Dithering + Bit depth

[Taylor Livingston]

I'm a little confused about this subject. I have the ability to record in 24 bit (and possibly 32 bit float). Should I? Is there an advantage to recording at a higher bit depth when it will have to be converted to 16 bit eventually when it becomes a cd? When should I dither?

[hunta]

Originally Posted by conical johnson

I'm a little confused about this subject. I have the ability to record in 24 bit (and possibly 32 bit float). Should I? Is there an advantage to recording at a higher bit depth when it will have to be converted to 16 bit eventually when it becomes a cd? When should I dither?

Record in 24 bit. Or 32 bit if you want, but 24 bit is good. I'm not real good with the technical mumbo jumbo, but here's a link: (http://www.tweakheadz.com/16_vs_24_bit_audio.htm). That article doesn't go into 32 bit but it addresses the basic concepts. There is an advantage to recording in higher bit depth, although the exact science of why is debatable. Read up on the Nyquist theory.

If you record in 24 bit, you want to dither when you reduce bit depth down to 16 bit (or from 32 bit down to 16 bit). This is usually done in the mastering process. I do it in wavelab. Don't dither more than once, and it should be the absolute last thing you do, right before you burn your disk.

Search the gearslutz forums for 16 bit vs. 24 bit or 24 bit vs. 32 bit if you want to read much more scientifically minded folks than me discuss it

[lemur821]

Originally Posted by conical johnson

I'm a little confused about this subject. I have the ability to record in 24 bit (and possibly 32 bit float). Should I? Is there an advantage to recording at a higher bit depth when it will have to be converted to 16 bit eventually when it becomes a cd?

If you plan on doing any mixing, overdubbing, or even gain adjustments, 24-bit or higher is a must.

[warnergt]

There is an advantage to recording more than 16 bits. Any processing (mixing, dubbing, normalizing, effects) you perform prior to creating your 16-bit end result is going to add artifacts. Starting this process with more bits up front will reduce or eliminate the artifacts on your end product.

[DocBop]

Basically 24-bits gives you more headroom/larger dynamic range which should result in a cleaner recording. In general you should use 24-bit as your default. Higher than 24-bit gets used in film sound tracks because they want even more dynamic range for more effect in theaters.