Topic: Newbie Question Regarding External A/D

[ghellquist]

No need to worry. The 20 bit AD actually outputs a 24 bit signal already padded with zeros.

As to if you should run 24bit or 16 bits -- my suggestion is to run 24 bits. You get a few extra bits of signal that you might want to use in post to raise signal levels. Keep them and you have the option, throw them away and they are gone forever.

The whole point of 24bits is a bit inflated. No AD for audio use (hear me, none) really works out at 24 bits. It is a lot of marketing BS calling it 24 bits. The limiting factor is generally the analog circuits that add enough noise to effectively make it about 18 bits, perhaps up to 20 or 21 for the really expensive stuff. I even remember testing a unit (Edirol R1) claiming 24 bits that really only had about 15 bits due to high noise levels.

Boiled down to the bottom line, better go with the more honest statement of 20 bits.

// Gunnar

[Scuzzi]

No need to worry. The 20 bit AD actually outputs a 24 bit signal already padded with zeros.

As to if you should run 24bit or 16 bits -- my suggestion is to run 24 bits. You get a few extra bits of signal that you might want to use in post to raise signal levels. Keep them and you have the option, throw them away and they are gone forever.

My question is will I be losing some kind of quality by dithering down to 16bit in Audacity as compared to using the built-in UV22 dithering on the AD-1000.  Is it really the same process or not at all?

[ghellquist]

My question is will I be losing some kind of quality by dithering down to 16bit in Audacity as compared to using the built-in UV22 dithering on the AD-1000.  Is it really the same process or not at all?

You might see gains or losses. I believe that UV22 is considered a good dithering algorithm, what Audacity uses I am not sure of. In post you do have a lot of choices though, I believe r8brain (found for free from voxengo, alas only windows) is a respected choice.

You would definitely gain a lot more than any loss if going more than 16 bits allows you to adjust gain upwards in post. Going 16 bits, you get what you have including all safety margin on volume. Using the extra bits allows you to continue having a safety margin when recording, raising volumes afterwards and still not increase the noise floor level.

Still, a lot of these things are totally swamped by the decisions you make on location on choice of mics and positioning of them. I would not spend too much thought on them, or as the Nike slogan used to say, just do it.

// Gunnar