2.1.13. Compensation action

The compensation action button changes behavior based on the current compensation state.

• Learn starts measuring the selected compensation mode and updates a live learned correction.
• Lock stops learning and commits the current learned correction into compensation gain.
• Reset clears the stored compensation gain back to 0.00 dB.

Limiter learning is intended for a learn-then-lock render workflow: run a representative pass, lock the learned correction, then render normally.