Shouldn't we upgrade the 20-20 audible range ?!

[kyuu]

That's good for you, but anecdotal. What I'm suggesting is that given good conditions, the sound will actually be much more similar than the data may suggest, despite low SINAD and potential reduced frequency range (vinyl is capable of 20-20,000hz, but is often not in practice). Tons of tracks have very little energy below 30hz or so, and very few people will be able to hear lack of energy above 15,000hz in musical content.

Note: This is not me saying vinyl playback is as good as digital. My original comment was in reply to a few posts implying that vinyl had a distinct sound and sounded obviously different from digital. Which isn't necessarily the case.

So going back to this thread's title, one might suggest that the 20-20000Hz range is actually rather generous and one could subtract a few Hz from the bottom end and few kHz from the top end and still pretty much cover all the bases.

 

[Julf]

Similar to a dbx sub-harmonic synthesizer?

Pretty much, yes.

 

[Snarfie]

There was recently a thread/test about audible DAC/filter differences and 'everyone' had doubts that the tester was able to hear above 20kHz. So I did a short check to see if that is really so rare and incredible. On short: it is not!

Ultrasound hearing range:
Hearing thresholds for pure tones above 16kHz. Quick resume (from Table1):

• Test setup: 32 young ears tested, 19-25 years old. Max allowed/tested level ~100dB SPL.
• 29 of 32 heard 20 kHz .. some at the very low level of 66dB!
• 16 (50%) heard 24 kHz.
• 3 (9%) heard 28 kHz.

Infrasound hearing range:
Hearing at Low and Infrasonic Frequencies. Quick resume (Table3): at 100dB SPL the audible threashold is ~8Hz.

Quick & pretty clear conclusion: the audible range is actually 8Hz - 28kHz.
(at the fairly ok level of 100dB SPL)

But ASR measures distortion/etc at 114dB and at that level the infrasound threshold is 4-5Hz. No data for ultrasound but probably fair to assume a few extra kHz. If we do a bit of rounding-up, like good engineers do to make sure that everything is covered, the range becomes 3Hz-30kHz. That sounds like the absolute bare minimum to me!

And maybe we can stop building devices like shrewd salesmen (i.e. barely cover the minimum required). And start building them like good engineers (i.e. clean to double the specs, just to be sure).
In that case, we should have audio devices and recordings that are 100% clean approx 1Hz-60kHz.

And if someone thinks that 3-30 or 1-60 are "too much" or "unnecessary", here are two examples from other fields/senses:
- vision. For a long time and based on various studies, the 'consensus' was that ~150-200 PPI resolution was "more than enough for everyone". But then Apple came up with retina-screens at 300 PPI and declared it to be the minimum required. After a bit of controversy, nowadays we have smartphones at 500+ PPI.
- touch. In the gaming community the mouse sensitivity/speed is very important and for a long time it was considered that ~2000 DPI is "more than enough for everyone". Another round of controversy and nowadays gamer-mice are at 10000+ DPI and consumer ones at 2000+ too.

So, why are we in the audio world supossed to be happy with the heavily truncated 20-20 'audible' range?!


P.S.
a 'bit' of extra science

Studies (many hundreds) about the effects of inaudible infra & ultra-sounds on the human body: here and here.

An AES meta-study on ultrasonics in music: A Meta-Analysis of High Resolution AudioPerceptual Evaluation:


Living cells do react to both infra- and ultra-sounds over the entire 1Hz-1MHz range, with some interesting hotspots in the 10-150kHz area.

The Hypersonic Effect detailed in 30+ studies from various Japanese scientists/universities as summarized by AES. Or as a BBC podcast:

Apparently, those 'inaudible' 20-120kHz sounds are still perceived and do make a difference in listeners' mood an level of enjoyment. Their presence/absence is also visible in EEG measurements.
And maybe you will enjoy some hypersonic art: A Study about Emotional Digital Art using Hypersonic Effect.

Live orchestras do go above 20kHz, some instruments even go above 100kHz: There's Life Above 20 Kilohertz:

and


P.S.2
someone else posted a nice list of (technical) reasons why 44kHz recording/playback is actually not enough.

P.S.3
If you do not like reading looong posts, you may enjoy someone singing it for you

Why not exceeding 20khz going subsonic.

I can remember somewhere end 70ties a publication that tones exceeding 20 khz could or are responsible for brass characteristics/color of such instrument like trumpets trombones etc. So it is suggested that basicly we don't t hear it in pure frequency but in tones of such instruments.

