After rendering and bringing back in ... 24bit or 32bit?

Hi - I’m a bit confused about whether the 32-bit floating calculations done by Cubase mean that I should choose 32 bits when I have it place it at the bottom of the project, or 24 bit (which is actually what I recorded using).

Any thoughts are welcome -


I would leave it at 24bit, Cubase will do all it’s internal stuff at 32.

Thanks much, Mr. Beer. That’s what I’ll do.

Parenthetically, who records at 32-bit, I wonder …

It’s pointless to record at 32-bit. There’s no 32-bit A/D converters (well, there is, but they are not widely available) and even the best 24-bit converters “only” produce something like 21 to 22-bits of significant data (last 2-3 bits are just random noise).

However, if you bounce/process/render your audio dozens of times in a row you may get some advantage from keeping everything in 32-bit (floating point) domain, because accumulated quatization errors of 24-bit rendering may become audible at some point if you have some highly sensitive material during low volume parts of your music.

Got it, thanks.

More disk space needed also, 1/3 more if going from 24- to 32-bit word length, does it work that way?