Am I Losing Sound Quality If My Gear Isn’t Spec’d to 192 kHz?

Got a tech question for Sound & Vision? Email us at AskSandV@gmail.com

Q My Sony Blu-ray player can play DTS-HD Master Audio soundtracks with up to 192-kilohertz/24-bit resolution. However, my receiver’s specified bandwidth is 20 Hz to 20 kHz, and my Polk 330T speakers only cover a frequency range of 33 Hz to 25 kHz. Am I losing sound quality because the receiver and the speakers aren’t spec’d to handle frequencies up to 192 kHz? — Javier Feliciano / via email

A Quick answer, no, but let’s step back and walk through your question. Audio tracks with sampling rates higher than the CD format’s 44.1 kHz are available on Blu-ray Disc and from high-resolution music download sites like HDtracks.com. But high sampling rates don’t necessarily translate into useful sonic information that the human auditory system can process.

In signal processing, an operating principle called the Nyquist Theorem specifies the minimum sampling rate required to adequately capture an analog waveform without introducing a particularly nasty-sounding form of distortion known as aliasing. For digital audio, that rate is usually 44.1 kHz, which translates to a Nyquist Frequency of 22.05 kHz—enough to cover the full range of frequencies the human ear can hear, which is generally cited as 20 Hz to 20 kHz.

Meanwhile, that 20-Hz-to-20-kHz spec you’re seeing for your receiver has nothing to do with its frequency response. Rather, it’s the audio band used by the manufacturer to measure and specify amplifier output power. (Receiver frequency response specs typically extend into the 50-kHz range.) However, the 33-Hz-to-25-kHz range you cite for your speakers in all likelihood is its manufacturer supplied frequency response.

So, to return to the question of whether you’re missing out on sound quality because your receiver and speakers aren’t spec’d to handle frequencies up to 192 kHz, the answer is no because your ears aren’t capable of hearing frequencies higher than 20 kHz. The main benefit to higher-than-44.1-kHz sampling rates in digital audio is to extend the Nyquist Frequency well outside the human hearing spectrum—so far out that there isn’t even a remote possibility of anomalies associated with the required anti-aliasing filters creeping into the audio band. While many might say that makes hi-res audio overkill, some astute listeners maintain they can hear a quality difference when higher sampling rates are used.

COMMENTS
Old Ben's picture

You wrote: "The main benefit to higher-than-44.1-kHz sampling rates in digital audio is to extend the Nyquist Frequency well outside the human hearing spectrum—so far out that there isn’t even a remote possibility of anomalies associated with the required anti-aliasing filters creeping into the audio band. While many might say that makes hi-res audio overkill, some astute listeners maintain they can hear a quality difference when higher sampling rates are used."

I think another part of the explanation is that these higher frequencies portions of the signal could interact - constructively and/or destructively - to result in additional audible signals.

thehun's picture

Keep reading about this theory all the time, how the brain "responds" to ultrasonic signals under some laboratory conditions. It neither proves anything nor it is useful for actual music listening that has nothing in those upper registers but noise, and even that's like 100db below the signal itself.

Old Ben's picture

I'm not talking about how the ear/brain responds to ultrasonic signals. Rather, I am talking about how two different ultrasonic frequencies can interact constructively or destructively, resulting in "beats" with a frequency in the audible range. If the higher sampling rate information is stored in the data and reproduced, that higher frequency data can interact with other frequencies to result in additional audible signals that would be "lost" at lower sampling rates.

thehun's picture

But if it's ultrasonic how could that be "audible" ?

Old Ben's picture

Sound waves - sonic or ultrasonic - are pressure waves. If two difference sound waves interact, they add or subtract from one another. If two sound waves have different frequencies, then the amount of addition or subtraction with respect to one another changes over time (as a separate frequency). An ultrasonic sound wave can interact with another, different ultrasonic sound wave such that the resulting additive/substractive oscillation has a frequency within the audible range.

thehun's picture

Then you shouldn't have any issues to name a recording that this is happening right?

Old Ben's picture

I cannot name a recording in which this happens. It's not like there's a label on a high res track saying "this recording includes audible interactions between ultrasonic frequencies." I simply raised a point that is valid. Simply put, if any audio recording is reproducing ultrasonic sounds, the ultrasonic frequencies may interact with other frequencies (including other ultrasonic frequencies) to result in an audible interference pattern.

thehun's picture

Of course you can't since your entire argument was theoretical and only concern nerds and frankly irrelevant to the original question on hand.

Old Ben's picture

I really don't know why I'm even responding to you at all - I'm not sure if you're being nasty because you're a deeply unhappy person or just a troll. Either way, I'm done after this. First, I'm not making an argument; I am asking a question. That question is whether the very well known property of sound waves adding to or subtracting from each other possibly leads to audible differences in a recording when that recording has a sampling rate high enough to include ultrasonic frequencies. Specifically, the question is whether the interactions of the ultrasonic frequencies with other frequencies could result in an interference pattern that has an audible component. Physics and math both say this is possible. Second, to the extent my question only concerns nerds, this is a site for audio enthusiasts, so audio "nerds" or audio "geeks" would be applicable to the audience here. Third, a recurring debate on this site and other similar sites is whether high res audio really leads to discernable differences. The question is related to sound quality and sampling rate. Therefore, my question is highly relevant to the original question.

HMB's picture

The purpose of making the Nyquist frequency high is to make the anti-aliasing filter before the ADC have more linear phase. A 192 KHz sampling frequency has a Nyquist of 96 KHz. Any frequency component above 96 KHz will be sampled in such as way as to fold-back to below 192 KHz - just like a tire rotation seems to go backwards or change direction on TV - that is an example of aliasing. The filter requirement would be to attenuate by 96 dB everying above 96 KHz. It is easier to do this if you start way below 96 KHz - the more the better, but you cant go below the audio range or else you eliminate the audio. Anyway - the greater the distance between 20 KHz and Nyquist, the easier your filter is to design and implement, and the more linear the phase will be in the audio band - better audio transient response will result.

X