Never found a scientific justification for that or is there? So to fresh up my mind i ask Chatgpt

ChatGPT:​

You're right in recalling discussions from the late 1970s and early 1980s around the idea that ultrasonic frequencies (above 20 kHz) could play a role in the tonal qualities of instruments like trumpets, trombones, and other brass instruments. This concept was part of a broader conversation about how sound perception works, particularly in relation to harmonics and overtones, and how high-frequency components might influence the perceived "color" or timbre of sounds.
However, there is no widely accepted scientific evidence to support the idea that we directly "hear" frequencies above 20 kHz in a conventional sense, as these frequencies are typically beyond the range of human hearing. The basic theory behind the suggestion you mentioned was that while we might not consciously hear frequencies above 20 kHz, their presence could still influence the overall timbre or perceived character of an instrument. The reasoning often involved the idea that the human auditory system could perceive the effects of these high-frequency overtones indirectly, through non-linear interactions in the ear or through mechanical vibrations in the body, even if we don’t "hear" them as discrete tones.

Some Key Points to Consider:​

1. Inaudible Frequencies and Timbre: While we cannot hear frequencies above 20 kHz, it is possible that they influence how we perceive lower frequencies. Instruments like trumpets and trombones create complex waveforms that include a broad spectrum of harmonics, some of which extend into the ultrasonic range. It's theorized that these high-frequency components could contribute to a fuller, richer sound or alter the perceived texture of the sound, even though we cannot consciously detect them.
2. Ultrasonic Harmonics and Nonlinearities: The idea that high-frequency overtones contribute to brass timbre might also relate to non-linear distortion effects in the ear or through the acoustics of the instrument itself. These interactions could generate audible effects at lower frequencies, which might explain why we associate certain "colors" with brass instruments, even though the frequencies themselves are beyond our hearing range.
3. Scientific Studies: Research on this topic has been somewhat limited. Most studies on musical acoustics focus on the harmonics within the audible range (20 Hz to 20 kHz) and how they contribute to timbre. Some studies have explored the role of ultrasonic frequencies in other domains (like audio recording and reproduction), but there's no strong consensus that these frequencies are critical to the perception of timbre in musical instruments.
4. Psychoacoustics: Psychoacoustics, the study of how humans perceive sound, suggests that while we may not hear ultrasonic frequencies directly, they could influence our perception of sound through mechanisms like psychoacoustic masking or the interaction of the sound waves with the resonance of the ear or the auditory pathway.

In Conclusion:​

While there has been speculation and some early theories regarding the role of ultrasonic frequencies in shaping the timbre of brass instruments, there isn't conclusive scientific evidence to support the idea that tones exceeding 20 kHz are directly responsible for the "brassiness" or color of such instruments. The interaction of these high frequencies with the rest of the sound spectrum and the complex acoustics of both the instrument and the ear might play a role, but more rigorous research would be needed to fully understand the phenomenon. It remains an interesting and somewhat speculative area within the field of acoustics.

 

[Julf]

Never found a scientific justification for that or is there? So to fresh up my mind i ask Chatgpt

Always remember the thing at the end, "ChatGPT kan fouten maken. Controleer belangrijke informatie." - "ChatGPT may make mistakes. Please check important information."

ChatGPT and other large language models are not "intelligent" in the way we usually think about it. They are just statistical text generators that generate text based on whatever text they have been fed with. Fine if they have been fed with peer-reviewed, verified scientific reports, less fine if they have been fed with all of the Internet...

 

[Snarfie]

Always remember the thing at the end, "ChatGPT kan fouten maken. Controleer belangrijke informatie." - "ChatGPT may make mistakes. Please check important information."

ChatGPT and other large language models are not "intelligent" in the way we usually think about it. They are just statistical text generators that generate text based on whatever text they have been fed with. Fine if they have been fed with peer-reviewed, verified scientific reports, less fine if they have been fed with all of the Internet...

You are correct but reading this specific analisys it is quite precise. First of all i would know if what i once read many years ago late 70ties if that was the case.

 

[Julf]

You are correct but reading this specific analisys it is quite precise. First of all i would know if what i once read many years ago late 70ties if that was the case.

That is the danger of LLM:s ("AI"). They sound very precise and credible, but could be spouting pure rubbish. Garbage in, garbage out. Only way to be sure is to ask ChatGPT for references and sources, and check them.

 

[Snarfie]

That is the danger of LLM:s ("AI"). They sound very precise and credible, but could be spouting pure rubbish. Garbage in, garbage out. Only way to be sure is to ask ChatGPT for references and sources, and check them.

