24 bit v. 16 bit. More resolution or lower noise?

[Flounder]

I'm wondering. Does 24 bit have higher resolution or lower noise? Or both? Say in the 40db below 0dBFS. Does the wave get represented more accurately? Or the same as 16 bit.

Flounder.

[bap]

Lower noise.

http://www.digido.com/portal/pmodule...34d0f4bfff9ba4

Look under articles and find 'More Bits Please'. There are other worth while things to read as well.

[genericperson]

lower noise, less distortion and better resolution. sample rates actually get noiser at higher rates, so it's a tradeoff. but bits are better at higher rates.

i'm personally much more interested in bits than super high sample rates. i'd do 32 bit/48k if this was some sort of viable reality.

[mpr3]

Originally posted by genericperson
but bits are better at higher rates.

i'm personally much more interested in bits than super high sample rates

Perhaps you are misunderstanding how the word resolution relates to digital audio? Currently, 24 bits has more than enough dynamic range to encompass the entire range of any analog preamp on the planet. I would push for cleaner electron flow technology before 32 bit converters.

[genericperson]

it's not just dynamic range, it's also resolution.

[jabney]

I'd rather record at 32 bits (integer) than at 24 bits just to avoid tossing away or otherwise dealing with the least significant 8 bits in each 32 bit word. In fact, I've read some people's claims that 32 bits is easier for them to record than 24. I'm not aware of any 32 bit converters. Nevertheless many software environments run at rates even higher than 32 bits, so recording at 32 bits can make sense.

It's also easier to combine different bit-rate recordings in the digital realm than it is to combine different sample rates in the digital realm.

[LTA]

Originally posted by genericperson
it's not just dynamic range, it's also resolution.

Well, when somebody makes a 32 bit converter in a new format that doesn't follow the current 6dB a bit philospophy, we'll be able to know. Higher resolution = more discrete volume possibilities and a more precise conversion to digital.

Currently, i feel 24 bit offers more than enough dynamic range for any practical situation short of recording whispering with an sm57 while a space shuttle launch is going on a mile away. During mixing and manipulation, a higher bit rate using either double precision or float formats becomes important, if only to maintain the extra signifigant digits through the final calculation. Think of it as doing a math problem on a 16 signifigant figure calculator that only has an 8 number display. If you write down each individual calculation to 8 digits, and are dealing with 16 digit numbers (like PI, trig functions, logs), you will come out with a slightly different answer than somebody that just worked straight through the problem never having hit clear. Truncation, they call it.