Does inaudible above 20khz really matter?
- Thread starter Mike
- Start date
Speedskater
Member
Heck, from a listening to music, nothing above lets say about 16 kHz matters. But from an equipment point of view above 20 kHz can be very important. If a hi-fi component doesn't handle these frequencies well, it can generate Intermodulation Distortion which shows up way down in the audio band.
audio.bill
Active member
While it is true that most people can't hear fundamental frequencies above 20kHz and even lower as we age, testing has determined that we are able to discern the difference when harmonic frequencies well beyond that limit are truncated. It has also been determined that in order for electronics to be able to produce the audible frequency range without introducing significant phase distortion requires a bandwidth many times greater than that upper limit.
CPP
Active member
Well I can't hear squat above 12.5 as proven by doctor, but I still feel that equipment needs to be capable to allow clear and details freqs to be passed so support the requirements of the musical instruments that made the music even if I can't hear those dog capable notes.
Attachments
AJ Soundfield
Well-known member
Hi Bill,
I follow perceptual science research pretty closely. Could you cite the sources of your claims above? Thanks.
cheers,
AJ
audio.bill
Active member
Hi AJ,
A couple of appropriately titled articles found with a quick search are: "The World Beyond 20kHz" and "There's Life Above 20 Kilohertz". As far as electronics specifically requiring wide bandwidth designs in order to maintain phase integrity of the audio signal, I've read about such research related to the design of several high end electronics manufacturers but am currently unable to locate them. The references I saw were several years ago, but companies like Spectral and Luminance Audio discuss why they implement such wide bandwidth amplifiers in their designs. It's very expensive to design and implement stable instrument grade MHz bandwidth electronics but some companies elect to since they believe it ultimately impacts the resulting performance of their products. Sorry I couldn't find more accredited scientific proof but hopefully the sources I provided will generate some further discussion on the topic.
Yours in higher fidelity,
- Bill
AJ Soundfield
Well-known member
Hi Bill, thanks for the quick search links.
Also unfortunately, Mr Oohashis research is debunked nonsense. Mr Boyk neglected to mention the name of the paper:
Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect
He also forgot to mention that a follow up study, critical for all science, failed to replicate the results...unless IM distortion into the audio band <20k was introduced.
Even funnier is that a later follow up study by same group found a negative "feel bad" effect to inaudible hf
Like I mentioned, I follow this stuff closely. It's sort of critical to what I do.
cheers,
AJ
Unfortunately, Mr Blackmer doesn't seem to understand Intermodulaton distortion and presents zero listening test data to support his beliefs, neither acceptable for perceptual research.
Also unfortunately, Mr Oohashis research is debunked nonsense. Mr Boyk neglected to mention the name of the paper:
Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect
He also forgot to mention that a follow up study, critical for all science, failed to replicate the results...unless IM distortion into the audio band <20k was introduced.
Even funnier is that a later follow up study by same group found a negative "feel bad" effect to inaudible hf
Like I mentioned, I follow this stuff closely. It's sort of critical to what I do.
cheers,
AJ
a.wayne
Well-known member
So anything above 20K is only good for debunking .... 
audio.bill
Active member
AJ - I don't claim to be a perceptual science researcher so I yield to your expertise on this topic. Good listening (that we both enjoy!) 
- Bill
- Bill
AJ Soundfield
Well-known member
D
DSkip
Guest
The Rosso Fiorentino speakers in the Reference and Flagship line go all the way up to 100khz with their super tweeter. I'm not sure what impact this has, but I can say that the most common compliment I get with these speakers is the extended soundstage they present. This includes all three axis.
There is something happening up there that definitely impacts the sound. I believe being able to reproduce those frequencies is definitely a positive.
There is something happening up there that definitely impacts the sound. I believe being able to reproduce those frequencies is definitely a positive.
Jim Smith
Member
Generally, speaker drivers will start beaming a frequency having a wavelength equal to the diameter of the radiating cone.
I have always wondered how frequencies above 30 kHz or more could possibly have any effect, since they are (depending on the width of the tweeter driver) less than 1/2 of an inch in width, therefore being reproduced in a constant tiny (1/2" wide or less) beam at 90 degrees from the driver. As such, there is almost no chance that they can be heard. It's hard enough to find a spot to place a microphone exactly so that it can measure the extended response, especially in the listening area.
