Ledger-Preserving Forms of Explicit Formulae: Typed Bookkeeping and Instrument Calibration

Two expressions for the same quantity can agree as outputs while preserving different internal structure: a reduced scalar can omit the factorization, boundary type, regularization prescription, or cancellation path that produced it. This note offers a modest methodological vocabulary for auditing such cases. We call a representation "ledger-preserving", relative to a stated question, when it keeps the typed constituents of a calculation visible before reduction to a scalar or an undifferentiated remainder; we stress that such a ledger is a chosen factorization adequate to the question, not a canonical decomposition of the output. The aim is not a theorem but an audit protocol: state the question, declare the regularization prescription, separate the typed ledgers, record provenance before cancellation, calibrate the measuring instrument, and only then estimate the remainder.

We illustrate with two standard objects. First, the elementary ball--cube--ball recursion: in dimension three the full cycle reduces to the scalar \(1/(3\sqrt3)\), while the factored ledger
\[
\left(\frac{2}{\pi\sqrt3},\frac{\pi}{6},\frac{1}{3\sqrt3}\right)
\]
records where \(\pi\) enters and cancels and separates a curved Euclidean boundary from a flat/cornered cubical one. Second, the Riemann--von Mangoldt explicit formula, written as a prime/zero/archimedean ledger on \(u=\log x\); Gaussian smoothing then supplies an instrument law
\[
H_\sigma(\gamma)=e^{-\sigma^2\gamma^2/2}
\]
with visibility gate
\[
\gamma_{\mathrm{trust}}=\frac{\sqrt{2\log(1/\varepsilon)}}{\sigma},
\]
which separates instrument geometry from arithmetic content before residuals are read. The numerical examples are diagnostics for the explicit formula, not evidence for the Riemann Hypothesis: critical-line reconstructions fix \(\rho=\tfrac12+i\gamma\) and therefore test ordinates, not abscissas. The contribution is bookkeeping discipline, not a new theory, geometry, or identity.