32 DAW x 24 bit audio samples

On 2011-05-21 16:25, monno wrote:
24 Bit is the "de facto" industry standard.
You already have ample dynamic headroom and very, very low noise floor with 24 Bit.
The reason why most programs do their calculations in 32 Bit float, is to keep some additional mathematical headroom if you will, to ensure a minimum of truncation and consequent dithering when bouncing down even to 24 bit (Adding many streams of audio results in longer wordlenghts bigger than 24 bit and working with 32 Bit files would theoretically exacerbate this issue)

What can be theoretically gained might just as easily be negated. 32 Bit files are larger and takes up more space with no proved advantage. You are better off working at higher sampling rates as the extra space taken by such a move will actually benefit you more than going from 24 to 32 bit.

Additionally i might add, that i always frown upon people sending me 32 bit files for mastering, as it involves one extra step of conversion (= unnecesary processing with potential loss of definition) to load such files into my system (which operates strictly on the accepted standards for reasons of compatibility)

On 2011-05-21 20:29, PoM wrote:
the bitrate in the settings are for the file you record or bounce ,all daw i know work are 32/64 bit float or 48 bit.
for daw like cubase that work at 32 bit float exporting files at 32 bit float avoid multiple convertion when you bounce file like synth and other stuff.
for final rendering 24 bit is the standar but 32 bit float is maybe marginally better if the mastering engeinner use a software that can open 32 bit float files.(for avoiding 32>24>32 )