What is the significance and origin of 44,100 Hz?

[ulillillia]

Sound cards natively support the same sample rates as MP3s support - 8000, 11,025 (weird), 12,000, 16,000, 22,050 (weird), 24,000, 32,000, 44,100 (weird), and 48,000 Hz. Some sound cards extend this pattern clear out to 192,000 Hz. My question is, what is the significance and origin of the rather weird 44,100 Hz sample rate (which derives from 11,025 and 22,050). Why aren't sample rates powers of 2 like computers tend to use for all data sizes (such as 65,536 Hz or 16,384 Hz))?

 

[hunter315]

FM radio tends to sample at 22050 Hz, the other two are derivates of that, most audio CDs are recorded at a sampling rate of 44.1 kHz so thats where those two come from now, im not sure how they were picked originally.


As for why audio sampling rates arent even multiples of 2, its for convience and due to sampling theory, you need to sample something from analog to digital at twice the highest frequency or the high frequency components look like lower frequency components and distort the true signal. Most voice is less than 4 kHz so 8 kHz is a nice low bandwidth sample rate that isnt going to clip voice conversations, the higher ones are for better quality audio, higher sample rates give you a smoother signal which is needed for higher quality to be maintained, but for most things 40k and up is more than sufficient and can be smoothed well enough by a simple low pass filter that you wont be able to tell it got sampled through most speakers.

 

[ulillillia]

Not multiples of 2, powers of 2 like 2, 4, 8, 16, 32, 64, etc. Multiples of 2 are expressed as 2x and powers of 2 are expressed as 2^x (or "pow(2.0, exponent)" in C programming where exponent doesn't have a fractional value). 48,000, 32,000, 44,100, etc. are not powers of 2.

I know computer audio very well, I was just wondering on the origins and significance of the rather odd 44,100 Hz sampling rate. If 20,000 Hz is the limit of human hearing, why isn't it 40,000 Hz for the sampling rate? Of course, a higher sampling rate reproduces high-pitched audio much better. 16,000 Hz, the supposed cap of MP3 files, requires 3 samples to reproduce at 48,000 Hz, but 12 at 192,000 Hz. 12 shows a fairly nice sine curve but 3 shows a very jagged shape (at 0, +0.866, and -0.866 repeatedly).

 

[pepe2907]

There was an explanation in Wikipedia. As I remember the specific rate was chosen initially by Sony for some technical reasons /besides the fact, that the maximum frequency, heard by an average person is considered to be ~20 kHz/.

 

[ulillillia]

Best answer selected by ulillillia.

 

[JAYDEEJOHN]

This topic has been closed by Jaydeejohn