberry Pi endpoints to special built devices like the Sonore microRendu to DIY NUCs with Audiolinux and Euphony to all sorts of power and USB and network and cable tweaks, I’ve stumbled into a high level strategy and hypothesis for digital optimization that has consistently pointed me in interesting (and usually better) directions: Invest in cabling, power, and digital hygiene to do as little harm as possible (directly or indirectly) to the DAC’s clock, ground plane, or reference voltage plane Invest in a digital endpoint that moves bit perfect digital data from ethernet to a USB DAC with as little timing variability and as much timing and signal integrity as possible (aka, as close to precision real time data streaming to DAC as possible) For me, the key issue isn't so much latency, as it is reducing the variation in latency. By way of example, playback that has a latency of 1 second +- 0.00000000001 second would be preferable to a latency of 0.001 sec +- 0.0001 sec. For my current reference system, I am using an i7 NUC running AudioLinux in RAM as my Roon server, connected through an Uptone Audio EtherREGEN to a second i7 NUC running Euphony in RAM with HQPlayer as my endpoint, connected to my DAC by USB via a SOtM tX-USBultra special edition. As my system has evolved to this point (informed predominantly by the experiences of the extended community here at Audiophile Style, especially the legendary Novel thread, and the soon to be legendary DIY thread), here have been my experiences that are consistent with this “reducing variability in latency” hypothesis: By using a NUC with highly integrated Silicon on a Chip (SoC) design (vs laptop or regular computer), sound quality is better (the SoC results in better timing coherence and optimized signal paths between system components) By going with a stripped down in-memory linux OS with everything else disconnected and disabled in BIOS, sound quality is better (near real time OS) By configuring linux to give the highest real-time priority to the endpoint software (Roon Bridge or Squeezelite or HQPlayer), sound quality is better (more near real time) By configuring linux to give the highest real-time priority to the IRQ for the USB port (low level interface between the operating system and the physical USB port), sound quality is better (even more near real time) By removing the USB boot stick so the only thing on the USB bus is the DAC, sound quality is better (fewer IRQs for USB storage) By configuring the music server to bridge ethernet traffic to the NUC endpoint, sound quality is better (isolating ethernet traffic and interrupts) By increasing network buffer for Squeezelite or other end point software so that all ethernet traffic is front loaded and the current song and next song are loaded into memory within a second or so, sound quality is better (fewer ethernet interrupts during playback) By increasing the size of the memory buffer from Squeezelite or Roon or HQPlayer to the Linux ALSA driver, sound quality is better (less variability on feeding the USB driver) By having a higher powered NUC (quad core i7 vs dual core i5 Celeron), sound quality is better (able to be closer to real time since it has more horsepower) By configuring linux to dedicate a CPU core to the USB IRQ, sound quality is better (less contention with other processes) By configuring linux to dedicate a CPU core to the endpoint software, sound quality is better By using industrial-rated wide temperature RAM (presumably tighter tolerances parts, and better performance even at room temperature), sound quality is better By playing music stored on Optane storage on the PCIe bus (vs SATA or SSD), sound quality is better (more direct transfer path from storage to the CPU) By regenerating USB and/or ethernet signals, sound quality is better (signal integrity improvements) None of these seem like they are RF noise related. These are all computer/OS/software configuration tweaks. What they all seem to share is the impact of reducing variability in the latency of the signal from the endpoint to the DAC and improving signal integrity, and thus (presumably) making the life of the PHY interface of the USB chipset a lot easier. Separately, any time I have mains isolated a digital component (batteries, super caps, etc), sound quality has improved. Any time I’ve inserted a better/faster power supply (Paul Hynes SR4, Uptone Audio LPS 1.2, LDPWR DXP supplies, etc) that delivers well-regulated voltage under very dynamic current loads, sound quality has improved. Similarly, better conductors and better shielding consistently improves sound quality, as does reducing mechanical vibrations on