32-bit Float & AI NR: Has your on-set gain approach changed?

Hey everyone,

I'm Erik Lindström, a sound mixer usually working on commercials and narrative shorts. With 32-bit float now pretty standard on my MixPre-6 II, and seeing what folks can do in post with iZotope RX, I'm genuinely curious if your on-set approach to gain staging has shifted.

I've been on a project recently with a C70 and a few PavoTube II 30X lights, where I experimented with being a bit more 'relaxed' with my input gains, trusting the 32-bit headroom. For talent, I'm still using my 4017 and 6060 lavaliers as usual. While post has loved the flexibility, I'm wondering if I'm creating bad habits.

Are you still meticulously riding gains as much as before? Or are you finding that 32-bit float and powerful AI tools in post are letting you focus more on other aspects of the mix, like mic placement or performance? What's your experience?