A zero-retune covariance test for precision QED windows Version: 4.0 Concept DOI: 10.5281/zenodo.20121952 Author: Ali Attar Website: quantumtraction.org Main book anchor: Quantum Traction Theory: Main Book v10.01 This paper tests whether one frozen Quantum Traction Theory source packet can pass a multi-row precision-QED comparison without being retuned by the observations it is judged against. The central object is not a fitted least-squares model. It is a source packet printed first, then pushed through a declared covariance row packet: \[ \mathcal P_{\rm QED}^{\rm QTT} \longmapsto \left\{O_i^{\rm QTT},\,O_i^{\rm ref},\,\Sigma_{ij}\right\}_{i=1}^{13}, \qquad N_{\rm fit}=0. \] The v4.0 master firewall is \[ \frac{\partial\mathcal P_{\rm QED}^{\rm QTT}} {\partial\{O_i^{\rm ref},\Sigma_{ij}\}} =0. \] The release adds the Rydberg-orthogonal transport theorem. For a homogeneous QED observable, \[ O_i^{\rm hom} = C_i\,m_e^{\rm source}\alpha^{p_i} = C_i\,\frac{2hR_\infty^{\rm QTT}}{c}\,\alpha^{q_i}, \qquad q_i=p_i-2, \] so the row transported into the covariance packet is \[ \frac{O_i^{\rm QTT}}{O_i^{\rm ref}} = \exp(\delta r_i+q_i\delta a_i). \] This is the load-bearing v4.0 discipline: the electron-source and Rydberg scales are not counted twice. The comparison is made in the dimensionless orthogonal packet that precision spectroscopy actually tests. For the declared 13-row packet, the no-retune covariance verdict is \[ \chi^2=8.164204027853675,\qquad \nu=13,\qquad \chi^2/\nu=0.6280156944502827, \] \[ p=0.8327404747638795, \qquad \max_i |z_i|=1.7588374430584028. \] The covariance matrix is positive definite, every leave-one-row and leave-one-block audit passes, the omitted-normalized-covariance stress envelope remains inside the declared gate, and an illegal \(+5\sigma\) photon-edge displacement is rejected: \[ \chi^2_{\rm illegal}=54.2060, \qquad p_{\rm illegal}=5.5667\times10^{-7}. \] The scientific status is therefore: SIGMA-QED-13-ROW-NO-RETUNE-COVARIANCE-GREEN SIGMA-QED-RYDBERG-ORTHOGONAL-TRANSPORT-CLOSED SIGMA-QED-COVARIANCE-POSITIVE-DEFINITE-CLOSED SIGMA-QED-LEAVE-ONE-ROBUSTNESS-GREEN SIGMA-QED-ILLEGAL-PHOTON-EDGE-NEGATIVE-CONTROL-REJECTED SIGMA-SOURCE-ONLY-SI-CLOSURE-NOT-CLAIMED SIGMA-PLANCKIAN-MATERIAL-PROTOCOL-ONLY The scope is deliberately narrow. Version 4.0 closes the frozen-source, multi-row QED covariance audit for the declared packet. It does not claim a source-only derivation of the SI metre, and it does not promote the separate Planckian-material protocol beyond its protocol status. Related QTT anchors: Main QTT book: 10.5281/zenodo.17527179 Artian Electron Source-Lab Rydberg Theorem: 10.5281/zenodo.20800371 QTT Photon-Edge Gate: 10.5281/zenodo.20628735 Fine-Structure Constant in QTT: 10.5281/zenodo.20330268 Charged-Lepton Family-Rank Constructor Theorem: 10.5281/zenodo.20698191 Corpus Tree family: quantumtraction.org/doi-map/#concept-10-5281-zenodo-20121952 Observatory QED row: quantumtraction.org/observatory/#obs-qed-no-retune Lexicon QED no-retune entry: quantumtraction.org/lexicon/#term-qed-no-retune-window-suite Two Universes source/readout page: quantumtraction.org/two-universes/ Included files: Main PDF paper, Version 4.0, uploaded first for Zenodo preview. Complete reconstruction ZIP containing LaTeX source, source and comparator packets, covariance matrices, scripts, machine certificate, render audit, and SHA-256 manifest.

Version 4.0 executes the declared 13-row precision-QED covariance packet with zero fitted source parameters. It adds the Rydberg-orthogonal transport theorem, keeps the source packet frozen before comparator access, and reports the positive-definite covariance verdict, leave-one robustness checks, and illegal photon-edge negative control.