Requantise to 24-bit
Drop an audio file and export a WAV whose samples sit on the 24-bit integer grid and whose header declares that width. Triangular dither is applied before rounding.
The working depth
24-bit is where mixing and mastering happen. The reason is accumulation: every gain stage, every filter, every summing operation rounds its result, and those small errors add up across dozens of processes.
At 16 bits that accumulation can eventually reach the point of mattering. At 24 bits the theoretical floor sits around -146 dB, so the errors stay buried no matter how long the chain is.
Headroom without penalty
Recording at 24 bits means you can leave 18 or 20 dB of headroom above the loudest expected peak and still have a noise floor far below anything the microphone or preamp contributes.
At 16 bits the same caution costs real resolution, which is why engineers working to tape-era habits used to track much closer to the ceiling. The extra depth removes the tradeoff entirely.
Where the real floor sits
-146 dB is theoretical. No microphone, preamp, or converter approaches it; their own self-noise sits far higher. The format stops being the limiting factor well before the equipment does, which is the point.
Archive and delivery
Many delivery specifications ask for 24-bit masters, and archives generally prefer them. Keeping the extra depth means a future remaster starts from the same resolution the mix was finished at rather than from a file already reduced for distribution.
Not the depth for CD
CD and most consumer delivery require 16-bit. Convert to 16-bit as the final step, once all processing is complete, so the dither is applied only once. For 8-bit lo-fi or to change the dither and shaping, use the full Bit Depth Converter.