Minimum norm and realizability: C2⋆ vs C3
The comparator that separates the mathematically valid solution from the physically executable one: move the line width, the offset and the distortion, and watch why C2⋆ holds while C3 demands a resonator 45% wider. Every figure cites its origin.
Realizability oscilloscope
Where ρ_crit is decided
Ratio |h̃|²/F along the pulse. ρ_crit is its maximum where F exceeds the tail threshold; the dot marks where it is reached.
Capture vs offset Δ
The flat line is C2⋆'s invariance: a common phase ramp leaves the Gram unchanged.
Capture vs distortion β
C3 matches C2⋆ across the sweep: in this channel it follows from the parity of the modes.
ρ_crit vs tail threshold
C3 does not move while the threshold is ≤ 10⁻⁴: its limit sits on the pulse's leading edge. C0, C2 and C2⋆ rise towards 1 as more tail is included, bringing the calculation close to the manuscript's ≈ 1.03.
ρ_crit vs distortion β
The dashed line is the receiver's ρ. Above it, the family cannot be executed.
The same figures, computed now
Engine tests
Published v1 engine vs corrected v2
Default configuration: mode f₁, β = 0.30, Δ = 3.5 MHz, threshold 10⁻⁶. C0 and C2 captures do not change; C2⋆ and C3 do.
Auditor's findings
Engine
Hypotheses and limits
Reference figures and their origin
Export and seal this experiment
JSON with the configuration and the results of the four families, and its SHA-256 fingerprint to pre-register a prediction.