SOULNOTE has introduced a new reference music server, the R-3, alongside two network transports, the Z-3 ver.2 and Z-3X. At the centre of all three products is USS, or Ultimate Sound Stream, a proprietary LAN audio transport developed jointly with SFORZATO.
The significance of the new lineup goes beyond adding another server or streamer. SOULNOTE is addressing the entire digital playback path — from server to network transport, and from transport to DAC — as one integrated system.
In a typical configuration, the R-3 sends audio over USS to a Z-3 transport, which then forwards the signal to a compatible DAC via ZERO LINK or USB.
USS Focuses on Processing Load, Not Data Accuracy
USS is SOULNOTE and SFORZATO’s new LAN audio transport.
The idea is not that conventional Ethernet or TCP/IP somehow fails to deliver the correct music data. Digital integrity is already maintained in standard network transmission.
Instead, SOULNOTE focuses on what happens inside the playback hardware while that data is being received and processed.
The company argues that CPU activity, packet handling and changing processing load can create variations in power consumption and electrical noise inside the transport. Those effects may then influence clocks and sensitive audio circuitry.
USS is designed to simplify the receive-side workload as much as possible, with the aim of reducing electrical activity around the audio path.
A dedicated R-3 server is not strictly required. Windows PCs can also transmit USS after installing the company’s USS ASIO driver.
Does This Sound Familiar? Diretta Comes to Mind
The basic idea behind USS inevitably recalls Diretta.
Diretta also separates a Host from a Target and aims to minimize processing activity and power fluctuations on the playback side, rather than treating network audio purely as a question of bit-perfect data delivery.
The two technologies are not the same, however.
Diretta is a broader Host-to-Target network audio protocol, while USS is SOULNOTE and SFORZATO’s dedicated transport between a server and a network transport.
After the Z-3 receives the audio over USS, another technology — ZERO LINK — handles the connection from transport to DAC.
So while the philosophy is similar, the scope differs. Diretta focuses on simplifying the network-renderer side, whereas SOULNOTE uses USS plus ZERO LINK to simplify different stages of the digital playback chain from server to DAC.
R-3: A Music Server Dominated by Its Power Supply
The R-3 is SOULNOTE’s reference music server and the primary source component for USS.
It includes a 1TB internal SSD and can also access music stored on USB drives or NAS devices on the same network.
SOULNOTE says the CPU, operating system and motherboard were selected through listening tests, while its own User Optimizer software is used to reduce unnecessary processor activity.
The most striking part of the design is the power supply.
According to the company, approximately 75% of the internal volume is devoted to power-supply circuitry.
The R-3 uses a newly developed 300VA transformer, eight high-speed SiC rectifier diodes and 42 low-multiplier electrolytic capacitors in a non-negative-feedback discrete linear power supply.
SOULNOTE describes this as the world’s first non-feedback discrete linear supply used in a music server, although that claim is based on the company’s own research.
The chassis measures 454 × 174 × 393mm and weighs approximately 22kg, making the R-3 physically closer to a large high-end component than a conventional computer-based server.
Z-3 ver.2 and Z-3X: General-Purpose or USS-Only
The network transport comes in two versions.
The Z-3 ver.2 supports USS as well as RAAT for Roon Ready operation, DLNA and UPnP, making it the more flexible option.
The Z-3X, by contrast, limits its LAN input to USS only. SOULNOTE’s intention is to remove general-purpose network protocols and simplify the software environment further for users who want to build a dedicated USS system.
The Z-3X is therefore not a Roon Ready endpoint.
Both models can be switched between Z-3 ver.2 and Z-3X software configurations, and existing Z-3 owners will be offered a paid upgrade path.
Both are priced at ¥1,650,000 in Japan and are scheduled for release in November 2026.
USS and RAAT Occupy a Similar Role — But Serve Different Goals
In practical system terms, USS sits in roughly the same part of the chain as protocols such as Roon RAAT.
For example:
Roon Server → RAAT → Z-3 ver.2
or:
R-3 → USS → Z-3 / Z-3X
Both are network audio transports between a server and an endpoint.
The difference is in priorities.
RAAT is designed as a broad ecosystem protocol that handles device discovery, format negotiation, synchronization and playback control across many manufacturers.
USS is much narrower. Its design is focused specifically on reducing processing activity at the receiving transport.
So USS should not be described as a direct RAAT replacement, but it does occupy a comparable functional layer in the playback chain.
USS and ZERO LINK Are Different Technologies
Another important distinction is the relationship between USS and ZERO LINK.
USS handles network transmission between the server and the transport.
ZERO LINK handles the digital connection between the transport and a compatible DAC or player.
SOULNOTE and SFORZATO designed ZERO LINK to move asynchronous processing away from the DAC and allow the transport to operate from the DAC’s clock reference.
A complete SOULNOTE system might therefore look like this:
R-3 → USS over LAN → Z-3 → ZERO LINK → D-3
or, with a conventional USB DAC:
R-3 → USS over LAN → Z-3 → USB → DAC
The Z-3 can also act as a bridge from USB input to ZERO LINK output.
SFP Networking and Support Up to PCM 768kHz
The Z-3 ver.2 and Z-3X provide both standard RJ45 Ethernet and an SFP port.
With an optical transceiver installed, users can connect the server and transport over fibre, while a copper RJ45 SFP module can also be used.
Supported audio formats extend to PCM 32-bit/768kHz and DSD512, with digital output available through ZERO LINK and USB.
The power supply uses a 260VA transformer, 16 SiC rectifier diodes and a four-rail non-feedback discrete design with a total of 70 electrolytic capacitors.
Each transport weighs approximately 20kg.
How Far Should Network Audio Be Separated?
What makes the R-3 and Z-3 platform interesting is not simply the cost or the unusually large power supplies.
Most modern network players combine server functions, network reception, DAC processing, clocking and analogue output into a single chassis.
SOULNOTE is moving in the opposite direction.
It separates these functions and treats processing load and electrical noise at each stage of the digital chain as engineering variables worth controlling independently.
RAAT and UPnP/DLNA prioritise flexibility and compatibility. USS prioritises a narrower and more controlled playback path.
Whether that design produces an audible improvement will ultimately require independent comparison. Changing the network protocol does not make the underlying music data more accurate.
The claim is instead about electrical behaviour inside the playback hardware.
Still, the fact that SOULNOTE is designing everything from the server-to-transport protocol to the transport-to-DAC interface illustrates how far high-end network audio has moved beyond simply adding more streaming services.
Key Specifications
| R-3 | Z-3 ver.2 / Z-3X | |
|---|---|---|
| Role | Reference music server | Network transport |
| USS | Output | Input |
| Roon | Can feed Roon-compatible endpoints | Z-3 ver.2 only: Roon Ready |
| Internal storage | 1TB SSD | None |
| Network | RJ45 | RJ45 / SFP |
| Digital output | USS over network | ZERO LINK / USB |
| PCM | Server output | Up to 32-bit/768kHz |
| DSD | Server output | Up to DSD512 |
| Power transformer | 300VA | 260VA |
| Weight | Approx. 22kg | Approx. 20kg |
| Availability | November 2026 | November 2026 |
| Japan price | ¥2,090,000 | ¥1,650,000 |
Source : SOULNOTE
https://www.soulnote.co.jp/detail_r3.html
https://www.soulnote.co.jp/detail_z3ver2.html