I am not saying that the effects cannot be experienced, I simply wonder how it can happen.
I have always wondered how frequencies above 30 kHz or more could possibly have any effect, since they are (depending on the width of the tweeter driver) less than 1/2 of an inch in width, therefore being reproduced in a constant tiny (1/2" wide or less) beam at 90 degrees from the driver. As such, there is almost no chance that they can be heard. It's hard enough to find a spot to place a microphone exactly so that it can measure the extended response, especially in the listening area.
I am not saying that the effects cannot be experienced, I simply wonder how it can happen.
Jim Smith
Member
A few good measurement mics do. I wasn’t very clear in my question above. If the tweeter driver isn’t pointing EXACTLY at the mic, the extreme high frequencies cannot be measured. So how can we hear missing high frequencies at our seat? Even if our hearing went out to 100K, and you had the speakers aimed precisely (horizontally & vertically) at each eardrum, if you moved your head in any direction a half-inch to an inch, those extreme high frequencies would be totally gone. So how does this technology overcome the basic laws of physics?
Sent from my iPhone using Tapatalk
Thank you Jim.
And I'm sorry, I wasn't clear either in my question.
I also meant: do recording engineers use microphones that capture sounds far above 20 kHz?
Thus meaning: do mastertapes really have a lot of content beyond let's say 25 kHz?
Jim Smith
Member
IME, no - some higher-end recordings go out a bit further, but relatively few.
But I am no expert on this - it's just my opinon.
audio.bill
Active member
After some further searching I found an interview with Raphael Pasche from CH Precision which explains their thoughts on the advantages of speed and high bandwidth design in their electronics. I'm including the applicable excerpt below for easy reference and hope that some members will find it helpful.
The advantages of speed
Companies like Spectral Audio and Soulution - like CH Precision - all favour High Speed designs. What are the benefits and are there any disadvantages as well?
‘High speed design means (at least) two things: wide frequency response and high slew rate. The main advantage of having a wide frequency response is that it keeps the group delay constant for as long as possible. This ensures that the phase relationship between various frequencies of an audio signal is kept as identical as possible before and after having been through the unit. In order for music to sound as natural as possible, the phase relationship (and by mathematical extension the group delay) of the signal mustn't be altered.
High speed design allows it. We talk of the audio domain being between 20Hz and 20kHz. This is right from a level point of view, but from a phase point of view, if the amplifier frequency response was only 100kHz, we would see that the phase (associated to the low pass filter forming the bandwidth of the device) would start rolling off far below 100kHz, well into the audio range mentioned above.
The slew rate is just as important as wide bandwidth. As you know, the slew rate is the maximum rate at which a device can respond to an abrupt change of input level. A low slew rate would mean that a transient signal would have the abrupt change impaired, to be avoided at all costs.
The main disadvantage of high speed designs is that they are more susceptible to RF perturbations which means it is a much tougher job to ensure that RF perturbations would disrupt the unit, both from a functioning and from a qualitative/performance point of view.'
The advantages of speed
Companies like Spectral Audio and Soulution - like CH Precision - all favour High Speed designs. What are the benefits and are there any disadvantages as well?
‘High speed design means (at least) two things: wide frequency response and high slew rate. The main advantage of having a wide frequency response is that it keeps the group delay constant for as long as possible. This ensures that the phase relationship between various frequencies of an audio signal is kept as identical as possible before and after having been through the unit. In order for music to sound as natural as possible, the phase relationship (and by mathematical extension the group delay) of the signal mustn't be altered.
High speed design allows it. We talk of the audio domain being between 20Hz and 20kHz. This is right from a level point of view, but from a phase point of view, if the amplifier frequency response was only 100kHz, we would see that the phase (associated to the low pass filter forming the bandwidth of the device) would start rolling off far below 100kHz, well into the audio range mentioned above.
The slew rate is just as important as wide bandwidth. As you know, the slew rate is the maximum rate at which a device can respond to an abrupt change of input level. A low slew rate would mean that a transient signal would have the abrupt change impaired, to be avoided at all costs.
The main disadvantage of high speed designs is that they are more susceptible to RF perturbations which means it is a much tougher job to ensure that RF perturbations would disrupt the unit, both from a functioning and from a qualitative/performance point of view.'