The default Ducking attack speed and placement makes it hard to hear the first word of the voice-over sentence. I would like to control the attack so that the background music (a) fades out more quickly; and (b) starts fading slightly before the transition, so it’s already halfway down before the voice-over starts.

Not being able to do this, I am manually fading every audio sample in and out, and for a music podcast with sometimes 100 examples per episode, this is hours of work.