Filter Curve Converts 16-Bit to 32-Bit Float

I’ve noticed that applying the filter curve on a 16-bit imported WAV file converts it to 32-bit float. Is there a way to prevent that and keep it at 16-bit for ACX when applying the filter curve? I’m using version 3.0.4 on a MacBook Pro, running OS 11.5.2. Thanks!

Yes there is, but it’s better to work in 32-bit float format from start to end.

Audacity works internally in ultra-high precision 32-bit float format. If you force Audacity to convert to 16-bit every time you edit or process, there will be a conversion from 16-bit to 32-bit float, then back to 16-bit. Better to just convert to 16-bit at the end.

ACX requires MP3 format, and that’s not really 16-bit or 32-bit - it’s encoded in a different way which uses “bits per second” rather than “bits per sample”. Converting from 32-bit float to MP3 is better than converting from 16-bit to MP3.