# Carpet star

- 2026-08-28 [Conjecture] The carpet-star decay coefficient is not exactly `-1/8`: it depends on the registration and band-normalization frame, the family frame through `L = 400` giving slope `-0.1242` and the ideal frame through the same `L = 400` giving `-0.18` to `-0.19` and still drifting; only the `(log L)/L` decay order survives the frame change.