It’s a known issue OP-1f doesn’t play well with Ableton bc of the float issue. Gonna have to export from Ableton then throw it into something like audacity to convert it again.
Yes, there are 32bit files that are not float. You’ll need to convert them to 32bit integer outside of Ableton. I had the same problem with Assimil8or and didn’t realise 32bit integer existed.