Hey!

I recorded some vocals that are really quiet via my Apollo Twin X Gen 2 and a SM7dB. The Noisefloor is below -60dB.
In the gain staging and processing the noisefloor rises to a audible level though.
If I wanted to remove the noise fully I tried to use Izotope RX but that also removed some of the „whisperness“ from the vocal… Any Idea how Finneas from Billie Eilish got the vocals so clean without deleting the breath? I also tried this on a Gefell UM70 (Vintage equivalent to a U87) and with extra whisper takes…
Those extra whisper layers seemed to help a little but not that much…

Do I just accept, that I need to use less compression with those quiet vocals?

I would really love some recommendations and help!

Thanks in advance and have a great day!
-Z0RY

when recording a whisper voice, you'll want to stick to the mic as close as possible, use a ldc with very high output and very low self noise (such as the tlm-103 which for this particular application i find very useful, much less for conventional applications) and maybe also a high quality preamp.

in terms of processing, normalize the track and use a comp/lim for bringing up the the level and an expander (not a gate!) for ducking a bit of noise before you resort to rx or similar tools.

oh, and room in which you're recording better be very quiet...