# Aligned contact numerators and physical detector scales The physical relation `lambda=4/(rho*T^2)` gives `a*lambda=lambda+4`. The two contact numerators coincide, and the original and normalized weights have the same rational square class. This also makes the actual pole inputs have the same square class whenever both traces are nonzero. On the stated positive domain, the physical contact normalizers still make the two order-one detector coefficients strictly different. The normalized equality threshold is greater than `2/3`, whereas the actual weight `u0` is less than `rho<2/3`. The complete proof and precise scope are in [contact_alignment_20261117_v76.tex](../../../papers/pole_parameters/contact_alignment_20261117_v76.tex). This excludes coefficientwise identification with the unchanged field tests. It does not exclude accidental full-product equality, rational points on either square cover, or a different correspondence between pole surfaces. Fixing this slope also requires a new completeness argument: the original order of prime preparation does not supply it. Run from the repository root: ```sh python3 computation/pole_parameters/general21_alignment_v75/check.py python3 computation/pole_parameters/general21_alignment_v75/check_independent.py ``` The first checker uses exact symbolic identities. The second imports no producer code or symbolic library; integer convolution independently verifies the decisive positive factorization, with exact rational scalar fixtures as supplementary checks. These fixtures are not actual elliptic witnesses, and neither checker claims twenty-one-variable completeness.