65
$\begingroup$

I read in some places that music is mostly sampled at 44.1 kHz whereas we can only hear up to 20 kHz. Why is it?

$\endgroup$
7
  • 1
    $\begingroup$ Younger people can hear higher frequencies. Other recording techniques use up to 48 kHz. $\endgroup$ Mar 6, 2017 at 13:42
  • 20
    $\begingroup$ Nyquist theorem : you need two samples every swing to tell the frequency of a wave. $\endgroup$ Mar 6, 2017 at 20:43
  • 1
    $\begingroup$ Because processors are faster, memory is cheap, but good analog filters are still tricky, even higher sample rates can make sense as well (96 or 192 kHz) $\endgroup$
    – Nick T
    Mar 6, 2017 at 22:25
  • 3
    $\begingroup$ @ThorbjørnRavnAndersen I think 48 kHz is common because it's divisible into 24, 25, and 30 fps used in video production. 24 doesn't evenly go into 44100. That's what Wikipedia mentions. $\endgroup$
    – Nick T
    Mar 6, 2017 at 22:28
  • 7
    $\begingroup$ @SohamDe This is because if you sample a 20 kHz audio signal at exactly 20 kHz, you would hear nothing at all. Picture it, a sine wave that peaks every 1 / 20,000 second. Well if you sample that at the exact same rate, then you would only sample the peaks (or nodes, or whatever level you happen to sample it at). So when you recreate the signal from digital, all you get is a flat line. This concept is called aliasing and it makes it so that you must sample at at least twice the maximum frequency you want to be able to hear. 44 100 Hz is convenient because is divisible by a power of 2. $\endgroup$
    – MichaelK
    Mar 8, 2017 at 10:44

6 Answers 6

96
$\begingroup$
  1. The sampling rate of a real signal needs to be greater than twice the signal bandwidth. Audio practically starts at 0 Hz, so the highest frequency present in audio recorded at 44.1 kHz is 22.05 kHz (22.05 kHz bandwidth).
  2. Perfect brickwall filters are mathematically impossible, so we can't just perfectly cut off frequencies above 20 kHz. The extra 2 kHz is for the roll-off of the filters; it's "wiggle room" in which the audio can alias due to imperfect filters, but we can't hear it.
  3. The specific value of 44.1 kHz was compatible with both PAL and NTSC video frame rates used at the time.

Note that the rationale is published in many places: Wikipedia: Why 44.1 kHz?

$\endgroup$
17
  • 11
    $\begingroup$ Hi, I really agree with your answer, but the "..twice the highest frequency" thing bites beginners very soon, because Nyquist is about bandwidth, not highest frequency; I went ahead and slightly modified your answer. Please check if it's OK with you. $\endgroup$ Mar 5, 2017 at 6:28
  • 2
    $\begingroup$ @Ruslan: Wikipedia is quite good about it. $\endgroup$
    – jojeck
    Mar 5, 2017 at 11:58
  • 2
    $\begingroup$ @BrianDrummond So edit it? $\endgroup$
    – endolith
    Mar 5, 2017 at 17:20
  • 4
    $\begingroup$ @MarcusMüller the beginner who's bitten by “Nyqvist is highest allowed frequency” will get bitten anyway, by aliasing artifacts... After that, they'll also understand how any range of frequencies of bandwidth $Δf$ is demodulated to one between $0$ and $Δf = f_s/2$. $\endgroup$ Mar 5, 2017 at 22:12
  • 2
    $\begingroup$ @supercat even in those cases, if it were possible the brick wall filter would still be desirable as the sampling filter. it's job is just to avoid aliasing. what you're talking about is the job of an equalization filter $\endgroup$
    – Steve Cox
    Mar 6, 2017 at 0:57
73
$\begingroup$

44,100 was chosen by Sony because it is the product of the squares of the first four prime numbers. This makes it divisible by many other whole numbers, which is a useful property in digital sampling.

44100 = 2^2 * 3^2 * 5^2 * 7^2

As you've noticed, 44100 is also just above the limit of human hearing doubled. The just above part gives the filters some leeway, therefore making them less expensive (less chips rejected).

