Audio 16 bit & 24 bit what’s the deal

[Videomansf]

16bit audio uses 65K levels of volume for ever sample of audio. 24bit audio uses 16 million levels per sample. The steps between samples can be more accurate in 24bits. think of making an audio fade out over 1000 seconds from 0db to -100db. at 16 bits the audio will fade out at 65 volume changes per second, and at 24 bits you would have 1600 changes per second. Now 1000 second fades are not useful but, i am using this to demo a point. You will not get “better” sound capturing 16bit 48K at 24bit 48K, but if you augment the 16 bit signal (say add an eq) in terms of 24bits, that newly created audio will be marginally better than the 16bit eq version. The more you play with the signal, the less the audio will break down in 24bit. Think of it like smooth gradations in video at 8bit vs 10bit. Bit depth can be changed mostly harmlessly, due to the fact that your representing voltage between off and full on or 0 and 1 volt. a value of 74% is the same in both 16bit and 24bit audio, but a value of 74.00423712345% in 24bits might become 74.01% in 16bits. Bits, in other words, are rounded off as bit depth decreases, and will not change as bit depth increases.

In fact, final cut pro processes all audio in 32bits then rounds down to the operating bit depth (16 or 24). Logic 7 works in 64bit audio until processed down. Bit depth will not through you out of sync like sample rate will, and making sure that your sample rates are all the same is much more important. New audio systems even run at 1bit (on or off) at a sample rate of 24MHz, and due to the huge sample rate on and off become almost perfectly smooth waveforms.

VM

[Dave Jenkins]

Thanks, lots of info to absorb.

Dajen Productions
Santa Barbara, CA
G5 Dual 2 Gig – AJA IO & LA
Huge 1.2 Raid
FCP 5-OS X 10.4.1-QT 7