power supplies and digital components. I'm not drawing conclusions nor pointing at a core root cause. However, I'm finding that any change I make that is consistent with improving power (stability and speed), reducing RF, improving ground, and reducing variability in the latency of the digital signal seems to have a positive impact on sound quality. Interestingly, I have also experienced the inherent tradeoff between having a higher power system that is capable of delivering lower variations in latency, but having sound quality suffer from requiring more power (and the difficulty in keeping that higher power level stable and clean). My i7 NUC consistently sounded better than my i5 NUC, but sound quality suffered as I increased CPU speed. The better and more dynamically responsive the power supply that I used with my i7 NUC, the more I could push CPU speed (and the better things sounded) before sound quality peaked and started to suffer. So as I started looking for my next (final?) digital server, a core question for me was how powerful and near real time a digital server can one build, while still being able to deliver high quality power and not requiring additional noise generating components like fans? Can functions of music server and music end point be combined on a single powerful box (with more direct high performance connection between them) while also being clean enough to not adversely impact the signal output with power-related noise? How much control could I have over where processes were running and what system priority each was getting? As I researched DIY options to push the boundaries on my NUC servers, I began to hear whispers about a digital music server that was well beyond anything else available out there, and checks all the boxes I’ve come to find as important (and a whole lot more). The more I dug in, the more I realized that through the rabbit hole was a world of digital audio optimization that was WAY beyond what I was thinking was practical or doable, and that every detail somehow and some way still seemed to matter. In Part 2, I’ll dive into the guts (real and virtual) of the Extreme, and set the stage for Part 3, where I share what I discovered in the world of Summit-Fi digital audio optimization. Equipment As Tested Digital Source – Review Setup Taiko Audio SGM Extreme with 8TB storage option ($27,000 as reviewed) HQPlayer Desktop ($235) Digital Source – Reference NUC Setup Stock Intel NUC7i7DNKE in standard case with 16GB consumer RAM and 32GB Optane stick ($800) Audiolinux ($120) and Roon Server ($120/year) NUC7i7DNKE motherboard in a fanless Akasa case with 4GB wide temperature industrial Apacer RAM and a 16GB Optane stick ($1,000) Euphony ($289) and HQPlayer Embedded ($235) PowerAdd Pilot 2 battery ($90) Stock Mac Mini file server ($500) DAC setup Chord DAVE ($13,000) – 2 channel For two channel for Chord DAVE, PCM+ mode, HF filter off, no crossfeed Chord Hugo TT2 ($5,500) – Headphone For headphones for Chord Hugo TT2, crossfeed set to 2 Direct Two Channel Setup RCA direct to Voxativ 9.87’s with Voxativ 4D drivers ($50,000) Custom Furutech RCA to female banana adapters Custom Furutech XLR to female banana adapters Voxativ Ampeggio Speaker Cables ($4,000) Iconoclast SPTPC speaker cables ($2,500) Headphone setup Abyss AB-1266 Phi TC ($5,000) Sennheiser HD800 (with SR mod) ($800 used) Cables Monoprice USB SlimRun optical USB extender cables ($150) Sablon Audio 2020 USB cable ($750) Uptone Audio USPCB ($35) Ghent Audio JSSG360 DC cables (various) USB SOtM tX-USBultra Special Edition ($1,300) Paul Hynes SR4-19 power supply ($500) Network AT&T fiber to the home, with Alcatel-Lucent G-010G-A ONT DXPWR DXP-1A5S single stage power supply ($109) PowerAdd Pilot Pro2 Battery ($90) Pace 5268ac Gateway/Router (AT&T supplied) Ubiquiti EdgeRouter 10X router/gateway ($109) DXPWR DXP-1A5DSC dual stage power supply ($159) PowerAdd Pilot Pro2 Battery ($90) Planet Technology MGB-TLX SFP transceivers ($30 each) Sonore opticalModule ($249) Uptone Audio LPS 1.2 power supply ($435) Uptone Audio EtherREGEN ($640) Uptone Audio LPS 1.2 power supply ($435) Monoprice and Cable Matters Cat 8 Cables ($10) Sablon Audio 2020 Panatela Reserva ethernet cable ($375, provide as demo) Sablon Audio router noise dongle (provided as demo) Vibration Control Daiza vibration Isolation platforms ($600) Custom precision rollerball isolators Music As Tested My Go-To demo tracks when demo’ing new pieces of equipment, and what I look for in each: Pink Noise (mono), from “The Ultimate Headphone Demonstration Disc” by David Chesky (Chesky, 24/192 FLAC) Yeah, its pink noise, but it tells you a lot about a