As Russell points out in the comments, the divisible by many other whole numbers aspect had an immediate benefit at the time the sample rate was chosen. Early digital audio was recorded on existing analog video recording media which supported, depending on region, either the NTSC or PAL video spec. NTSC and PAL had different Lines per Field and Fields per Second rates, the LCM of which (together with the Samples per Line) is 44100.

$\endgroup$
6
  • 12
    $\begingroup$ The choice wasn't simply about getting many prime factors, but specifically to make good use of NTSC and PAL video recording equipment to store digital masters. en.wikipedia.org/wiki/44,100_Hz#Recording_on_video_equipment $\endgroup$ Mar 6, 2017 at 0:59
  • 3
    $\begingroup$ @RussellBorogove: Thank you. As per the Wiki link, 44100 is the LCM of the sample rates of the NTSC and PAL video features' rates . That is quite a direct consequence of being a number with so many factors, and I do believe that you are right that the horse led the cart on this spec. $\endgroup$
    – dotancohen
    Mar 6, 2017 at 6:32
  • 2
    $\begingroup$ Divisible by many numbers, but not by 8 :) $\endgroup$ Mar 6, 2017 at 16:02
  • $\begingroup$ (Wikipedia says a variety of rates from 40.5 to 46.8 kHz would have met these criteria, and 44.1 kHz was chosen to provide a transition band for antiliasing filter) $\endgroup$
    – endolith
    Mar 7, 2017 at 19:12
  • 3
    $\begingroup$ @BogdanAlexandru Also not divisible by 1 ms USB frames :D $\endgroup$
    – endolith
    Mar 8, 2017 at 20:15
13
$\begingroup$

The Nyquist rate is above twice the bandlimit of a baseband signal that you want to capture without ambiguity (e.g. aliasing).

Sample at a lower rate than twice 20kHz, and you won't be able to tell the difference between very high and very low frequencies just from looking at the samples, due to aliasing.

Added: Note that any finite length signal has infinite support in the frequency domain, thus is not strictly bandlimited. This is yet another reason why sampling any non-infinite audio source a bit above twice the highest frequency spectra (in a baseband signal) is required to avoid significant aliasing (beyond just reasons of finite filter transition roll-off).

$\endgroup$
4
  • $\begingroup$ Hi, I really agree with your answer, but the "..twice the highest frequency" thing bites beginners very soon, because Nyquist is about bandwidth, not highest frequency; I went ahead and slightly modified your answer. Please check if it's OK with you. $\endgroup$ Mar 5, 2017 at 6:29
  • 6
    $\begingroup$ @MarcusMüller, because "beginners" to sampling start with sampling baseband signals and not passband signals, it really is about the highest frequency (sometimes called "bandlimit") and not bandwidth (which has an additional ambiguity regarding one-sided or two-sided bandwidth). $\endgroup$ Mar 5, 2017 at 6:43
  • $\begingroup$ @robertbristow-johnson haven't looked at that ambiguity. Hm; I like the bandlimit approach! $\endgroup$ Mar 5, 2017 at 6:46
  • 3
    $\begingroup$ in the Wikipedia article we call it "$B$" and, although Shannon said $f_\text{s} \ge 2B$ is sufficient, he was assuming finite energy, so no sinusoids (which have infinite energy and can also put dirac deltas at $\pm B$). if you allow for a sinusoid right at frequency $B$, then it's the more oft-stated $f_\text{s}>2B$ . $\endgroup$ Mar 5, 2017 at 6:54
10
$\begingroup$

Basically, twice the bandwidth is a common requirement for signal sampling, thus $2\times 20 = 40$ kHz is a minimum. Then, a little more is useful to cope with imperfect filtering and quantization. Details follow.

What you need in theory is not what is required in practice. This goes along the quote (attributed to many):

In theory there is no difference between theory and practice. In practice there is.

I am not an expert in audio, but I have been trained by high-quality audio sampling/compression people. My knowledge might be rusty, take it with caution.

First, standard sampling theory works under some assumptions: linear systems, and time invariance. Then, a continuous bandlimited phenomenon is known, in theory, to be possibly sampled at about twice the bandwidth (or twice the maximum frequency for baseband signals) without loss. The "Nyquist rate" is often defined as:

the minimum rate at which a signal can be sampled without introducing errors

This is the analysis part of the "sampling theorem". The "can be" is important. There is a synthesis part: the continuous signal "can be reconstructed" analogously using cardinal sines. This is not the only technique, and it does not take into account low-pass prefiltering, non-linear (such as quantization, saturation) and other time-variant factors.

