Belt tension pass on the printer

Placeholder log entry. Ghosting on the X axis had been getting worse for weeks, and it turned out both belts had drifted to noticeably different tensions.

Re-tensioned each until they plucked at roughly the same pitch, then reran the ringing tower. The artefacts on the leading edges are gone at 60mm/s and much reduced at 100mm/s.

Next: reprint the calibration cube and check whether the dimensional error on X tracked the tension change.