set of headphones as you get oriented to your system. For 2 channel audio, I’ve found that pink noise is a great way to see if your speakers are aligned and you’re in the right spot. Great baseline reference to make sure things aren’t wonky and you’re in a good place (and it is sort of fun to think about all this technology being used to play a 24/192kHz noise file. Murakkaz Ah Ya Muddasin, from “The Splendour of Al Andalus” by Calamus (MA Recordings, DSD64) Remarkable recording in what sounds to be a majestic and spiritual centuries old cathedral in Spain. With the right equipment, you are transported to a place you’ve never been to but always want to get back to. When the full group joins in, it is profoundly challenging to reproduce the mids and highs without sounding shrill and congested. When the reproduction is effortless, it is magical (to say the least…my jaw drops every time when it’s “right”) If the sound chain is able to maintain that glorious soundstage, it is off the charts. Todd Garfinkle is a magician behind the microphone. Noche Maravillosa, from “Salterio” by Begonia Olavide (MA Recordings, 16/44.1 FLAC) Another gem of a recording and performance from Todd. The precision and clarity of the instruments (particularly the percussive ones) is intoxicating and tangible. Voglia Di Vita Uscir, from “Buenos Aires Madrigal” by La Chimera (MA Recordings, 16/44.1 FLAC) Todd also is a master at capturing performances in a more orchestral setting. These performances by La Chimera are a joy, and the sense of being there with the performers is absolutely intoxicating. Mahler: Symphony #2 'Resurrection’, 1st movement conducted by Iván Fischer (Channel Classics, DSD64) Mahler: Symphony #2 'Resurrection’, 5th movement conducted by Iván Fischer (Channel Classics, DSD64) Near perfect performance with a perfect recording. The dynamics and power of this performance are vivid and real: the orchestra virtually screams with one voice during the climaxes. This recording beautifully captures the essence of horns and low strings, which are very difficult to reproduce. The closest I’ve heard to the experience of the dynamics of a live orchestral performance. Rimsky-Korsakoff: Scheherazade, 1st movement conducted by Fritz Reiner (Analogue Productions Remaster, DSD64) The most perfect recording of the most perfect performance I’ve ever heard. Listening to this recording on a transparent system is a life changing experience: you are standing with Maestro Reiner in Chicago as his orchestra reaches for a performance for the ages. A cultural treasure, and worthy of building a world class system around. Organ Prelude, JS Bach Magnificat by Dunedin Consort (Linn, DSD64) Motet, JS Bach Magnificat by Dunedin Consort (Linn, DSD64) A breathtakingly lovely recording! The dynamics and harmonics of the organ can range from a muddy “eh” to “holy crap!” depending on the quality of the reproduction. Standing in the middle of the choir is a lovely test of imaging and voice reproduction: the more precise the soundstage the more you can pick out individuals (including depth and height…remarkable). Handel: Messiah - Chorus. O thou tellest good tidings by Dunedin Consort (Linn, DSD64) Handel: Messiah - Hallelujah by Dunedin Consort (Linn, SACD) A magnificent recording, reconstructing the original version of Handel’s Messiah, with a total of 12 singers. The normal complexity of the piece is captured in a way where you can hear each voice in the chorus, and how it comes together into a larger whole. An amazingly intimate performance when the reproduction chain can manage the complexity and dynamics and not have the soundstage become muddy and flat. Arnesen: Magnificat - Fecit potentiam by TrondheimSolistene (2l, 24/192 FLAC) This is such a lovely recording at any quality level, but goes from incredible to other worldly as the chain scales up. The orchestra, choir, and church should all have equal contribution to something far greater than the sum of its parts. When it all comes together, you can feel the three core elements feeding off each, creating a profound joy that sweeps you into euphoria. Stardust, from “Duets” by Rob Wasserman (16/44.1 lossless) Every track on this album is a gem, but this one is particular is a fantastic test of sound stage and imaging. At its best, you hear each backing voice precisely in space, but still presenting as a harmonious whole. In real life, detail and precision spatial placement isn’t hard and clinical, why should it be in reproduction? Oh, Lady Be Good, from “Plays Gershwin” by the Bassface Swing Trio (DSD64) Night and Day, from “Tribute to Cole Porter” by the Bassface Swing Trio (DSD64) These direct to disc Stockfisch recordings are extraordinary. Imaging and dynamics FTW. Recordings like these are why we obsess over the things we obsess about. I’m looking to get lost in the music, and the band appearing to be sitting right there. When tonal balance is just right, these performances just jump off the SACD. Shamas-Ud-Doha Bader-Ud-Doja, from “Shahen-Shah” by Nusrat Fateh Ali Khan (16/44.1 lossless) The first track from what was my surprise 2012 album of the year (see this video for the surprise ending). A remarkable supremely spiritual performance by a remarkable man, captured in an “eh” recording. The question for me when I’m evaluating new equipment is what it can do to elevate a middle of the road recording that is worthy of elevation. Let Me Touch You For Awhile, from “Live” by Alison Krauss (DSD64) I adore Alison Krauss. Having equipment that can reproduce the wonderful emotion and musicality of these amazing artists is why I spend so much time looking for the right speakers/cans/etc. Their Live album is special, and you can feel the humanity and emotion in this track. Fields of Gold, from “Nightbird” by Eva Cassidy (16/44.1 FLAC) An absolutely amazing and engaging performance, with a moderate recording and mics that were over driven. The dynamics of the performance unfortunately saturate the recording, and the breaks and near breaks can be painful in some chains. A wonderful song to listen to, but also a great song to see how much of the song is wonderful to listen to. Tenderly, from “While She Sleeps” by Art Lande (Blue Coast, 24/88.2 FLAC) Cookie Marenco has a gift for capturing piano, guitar, and voice as if you’re sitting in the room with the artist. If you’ve ever sat next to a wonderfully tuned piano with an extraordinary player, you know how magical that experience can be. The best pianos sing with resonances that envelop you. The best musicians know how to coax beauty and life out of the instrument. This recording from Art Lande captures that magic. The stronger the dynamics, soundstage, and precision of the system, the more lifelike this track becomes for me. I haven’t experienced this track topping out: the better the reproduction chain, the more lifelike it becomes. One World, from “Session 1” by Sareena Overwater (Blue Coast, DSD64) One World (Instrumental), from “Session 2” by Sareena Overwater (Blue Coast, DSD64) Real magic from Cookie. These tracks are wired directly to deeply held memories for me, and the stronger the reproduction chain, the stronger the emotion that they evoke. There are better examples of piano performance and better examples of vocals, but the emotional truth and power of this performance is unmatched. 99, from “Blue Coast Special Event 43” by Meghan Andrews (Blue Coast, 16/44.1 FLAC) Cookie knows how to record guitar too, and Meghan Andrews knows how to bring a performance that is worth catching in a bottle. Vous et Moi, from “Night and Day” by Willie Nelson (SurroundedBy Entertainment, 24/96 FLAC) What if Willie Nelson was in a dispute with his label, got pissed off, and invited the best musicians he knew to the studio to record an instrumental album in full surround? Yeah, this actually happened, and it is as awesome as you think it is. The album is amazing in a musical surround setup, but a proper 2 channel system puts you right in the middle of the band. Incredible stuff. Music in My Room, from “The Folkscene Collection, Vol. 3” by Cheryl Williams (Redhouse Records, 16/44.1 FLAC) We’ve all had the experience of being in a coffee shop or small venue, when someone with a guitar and something that has to be shared commands the attention of everyone in the room, and you have a moment where the whole room is one. These CDs where engineer Peter Cutler captured intimate in-studio performances at KPFK in Los Angeles are replete with those moments, but this performance by Cheryl Williams stands out for me. With a great reproduction chain, the guitar is real and present, and the voice and singer connect at a deep emotional level. A jewel of a moment, waiting for the right equipment to be a moment again. All I Want, from “After Blue” by Tierney Sutton (BFM Jazz, 16/44.1 FLAC) Tierney Sutton has a striking clear and present vocal style, and that is on full display on her “After Blue” album of Joni Mitchell standards. I continue to be amazed how ever better DACs extract ever more nuance and subtlety of performance from top tier vocalists. It is a joy to hear the depth of craft and art of vocal performance on tracks like this. Rosa fresca, from “Il viaggio d’amore” by Arianna Savall and Petter Udland Johansen (Carpe Diem, 16/44.1 TIDAL