Engineering programme
What a credible prototype must show.
The v0.2 specification is a proposal. This page separates its present analytical evidence from the hardware and interoperability work still required.
Implementation status
No public O2AT prototype is claimed.
The O2AT v0.2 document defines proposed framing, packetisation, capability negotiation, receiver credits and clock states. This site publishes its calculations and a labelled illustrative model. No public reference implementation, bitstream or hardware bridge has been identified for this release.
Existing playback reports, including the author’s reported DSD512 playback using LAiV equipment, concern that equipment and do not demonstrate an O2AT transport implementation.
Prototype requirements
Evidence to collect, not results to assume.
- Capture and decode both Ethernet directions. Verify the 12-byte header, sequence handling, FCS rejection and exact capability descriptor values.
- Measure actual 1G-SX module interoperability, host SerDes/MAC behaviour, optical power margin, module power and loss/reacquisition.
- Exercise PCM and native DSD packing with bit-exact vectors, channel order, group counters, malformed payloads and endpoint rejection.
- Measure FIFO occupancy against packet delay, bursts, dropped frames, stalls and source rate error. Include data in flight in the credit bound.
- Measure conversion-clock frequency, uncertainty, relative drift, phase noise, spurs and time error in both local and external-reference modes.
- Interrupt and restore 10 MHz reference; verify declared holdover bounds, local mute/abort and fresh-session recovery.
- Use a physical reference plane and instruments to document uncertainty, temperature and reproducible test configuration.
- Publish independent encoder/decoder test vectors and confirm interoperability with at least two independently developed implementations.
System boundaries
Transport does not remove endpoint ceilings.
A 1G transport calculation does not prove that a host USB controller, bridge, FPGA, DAC, I²S path or analogue output supports the same rate. The source paper explicitly separates the transport envelope from installed output capability.
For fixed-rate live legacy inputs, a finite FIFO and receiver credits cannot force an unpausable producer to run at the receiver’s rate indefinitely. A bridge requires a genuine clock-control mechanism, an explicit conversion strategy or a defined safe rejection condition.
ADAT and S/MUX bridging, preservation of legacy encoded formats and full metadata mapping are separate engineering scope. The paper treats them as future work.
Conformance evidence
Release only against a named draft.
Any test result should identify the ODT hash, wire-version nibble, encoder/decoder revision, endpoint profile, instrument model and configuration, traffic conditions, environmental conditions, and raw output files needed to reproduce the claim.
Controlled testing of EtherType 0x88B5 is suitable only for an evaluation network. A commercial deployment requires the identifier and interoperability programme to be resolved.
Workflow diagrams
Review the proposed behaviour visually.
Thirty-four editable Mermaid diagrams cover the v0.2 transport and identify proposed clock-transfer and hub extensions separately. The pack’s grammar checks passed; diagrams remain open to visual review and engineering discussion.
Hardware identification
Use the compact mark at device scale.
The outlined superscript O²AT mark is used on the website; this baseline version is reserved for compact device identification. Neither file specifies minimum manufacturing dimensions.