16 bit or 24 bit?

[Sacred One]

is there a major difference, as a matter of fact what is the difference?

 

[TKNK]

basic more bits of audio per cycle are recorded, so a higher bit is a more accurate recording but u only notice the difference if the orignal audio was recorded at 24bits then converted down to 16bits.
Converting 16bits to 24bits theres no noticeable difference other then a larger file

 

[Sin:Aesthetic]

sample at 24 bit...then when you finish mastering the track, as the very last step, dither down to 16 bit....there will be an audible difference from just keeping it at 16 bit

 

[Equality 7-2521]

When signal passes through the A/D converter, the analogue voltage is converted to binary numbers. These binary numbers represent the amplitde of a waveform at any given time. Since there are an infinite number of analogue variations, the digital circuitry quantizes (moves to the nearest) to the bit depth you have chosen. The difference between the analogue signal and the quantized binary number is called quantization distortion. The more bits you have the less distortion because there are more levels of quantization and thus it is more accurate in its rounding. This effectively increases signal to noise ratio. The number of times a binary number is assigned per second is the sample rate (e.g. 44.1).

The difference in bit depth is most audible with low level signals like reverb tails etc and is also audible if the signal is truncated (extra bits are simply discarded) down to 16 bit instead of dithered down.

standard practice these days is to sample at 24 bit and dither (usually with some noise shaping) down to 16 bit for burning to cd.

 

[Architect]

Surely record at 24 bits or higher depending on your preference and what your working with, I record and track at 24 bits/88,200khz. Dither down to 16bits 44.1 when going to CD either using POW-R algorithm or Alesis Masterlink or whatever it is you got available. This is a complex subject it can get confusing.

 

[afriquedeluxe]

When signal passes through the A/D converter, the analogue voltage is converted to binary numbers. These binary numbers represent the amplitde of a waveform at any given time. Since there are an infinite number of analogue variations, the digital circuitry quantizes (moves to the nearest) to the bit depth you have chosen. The difference between the analogue signal and the quantized binary number is called quantization distortion. The more bits you have the less distortion because there are more levels of quantization and thus it is more accurate in its rounding. This effectively increases signal to noise ratio. The number of times a binary number is assigned per second is the sample rate (e.g. 44.1).

The difference in bit depth is most audible with low level signals like reverb tails etc and is also audible if the signal is truncated (extra bits are simply discarded) down to 16 bit instead of dithered down.

standard practice these days is to sample at 24 bit and dither (usually with some noise shaping) down to 16 bit for burning to cd.

very nicely explained, thanks for that. ive always wondered about what the true benefit was, a higher signal to noise ratio besides the clearer singal to begin with.