Lossless) “The journey of love” is a marvelous exploration of love through the ages, from multiple cultures and times. The whole album is a joyous wonder, but the opening track (“Fresh Rose”) of a traditional song from the 1500s is an invitation to join in joy and unbounded hope. The company of players is feeling it, and you do too. When you’re hearing every string pluck in the strums and the voices come together into something much larger than the sum of its parts, you’ll be glad you accepted that invitation. L’Amor, from “Bella Terra” by Arianno Savall (Alia Vox, 16/44.1 FLAC) If Rosa fresca makes you fall in love with Arianna Savall singing about love, you’ll want to seek out her “Bella Terra” album. An accomplished harpist and vocalist, Savall is at her best when she brings both together: voice and instrument are one, and evoke marvelous sound and emotional resonances in each other. Traveler, from “Little Crimes” by Melissa Menago (Chesky, Binaural 24/192 FLAC) Airplane, from “Little Crimes” by Melissa Menago (Chesky, Binaural 24/192 FLAC) A gem of a recording from Chesky: direct binaural recording, made in a church while it is raining outside. Like all Chesky binaural recordings, you are there sitting with the performers (Airplane), with special magic from the sound of the rain outside of the church (Traveler). Fantastic test of soundstage and spatial detail. Hold On, from “Sessions from the 17th Ward” by Amber Rubarth (Chesky, Binaural 24/192 FLAC) Don’t You, from “Sessions from the 17th Ward” by Amber Rubarth (Chesky, Binaural 24/192 FLAC) More Chesky magic. No rain this time, but Amber’s rich voice + violin + guitar + percussion are amazing on any system, but the sense of being there scales beautifully as the reproduction chain improves (it is magical when your system crosses some threshold of transparency…all of a sudden you are there). Stuck In A Moment You Can’t Get Out Of, from “Open Your Ears” by The Persuasions (Chesky, Binaural 24/96 FLAC) One last gem from Chesky. There is a profound difference to listening to a recording of a group of people sing, and being with a group of people that are singing. This is another recording that (at least for me), when you cross some magical threshold of transparency, the people become real. Ephemera, from “Ephemera” by Carla Lother (Chesky, 16/44.1 FLAC) A stunning interpretation of an incredible poem by William Butler Yeats. The recording is exquisite, but the incredible sense of space and balance amongst the performers is very fragile. The better the chain, the more perfect the recording becomes. Karamawari, from “Gamushara” by YAMATO the drummers (TIDAL MP3) Drums are notoriously difficult to reproduce in the way you experience them in person. There is a physicality that is lost in most systems. Hearing a group of percussion masters really bring it on a system that can approximate that in person experience? Amazing. Get Lucky, from “Random Access Memories” by Daft Punk (24/88.2 FLAC) Another track that transcends pop when played back through equipment that really reproduces the full range and dynamics of the recording. The subtlety and layers on Nile Rodgers’ guitar work is incredible, and the recording is outstanding so you should be able to hear it all. I listen for whether it is washed out, and how well I hear all the (considerable) nuances in his playing. Take Five, from “Time Out” by Dave Brubeck (Analogue Productions SACD) An excellent test of dynamics at the high end. As an aside, these Analogue Productions remasters are off the charts! No Love Dying, from “Liquid Spirit” by Gregory Porter (24/192 flac) Another lovely recording and performance, that on a balanced system hits a resonance that is next level for me (just sounds “right” and get the “wow!”) When things are not in balance or boomy/shrill, I hear it loud and clear. Beethoven: Symphony #9, 4th movement by Suitner (OG Denon, 16/44.1 lossless) This was the first CD I ever bought in 1984 (first CD ever made?) I know every second of this movement and every nuance. My current 2 channel setup was the first time I had ever heard the entire movement without a break (every other system I’d ever had/auditioned had some break at some challenging passage). So What, from “Kind of Blue” by Miles Davis (Japanese single layer SACD version) One of the finest recordings and performances of the 20th century. Always the last track I play during any audition. Until there is a time machine to take me back to March 2 and April 22 1959, I will buy every new remaster of this album, and play it on every piece of high end audio kit I can find. 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.