Reading this specific answer it was more or less the same answer i got at the time from high end shops ( scientific not proven very debateble) that i visited. But AI now went more in-depth but conclusions are still the same so noting changed in this case since around 45 years. Will ask if chatgpt can come up with the references.

Chatgpt answer:


Great follow-up. Here's a breakdown of the sources and basis for the answer I gave regarding the potential role of ultrasonic (and subsonic) frequencies in the timbre of brass instruments:
Main Sources and Background for the Response:

General Acoustics and Musical Instrument Research:

Fletcher, N. H., & Rossing, T. D. (1998). The Physics of Musical Instruments. This is a foundational text in understanding how instruments produce sound, including harmonic structures and the spectral content of brass instruments.

Benade, A. H. (1990). Fundamentals of Musical Acoustics. Benade discussed in detail how harmonic overtones contribute to perceived timbre and richness in brass and wind instruments.

Ultrasonic Frequency Influence on Perception:

Oohashi, T. et al. (2000). High-Frequency Sound Above the Audible Range Affects Brain Electric Activity and Sound Perception. Journal of Neurophysiology.

This study is often cited when discussing whether ultrasonic content affects perception despite being inaudible. The authors suggest that frequencies above 20 kHz can have a measurable effect on EEG activity and subjective listening impressions.

Meyer, J. (2009). Acoustics and the Performance of Music. Meyer discusses the spectral characteristics of different instruments, including brass, and how those influence perception.

Nonlinear Perception Theories:

Psychoacoustic literature often touches on how non-linearities in the ear or recording systems might allow inaudible frequencies to interact with audible ones. While not always directly linked to brass instruments, this is a common area of discussion in audio perception and reproduction research.

Moore, B. C. J. (2003). An Introduction to the Psychology of Hearing. This text covers psychoacoustic phenomena, such as masking and distortion products, that may help explain indirect perception of certain spectral content.

Speculative and Experimental Commentary:

Anecdotal discussions in audio engineering and music production circles—particularly among high-resolution audio proponents—often suggest that ultrasonic content may influence subjective sound quality. This is speculative, but it adds to the broader context.

Some ASR or Gearspace forum discussions over the years have touched on this topic, though more as thought experiments than scientific consensus.

To Clarify:

There is no mainstream scientific agreement that ultrasonic (or subsonic) components are critical to how we perceive brass timbre.

However, there are theoretical models and fringe studies suggesting they might play a role via indirect mechanisms, like ear nonlinearities or bone conduction, even if the listener can’t directly “hear” those frequencies.

If you're looking to dig deeper, I can help pull the specific studies or chapters where these topics are covered. Would you like links or excerpts from any of the major sources listed?

 

[Julf]

Reading this specific answer it was more or less the same answer i got at the time from high end shops ( scientific not proven very debateble) that i visited.

OK, "high end" shops make up more BS than even ChatGPT.

Will ask if chatgpt can come up with the references.
...
Oohashi, T. et al. (2000). High-Frequency Sound Above the Audible Range Affects Brain Electric Activity and Sound Perception. Journal of Neurophysiology.

Ah, of course, good old Oohashi, no matter how many times that stuff has been debunked...

 

[Snarfie]

OK, "high end" shops make up more BS than even ChatGPT.

I'm refering late 1970ties that i visited high end shops regarding this topic. Looking to the
answers they were probably right. But looking to your response regarding AI you are quite skeptical. I agree looking to it respons we have to be objective but not utterly defensive. Lots of things are improving for instance gathering fast amounts of information/data regarding numerous topics using AI’s Role in cancer diagnosis an treatment etc etc.
But now we go off topic

 

[Julf]

I'm refering late 1970ties that i visited high end shops regarding this topic. Looking to the
answers they were probably right.

Not sure what the high end shops told you at the time - all I know that there was a lot of hype about "supertweeters" etc., and attempts to justify them.

But looking to your response regarding AI you are quite skeptical. I agree looking to it respons we have to be objective but not utterly defensive. Lots of things are improving for instance gathering fast amounts of information/data regarding numerous topics using AI’s Role in cancer diagnosis an treatment etc etc.

Sure - as I wrote, LLM's ("AI" is a marketing term these days) are fine and useful *if* they have been fed with peer-reviewed, verified scientific reports, less fine if they have been fed with all of the Internet. I would have to add "and properly used". They are very good for spotting (and reproducing) patterns.

 

[MRC01]

I was part of a team running the sound and light at a disco back in the day. We did actually make a box of electronics that basically took the lower bass notes and added a copy of them, shifted an octave down.

