To my ears it sounds like the reverb reacts more when the singing is louder. I like how the vocals are "breathing", in quieter parts intimate and sort of dry, in louder parts airy and open.
Maybe there is some compressor placed in reverb b̶u̶s̶ send, that controls decay/mix depending on volume of incoming signal?
I have no idea, but I love it.

 

Last edited: Jun 2, 2024

I'd put the reverb on a dedicated send and write automation as needed/desired (like for send levels). Or automate levels between two different reverbs. IMO that's more interesting for the listener than hearing fully automatic modulation based on input levels.