  7. @Dutch can you please speak more about this? I'm experimenting with this currently (my Roon server on 192.168.4.x, my laptop and the rest of my home network on 192.168.3.x). I ended up having to setup port forwarding for UDP 9003 to see the Roon server from my home network, and TCP 9100-9200 to actually be able to connect to it. Looking at the Roon config screens, I'm not seeing anywhere obvious where one can manually enter a server's IP address? I have not looked to setting up routing between networks (I have an EdgeRouter X). I have a similar goal is to isolate the Roon server as much as possible from other network traffic
  8. Wonderful write up Rajiv! Very evocative, and I've enjoyed the tX-USBultra SE enough that I really appreciated that as the baseline for comparisons with the Phoenix. I continue to pretend that reference clocks don't exist as a category, precisely so I don't get tempted by things like your description of the REF 10 SE 120...you're making that a very difficult fiction to maintain!
  9. Purchased this new favorite today as part of the Chesky "May Sale". Heartbreakingly lovely...so beautiful!
  10. I did similar back of the envelope math before I purchased my Extreme. Of course, DIY allows you to calibrate how far you travel into the crazy pit, and to spread the pain over some period of time. With Extreme, you are all in and liberated from decision making, but definitely at a price. If you're willing and able to go all in, I think there is tremendous value in the Extreme (much more on that next week...). I also tremendously appreciate the amazing work and insights folks are sharing here...there is so much to learn and to figure out!
  11. I am continually being surprised by the impact of network traffic (and network topology) on SQ. The EtherREGEN started me down this path (surprise -> attention and study) but the rabbit hole is much vaster than signal regeneration. Honestly, I thought this would be at the minor tweak level, but there seems to be something fairly fundamental going on here. Having just switched from a 2 box i7 NUC setup to an Extreme, the impact of network traffic and topology has become at least an order of magnitude more obvious (more surprise -> more attention and study I'll have more to share next week, but when running the Extreme as an end point (Roon, HQP, etc) on the native Windows or booting to Linux from a USB stick, you are giving up a lot compared to running everything on Extreme in a one box solution.
  12. ray-dude

    HQ Player

    @miska I just saw the release notes for the new HQPlayer Embedded. Do you have information you can share on the new sinc-S filter? Looking forward to trying it when it makes its way to HQPlayer Desktop!
  13. Another old friend dropped by for a visit today!
  14. Lots of old friends dropping by the house tonight...it's been far too long
  15. So lovely to spend time with an old friend...
  16. My apologies Olaf, the unit has been sold (I prefer to sell within the US when I can, to avoid customs headaches) To your question, I found the tX to have more smoothness and naturalness and detail and depth and subtlety than the ISOregen (which I loved). The tX SE just took all that that much higher still. I strongly prefer a more transparent and detailed presentation (sense of reality from being in the same space as the performers). The tX gave me all that (and gave more with better power) and never felt analytical to me (neither did the ISOregen). If anything, the tX SE with great power was smoother and more relaxed and had WAY more special presence than the IR (IR is no slouch...loved it in my chain) I hope this is helpful to you...
  17. Over on the main ER thread @cat6man asked: (and @superdad appropriately redirected here) I run my LPS 1.2 at 12V for my EdgeRouter 10X. I don't have any PoE devices on it (unlike the EdgeRouter X, the 10X is passive PoE pass through...the ER X needs the extra voltage for PoE, but doesn't need it to run) I have not tried different voltages to see if it makes a difference.
