"We can't hear frequencies over ____Hz"

[MogwaiBoy]

I see this mentioned all the time - but I don't get the science behind it.

How come.. if I take a sinewave and pitch it upwards, it disappears (becomes inaudible) as it heads over 15kHz.

Yet I can add the Maag EQ's 40kHz or Museq's 35kHz boost to a snare, high-hat or full mix - and clearly hear the additional high end when A/B'd, even though 40kHz is much higher?

Is it a cumulative/decibel thing? So in actual fact we can hear these high frequencies?

I tamed some harshness in a mix with NI Solid Mix's lowpass filter at 35kHz.. I'm just confused because shouldn't we be able to hear anything up there anyway?

I just got my kids to do a test and they can hear a sinewave tone at 17.5kHz that people over 18 supposedly can't hear. I can hear it but it's not loud or annoying. According to my wife (over 40) it's loud for her.

I'm curious as to how this all fits together.

[Vospi]

What was your sound card working sampling frequency (44.1kHz? 48kHz?) when you did this absolutely nonsensical sound tweaks?

All these 40, 35 kHz "EQ" stuff — you're not supposed to do any of that things, ever. I've already used the best word to describe it in terms of audio production and composition: they are nonsence, you're just getting stuck at something that's not useful at all.

Answering your question. You (probably...) can hear A/B difference 1) because no EQ works at "single frequency" (and do you think you really have recorded something valuable at 40kHz in the first place?.. please don't) 2) because maths behind your DAW/plugins or physics behind your soundcard's D/A converter just gets wrong results; there are even test WAV files to see if your soundcard does wrong things at 22+ kHz, and a lot of cards fail the test.

Also, make sure you're aware of what Nyquist frequency is. That, combined with an answer to my first question, should dissolve a lot of wrong concepts.

[ATN69]

First of all, the hearing threshold is individual and all though age matters it is not the only factor. Probably what you hear when you boost anything above, lets say 20kHz, are the effects of the lower side tones increasing, i.e. tones within your normal hearing range. The normal 20Hz to 20kHz as stated is based on an old study and unfortunately, these days I would say that the upper threshold is coming down. Recent studies show that young people today are more exposed to loud sounds. The need for hearing aid devices is on the rise.

I know because I work as an acoustic engineer in the mobile phone industry and we make hearing aid compatible phones. The demand is high. So if you have kids you should remind them to take good care of their ears and their hearing because it will benefit them later in life.

[MogwaiBoy]

What was your sound card working sampling frequency (44.1kHz? 48kHz?) when you did this absolutely nonsensical sound tweaks?

All these 40, 35 kHz "EQ" stuff — you're not supposed to do any of that things, ever. I've already used the best word to describe it in terms of audio production and composition: they are nonsence, you're just getting stuck at something that's not useful at all.

Answering your question. You (probably...) can hear A/B difference 1) because no EQ works at "single frequency" (and do you think you really have recorded something valuable at 40kHz in the first place?.. please don't) 2) because maths behind your DAW/plugins or physics behind your soundcard's D/A converter just gets wrong results; there are even test WAV files to see if your soundcard does wrong things at 22+ kHz, and a lot of cards fail the test.

Also, make sure you're aware of what Nyquist frequency is. That, combined with an answer to my first question, should dissolve a lot of wrong concepts.

Not sure what you mean by my supposedly "absolutely nonsensical", "wrong concepts" - I haven't made any claims about correct/incorrect mixing and EQ methods. I'm not saying high end "air" boosting is a good thing or a bad thing. This is simply a curiosity about sound frequencies and perceptual differences between a pure sinewave tone and full mixes.

Speaking of tones, I hear the tone of your post loud and clear

I appreciate the input though

[Danno]

Regarding the change you hear when you EQ.. Keep in mind there's always a curve around the freq that you're boosting/cutting!

[MogwaiBoy]

Yep! I get that. Everything around the centre frequency also gets boosted, depending on the "Q" value. So if you boost at 20kHz with a wide band you are boosting lower frequencies too, that are also more audible.

Is that it though.. case closed?

[Michael L]

Is that it though.. case closed?

Here is a Wikipedia post on hearing ultrasound (which is your actual question) that can link to further investigation:
https://en.wikipedia.org/wiki/Ultrasonic_hearing
It says:

Ultrasonic hearing is a recognised auditory effect which allows humans to perceive sounds of a much higher frequency than would ordinarily be audible using the physical inner ear, usually by stimulation of the base of the cochlea through bone conduction. Human hearing is recognised as having an upper bound around 17-20 kHz, depending on the person, but ultrasonic sinusoids as high as 120 kHz have been reported as successfully perceived.

Two competing theories are proposed to explain this effect. The first asserts that ultrasonic sounds excite the inner hair cells of the cochlea basal turn, which are responsive to high frequency sounds.[1] The second proposes that ultrasonic signals resonate the brain and are modulated down to frequencies that the cochlea can then detect.[2]

Researchers Tsutomu Oohashi et al. have coined the term hypersonic effect[3] to describe the results of their controversial study supporting audibility of ultrasonics.

By modulating speech signals onto an ultrasonic carrier, intelligible speech has also been perceived with a high degree of clarity, especially in areas of high ambient noise. Deatherage states that what humans experience as ultrasonic perception may have been a necessary precursor in the evolution of echolocation in marine mammals.[4]

And speaking of which (from Wikipedia):

Animal_hearing_frequency_range.svg.png

And finally, a chapter on human hearing (in the context of a great tech book on DSP) that addresses the important perceptual relationship between Hz and dB:
http://www.dspguide.com/ch22/1.htm

[zethus909]

soundwaves are pressure variations in the atmosphere. so you can feel sound as much as hear it.....the same way you feel subs, you feel those ultra high freqs, and they all play into the realness and crispness of a sound

[MadBrain]

As outlined earlier, EQs aren't limited to a single frequency and affect a whole range. Especially with a very large bandwidth / very low Q, a 35khz cut is still rolling down 15khz a few dB's.

In addition to that, EQs behave weird near the sampling rate frequency. If your project has a 44khz sample rate, EQs that aren't oversampled will behave weird near the 22khz limit, in the 20khz-22khz range in particular. With higher sample rates, this effect kinda goes away though, and for normal usage it isn't very relevant.

[dermichl]

aliasing is one reason. nyquist won't let your average soundcard reproduce that frequencies too good...

[ZentralmassivSound]

Yep! I get that. Everything around the centre frequency also gets boosted, depending on the "Q" value. So if you boost at 20kHz with a wide band you are boosting lower frequencies too, that are also more audible.

Is that it though.. case closed?

Yes that's it, case closed. If you deal with normal mp3s or CD format, your system won't reproduce frequencies higher than 20kHz anyway. You would need direct playback on a soundcard with 96kHz plus an amplifier which doesn't filter above 20kHz plus speakers that reproduce sound in that range.

[xoxos]

modulation in higher frequencies may fall within the audible range eg. the "microwave auditory effect".. can't hear the signal, but changes in the signal, yes.

you may also hear the result of resonances in the cabinetry, i'd think especially on laptops.