Similar to a dbx sub-harmonic synthesizer?

Here's another approach, described in detail both acoustically & electrically:

bass – HeadWize Memorial

Posts about bass written by Jour

headwizememorial.wordpress.com

Jan Meier made a DSP implementation of this in one of his devices, the COUNTry: https://www.meier-audio.de/country.html

 

[sigbergaudio]

So going back to this thread's title, one might suggest that the 20-20000Hz range is actually rather generous and one could subtract a few Hz from the bottom end and few kHz from the top end and still pretty much cover all the bases.

At the bottom end I think you will find some movie enthusiasts that would like to keep 20hz and even below, and young people can hear up to 20khz - so I think the current frequency range is good. Modern (digital) 44.1khz sources are typically better than this. They don't really have a lower limit, and they typically go beyond 20khz. So you will get something like 5-22,000hz with little deviation. Which is plenty.

 

[kyuu]

At the bottom end I think you will find some movie enthusiasts that would like to keep 20hz and even below, and young people can hear up to 20khz - so I think the current frequency range is good. Modern (digital) 44.1khz sources are typically better than this. They don't really have a lower limit, and they typically go beyond 20khz. So you will get something like 5-22,000hz with little deviation. Which is plenty.

Sure, I wasn't actually suggesting we "downgrade" from 20-20000Hz as the de-facto standard for the audible range. Just sort of a retort to the premise of the thread, that the range is already pretty generous and most people wouldn't notice if some of the frequencies in that range were missing. Particularly for those most into the hi-fi space (generally middle-age or older, IME), they're certainly not missing anything with 44.1/48kHz sampling considering they (I should say we, I guess) could get by on ~30kHz sampling and not miss a thing.

 

[Julf]

Sure, I wasn't actually suggesting we "downgrade" from 20-20000Hz as the de-facto standard for the audible range. Just sort of a retort to the premise of the thread, that the range is already pretty generous and most people wouldn't notice if some of the frequencies in that range were missing. Particularly for those most into the hi-fi space (generally middle-age or older, IME), they're certainly not missing anything with 44.1/48kHz sampling considering they (I should say we, I guess) could get by on ~30kHz sampling and not miss a thing.

I still have (somewhere in the attic) a Yamaha TX16W sampler that a lot of late-1980s/early-1990s musicians used. It is 12 bits at 33 kHz when doing stereo. The Fairlight CMI, used by a huge number of artists, was 8 bits at first 24 kHz (Series I) and then 32 kHz (Series II).

 

[Vacceo]

Sure, I wasn't actually suggesting we "downgrade" from 20-20000Hz as the de-facto standard for the audible range. Just sort of a retort to the premise of the thread, that the range is already pretty generous and most people wouldn't notice if some of the frequencies in that range were missing. Particularly for those most into the hi-fi space (generally middle-age or older, IME), they're certainly not missing anything with 44.1/48kHz sampling considering they (I should say we, I guess) could get by on ~30kHz sampling and not miss a thing.

Below 20hz you don´t "hear", so at that point, distortion probably does not matter much. I mean, the difference between a 9hz "sound" and a 10hz "sound" is that the lower it is, the more pressure you feel. Measuring output? Yeah, that is productive, at least you´ll know how low and how hard the infrasonics will hit.

 

[MRC01]

Below 20hz you don´t "hear", so at that point, distortion probably does not matter much. I mean, the difference between a 9hz "sound" and a 10hz "sound" is that the lower it is, the more pressure you feel. Measuring output? Yeah, that is productive, at least you´ll know how low and how hard the infrasonics will hit.

Talking about 9-10 Hz is a full octave below 20 Hz, so of course we don't hear that. But most people can hear a bit below 20 Hz, and by "hear" I mean literally hear, not feel, like on headphones. So response down to 20 Hz is important for perceptual transparency, even if not important for musical transparency since not much music has anything that low.

 

[Vacceo]

Talking about 9-10 Hz is a full octave below 20 Hz, so of course we don't hear that. But most people can hear a bit below 20 Hz, and by "hear" I mean literally hear, not feel, like on headphones. So response down to 20 Hz is important for perceptual transparency, even if not important for musical transparency since not much music has anything that low.

Speakers are not only used to reproduce music, and in content such as films or videogames, that low end becomes important.

 

[Flaesh]

But ASR measures distortion/etc at 114dB

I seem to have missed this. Are there distortion measurements somewhere here at 114 dB SPL and frequencies below 10 Hz? But I hardly read subwoofer tests.