  18. For sale is an Uptone Audio EtherRegen audiophile switch and ethernet regenerator. This unit is from the first batch shipped late last year, and has a transferable 3 year warranty. Dollar for dollar, one of the best upgrades I’ve added to my system. I was so impressed by how changes to network gear could have such a huge impact on sound quality that I’ve completely rebuilt basically my entire home network to get more of what I was hearing with the EtherRegen (beware, it is that much of an eye opener). Unit is in great physical and working shape. For detailed specs see: https://uptoneaudio.com/products/etherregen For a whitepaper on how network tweaks can possibly be audible see: https://cdn.shopify.com/s/files/1/0660/6121/files/UpTone-J.Swenson_EtherREGEN_white_paper.pdf?v=1583429386 For a great in depth review of the EtherRegen see: https://audiophilestyle.com/ca/reviews/uptone-audio-etherregen-review-and-comparison-r887/ https://youtu.be/Io4SDi5hLxs These are perpetually back ordered, and for good reason. Retail price ~$640. Asking $550, buyer pays shipping and PayPal fees. CONUS only please. Please see pictures. PM with questions, happy to help however I can Spring cleaning time, so see my other posts for a SOtM tX-USBultra special edition and Acoustic Revive TB-38H vibration isolation platform (well matched to the Chord DAVE)
  19. For sale is a SOtM tX-USBultra, upgraded by SOtM Acoustic to special edition (it is fully loaded 12V model, and includes a 75-ohm master clock input, EVOX cap, and OCC silver DC cabling). The unit also includes clock tap points for a variety of clock frequencies (for using the high quality clock with other devices and PCs) This the top of the heap when it comes to USB regeneration. I had two Uptone Audio ISOregen’s that I used to use in series, then I upgraded to a SOtM tX-USBultra. I was so blown away by what I heard that I immediately upgraded to the fully loaded special edition. This is a device that sound incredible out of the box, but scales beautifully as you give it better and better power (I ran it with a Paul Hynes SR4). For sale is the unit only (you provide your own power supply), although if you’re looking to purchase a SR4 drop a note (I’m open to selling them as a bundle…perfect match between the tX-USBultra and the SR4). Units is in great shape (see photos). For detailed specs see: https://sotm-usa.com/products/tx-usbultra-regenerator-special-edition For reviews on the non upgraded model: https://audiobacon.net/2017/05/23/sotm-tx-usbultra-usb-regenerator-review-the-ultimate-digital-defuzzer/ https://www.6moons.com/audioreviews2/sotm4/1.html Having briefly owned the regular tX-USBultra before upgrading to the special edition, the upgrades have a huge positive impact on SQ (I was extremely happy with the upgrade). Selling because I’ve upgraded my music server and now have an internal USB solution, and no longer have a use for this gem. Retail price ~$1270. Asking $750, buyer pays shipping and PayPal fees. CONUS only please. Please see pictures. PM with questions, happy to help however I can Spring cleaning time, so see my other posts for an Uptone Audio EtherRegen and Acoustic Revive TB-38H vibration isolation platform (well matched to the Chord DAVE)
  20. For sale is an Acoustic Revive TB-38H Quartz Isolation Base with a Hickory Top Plate, and 4 Black Ravioli isolation pads. It is perfectly sized for the Chord DAVE DAC (which is where I used it), and the combination has fantastic synergy with the DAVE (see https://www.head-fi.org/threads/chord-electronics-dave.766517/page-301#post-12837752 for a detailed description and review of this combo). The combination is also great with power supplies and other audio items that benefit from vibration isolation.The Acoustic Revive uses a layer of quartz for vibration dampening, tempered by a hickory wood base. The Black Ravioli pads also do a wonderful job with the DAVE. Together, they significantly improve the amazing sense of space and depth that I hear with my DAVE, and bring out great detail and imaging.Aesthetically, the Acoustic Revive platform is basically a perfect match to the Chord DAVE DACFor detailed specs see:http://acoustic-revive.com/english/underboard/tb-38h.htmlhttps://www.thecableco.com/tb-38h-quartz-filled-platform-for-power-strips.htmlSelling because upgraditisRetail price ~$1000. Asking $600, buyer pays shipping and PayPal fees. CONUS only please.Please see pictures (DAVE shown for size, not included in purchase). PM with questions, happy to help however I canSpring cleaning time, so see my other posts for an Uptone Audio EtherRegen and SOtM tX-USBultra special edition
  21. Great report, thank you for posting! I've been doing similar experiments over the last couple weeks. In my chain, an LPS 1.2 on my EdgeRouter 10X had a surprisingly positive impact.
  • Create New...