I visited your room/system after a long time. I am impressed with the components changes and room acoustics, you’ve chosen wisely. As far as your NUC (streamer), if I can make one suggestion, if you haven’t already; try an Ethernet based optical isolation before you consider any other upgrade. Being streaming as your main source, I would do everything possible to optimize your source and cleanest signal into your amazing DAC.
Do streamers make a difference?
Just added the Nagra Streamer and I can wholeheartedly say yes, it does. Without buttons, remote control or a screen, it is elegant in its simplicity. On the other hand, its music selection is anything but simple as it accommodates Qobuz Connect, Tidal Connect, and Spotify Connect. It is Roon ready/Roon Tested, JPLAY Certified, and plays with Audirvāna, Airplay 2, UPnP/dlna for local files and vTuner for radio access.
It is a wonderful device and I hear more music and I hear the music more clearly. But in the interest of full disclosure, is it the streamer or the result of removing a USB connection?
My previous system used a Roon Nucleus Plus as server and streamer and was fed via an AudioQuest Vodka RJ/E ethernet cable from an Xfinity XB7-T modem/router. The Nucleus was then connected via an AudioQuest Dragon USB cable to a Playback Designs USB-X4 interface, which connected to a Playback Designs MPD-8 DAC using their proprietary high-bandwidth fiberoptic cable. In addition to the fiberoptic connection providing galvanic isolation, the USB-X4 also reduces jitter with a clock that is identical to the clock in the MPD-8.
The new system has the Nucleus and the Nagra Streamer each connected via Vodka ethernet cables to the Xfinity XB7-T. Nagra and Playback Designs share design technology and the Nagra Streamer connects to the MPD-8 with the same proprietary fiberoptic connection as the USB-X4 and also incorporates the identical clock, yielding the same sonic benefits in terms of noise and jitter reduction.
There is definite improvement in terms of detail and space. There is simply more to the music. I venture that the streamer contributes the lion’s share but acknowledge that I have removed a USB connection from the flow, which may also contribute to the difference.
While I believe it is a great addition to my system, the Nagra is not for everybody since its connectivity is limited to only two outputs, the second being an S/PDIF. The USB port is an input for hard and flash drives.
The matching Nagra Compact PSU almost doubles the purchase price so I have ordered a Teddy Pardo LPS to replace the included SMPS.
I do not use a network switch since I have only the two ethernet connections and see no reason to isolate one from the other. I also believe clocking is not an issue, but I am interested in what others think on both fronts.
Of note, EMM employ a similar proprietary fiberoptic cable set-up but utilize a different format that is not compatible with Nagra and Playback Designs.
- ...
- 102 posts total
Thank you, Lalitk. So... my understanding is that ethernet packets aren't even guaranteed to arrive in order—they're reassembled at the receiving end. Data arrives bit-perfect or gets retransmitted, and packet timing is completely decoupled from the DAC's audio clock. (USB is similar: no timing data in the stream itself.) More to the point, standard Ethernet is already galvanically isolated via the magnetics in every physical layer. So the only possible mechanism for additional optical isolation to help is preventing RF noise from polluting the streamer's internal ground or power rails. That might matter if I had: Severe ground loop problems creating audible hum My competently designed streamer already isolates the network interface from the audio circuits. Optical isolation would be masking a design flaw, not improving a working system. I've seen no controlled evidence showing audibility in systems that aren't already broken. For my setup, as I understand it, this is a non-problem. |
About the LDMS, I took a closer look, and while I'm sure it's a well-built piece of kit, I can't get past some of the core claims at this price point. Here are my comments, to the best of my ability. Jitter: The marketing emphasizes "virtually jitter-free playback," but jitter at the server is irrelevant in a properly designed system. The DAC reclocks the incoming stream—its oscillator determines jitter performance, not the source. This framing made sense for CD transports with S/PDIF, but it doesn't translate to network streaming. Galvanic isolation: Standard Ethernet already provides galvanic isolation via the magnetics in every PHY. "True galvanic isolation" sounds impressive, but I'd want to know what's being additionally isolated and why. Vibration control: Chassis damping matters for turntables and CD transports with spinning media. For a solid-state device with SSDs, I'm skeptical there's a problem to solve here. The vague bits: "State-of-the-art thermal optimisation" and "cutting-edge EMI/RF shielding" don't tell me anything. What specifically is being done, and what measurable or audible improvement does it yield? None of this means it's a bad product. The linear PSU is a legitimate design choice, the casework looks beautiful, and the feature set is sensible. But at £24k, I'd want the value proposition grounded in things that demonstrably matter for digital audio—not language borrowed from contexts where it made sense but may not apply here. Not for me, but I'd be genuinely curious if anyone's done controlled comparisons against more modestly priced streamers. |
@hilde45 This is valid only where the DAC receives signal input via USB. With SPDIF/AES EBU/I2S, there is no reclocking by the DAC, and jitter is indeed a concern for the server because its oscillators determine the timing. Even where the DAC does the clocking (USB), the jitter performance of its oscillators are oftentimes susceptible to some extent by infiltrating galvanic noise generated by the server. |
Great, interactive, and educational post. @fastfreight @tcutter and everyone else, I found this discussion to be very beneficial. I have a similar setup to some of you. My main DAC is the Naga Tube DAC with a separate power supply along with a Siltech Royal Double Crown AES cable connected to my Grimm MU1 streamer. Currently, I am demoing a Nadac D and Nadac C by Master Fidelity with a Kubala-Sosna USB connected to a "homemade NUC." I know using the AES cable to connect my Grimm (since it does not have USB out) to the Nadac D+C defeats the clock; however, initially that was the only way to test the Nadac D. I have not spent much time demoing with the new NUC. I did speak directly to Mr. Xu, who was incredibly helpful with setup and filter selection using the AES, and he was the one who recommended a NUC to test the Nadac C+D. |
- 102 posts total