DeVerberate 3.

DeVerberate 3 turns down, or up, the reverb already baked into a recording. On dialogue, a neural network trained on thousands of voice recordings and many different rooms separates speech from reverb without manual tuning, adapting as the reverb changes. On music and other material, a manual mode splits the signal into direct and reverberant parts, so you decide how loud the room stays and how long it rings. An Early Reflections Filter, available in either mode, targets the resonances that small rooms leave behind.