SYNTHETIC EXAMPLE

Digital-to-Wave Testbench

Synthetic dimensionless 32-value software stress vectors · synthetic example

Interpretation

In this synthetic software condition only, the existing nominal decoder passed 0 of 12,000 frames at +8 samples, while an oracle given the exact timing passed 11,346. The oracle is a diagnostic upper bound, not a timing-recovery algorithm or a fair competing decoder. This does not establish physical timing or device performance.

Run record

Project version: v5 exploratory timing-stress run (no release version stated)
Code version: v5 exploratory timing benchmark source snapshot
Code commit: Not recorded
Evidence class: synthetic
Dataset type: synthetic software test; six-vector panel
Dataset notes: Selected scope is POOLED_ALL_VECTORS at an arbitrary +8-sample timing offset (12.5% of a symbol) with synthetic IID Gaussian sample noise SD 0.45.
Seed: 20261007
Baselines: oracle_exact_timing (diagnostic upper bound; not a competing decoder)
Protocol: v5 integer-sample timing protocol · 5 (exploratory)
Protocol lock claim: Conditions were frozen before the full run, according to the source note; it is not described as a formal preregistration. (source-stated; not independently time-verified)
Protocol SHA-256: 8136de81b3bd1c399e093b5751043195393acacabdccd55158631edfa753e5cc
Protocol lock commit: Not recorded
Project source: Not supplied
Dataset source: Not supplied
Source note: The included output, protocol note, and source file are copied from local saved project artifacts; no public repository URL or commit is asserted.

Reported metrics

Values are copied from the declared run summary. This report does not calculate scores, rank methods, or combine metrics across scopes.

POOLED_ALL_VECTORS; offset +8 samples; noise SD 0.45

Method	Metric	Value	Unit	95% interval	Note
existing_nominal_decoder	frames passing	0	frames / 12,000	—	—
existing_nominal_decoder	frame pass share	0	fraction	[0, 0.00032001912]	Two-sided 95% Wilson interval, conditional on the fixed vectors and synthetic IID draws.
existing_nominal_decoder	empirical MAE	1.8717171	dimensionless units	—	—
existing_nominal_decoder	empirical RMSE	2.3016992	dimensionless units	—	—
oracle_exact_timing	frames passing	11346	frames / 12,000	—	—
oracle_exact_timing	frame pass share	0.9455	fraction	[0.94129408, 0.94942078]	Two-sided 95% Wilson interval, conditional on the fixed vectors and synthetic IID draws.
oracle_exact_timing	empirical MAE	0.063550505	dimensionless units	—	—
oracle_exact_timing	empirical RMSE	0.079653815	dimensionless units	—	—

Limitations

The vectors, waveform, timing offsets, and noise are synthetic and dimensionless; there is no physical channel or device measurement.
The chosen +8-sample offset is an arbitrary stress value, not a physical operating range.
The protocol is exploratory, not a formal preregistration or physical validation.
The uncertainty interval quantifies IID simulation variation conditional on the fixed panel and stated model, not model or physical uncertainty.
The oracle uses exact timing and is a diagnostic upper bound, not a timing-recovery method.

Artifact integrity

All referenced files match their manifest SHA-256 values. This confirms bundle consistency only; it does not establish authorship, timing, provenance authenticity, or measurement validity.

artifacts/digital-to-wave/v5-timing-protocol-lock.md — protocol; SHA-256 8136de81b3bd1c399e093b5751043195393acacabdccd55158631edfa753e5cc
artifacts/digital-to-wave/robustness-benchmark-v5-timing.json — results; SHA-256 10f5643e834c5df2a417b6bcca9d6c43a067156c2a9fcb8533c0af6bb2f7c722
artifacts/digital-to-wave/robustness_benchmark_v5_timing.py — source_code; SHA-256 1bd9500ce3ceadadae25ce6731bc8ea9058adefa2781adf4d16d403b6c383419

SHA-256 matching proves only byte-for-byte consistency between a referenced file and the digest recorded in this bundle. It does not prove authorship, when a file or protocol existed, provenance authenticity, or measurement or method validity.