Human hearing is not a simple topic. It is accepted that humans hear frequencies from 20 Hz up to 20,000 Hz. But such precise bounds in Hertz are not a trait of nature for all humans. A gradual loss of sensitivity to higher frequencies is frequent with age. On the other side:

Under ideal laboratory conditions, humans can hear sound as low as 12 Hz and as high as 28 kHz, though the threshold increases sharply at 15 kHz in adults

Hearing is not linear: there are audition and suffering thresholds. It is not time-invariant. There are masking effects in both time and frequency.

If the 20 Hz up to 20,000 Hz band is a common range, and a 40,000 Hz should theoretically suffice, a little extra is needed to cope with extra distortion. A rule of thumb says that 10% more is ok ($2.2\times$ signal bandwidth) and 44,100 Hz just does it. It goes back to the late 1970s. Why is not 44,000 Hz used? Mainly because of standards, set by the popularity of CDs, whose technology is as always based on a trade-off. In addition, 44,100 is the product of squares of first four prime numbers ($2^2 \times 3^2 \times 5^2 \times 7^2$), hence has small factors, beneficial for computations (like FFT).

So from $2\times 20 $ to $44.1$ (and multiples), we have a balance in safety, quantization, usability, computations and standards.

Other options exist: the DAT format for instance was released with 48 kHz sampling, with initially difficult conversion. 96 kHz is discussed with respect to quantization (or bit depth) in What sample rate and bit depth should I use? This is a controversial subject, see 24 bit 48kHz verses 24 bit 96kHz. You can check Audacity sample rates for instance.

$\endgroup$
3
  • 2
    $\begingroup$ 1. The answer to the question is that the Nyquist theorem dictates > 40kHz, not > 20kHz. 2. Neither human hearing nor the CD format is limited to 20Hz at the low end. Any large enough pipe organ can produce a 16Hz tone, and CD can reproduce it easily. Some organs go down to 8Hz, which starts to be perceived as individual vibrations, but which again CD can reproduce. $\endgroup$
    – user207421
    Mar 5, 2017 at 22:20
  • $\begingroup$ I do agree with your comment, except for "dictates" (this is an "if" condition). Could you point out where I have deviated from it? $\endgroup$ Mar 6, 2017 at 7:40
  • 1
    $\begingroup$ I have just one supplement to @LaurentDuval 's answer. Speech, music, and sound in general are non-stationary signals. Although these are effectively bandlimited but we yet do not know how the human ear is transducing the continuous time signal to nerve firings which facilitate our perception of sound. It is often argued that some people have "golden ears" and can make out difference between 44.1 kHz versus 96 kHz recordings. Also, I am yet to confirm on the following, it seems higher sampling rates benefits perception of additional cues, such as localization in binaural recordings. $\endgroup$
    – Neeks
    Mar 7, 2017 at 11:49
0
$\begingroup$

Why it is exactly 44.1 kHz has been already answered - but to focus on the aspect of your question related to the limit of human perception, the reason is quite simple.

The resolution in time has to be fine enough to be able to generate all possible wave forms up to the limit that is perceptible. According to the sampling theorem, the resolution has to be such that the sampling frequency is at least twice this frequency. Intuitively, at the highest frequency, you need at least 2 points to represent the maximum and the minimum of your signal - giving this Ascii-art square wave:

_   _
 |_| |_
$\endgroup$
-2
$\begingroup$

In order to faithfully reproduce a signal, the faster the sample rate the better. ~40 kHz was chosen, because it was a low sample rate that most people can't tell the difference for (when reconstructed). When audio sampling was introduced, memory and storage was expensive and higher sample rates were not cheaply possible.

At double the upper limit of human hearing two samples per cycle is very poor reconstruction, even if it meets the Nyquist criteria for sampling signals, a simple chart depicting a sine wave with two samples per cycle will show you how poor two samples per cycle is in reproducing a waveform. You can literally turn a sine wave into square wave; it is a good thing at 20 kHz nobody can tell. I bet a dog could though.

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.