Axiom Home Page Forums General Discussion The Water Cooler hertz, bits, and bitrates - demistifying the devil

Just to add a bit(sorry)more, Ravi, the bit rate is essentially a combination of bits and sampling rate. On CDs the sampling of the frequencies takes place at 44,100 times a second, and as BigBear pointed out, this allows frequencies up to half of that(i.e. 22.05KHz) to be sampled perfectly(it's called the Nyquist theorem). The 16 bits (BInary digiTS) long "words" composed of 1s and 0s can allow for 65,536 different volume levels(2 to the 16th power)for each frequency sampled. So, the sampling measures the frequency and the particular combination of bits needed to represent the volume that frequency had in the original performance taken together result in an exact reproduction of what was originally recorded.

No, 24 bit doesn't necessarily give audibly better sound than 16 bit. Yes, each sample is composed of 24 rather than 16 bits and allows for over 16 million(2 to the 24th power)different sound levels. 16 bits allows for a dynamic range of about 98dB(6.02n + 1.76)while 24 bits would allow about 146dB in theory. However, this actually can't be used and recordings don't have at most more than about a 90dB dynamic range(e.g. the 30dB noise level of a very quiet room up to a 120dB maximum level at orchestral peaks). The quality of sound depends on the original recording technique and the mixing and mastering then applied, rather than whether it winds up on a CD, DVD-A or SACD.