Recently, I was doing some recording, digitally, to study the harmonic content of real life sounds of musical instruments - especially in the higher frequency area of the spectrum. I took a couple of musical instruments and recorded some individual high pitch notes. I was very surprised to find significant harmonic content above 20Khz! and well into the 30Khz range as well! And that's with a fundamental frequency of just 1-2Khz of the recorded note.
Only because I was recording at 96Khz I could capture the spectrum all the way up to 48Khz, and hence the revelation.
Question # 1 -
Almost any recording or reproducing equipment microphone, pre-amps, amps, speakers, recording media(except digital), roll off heavily past 20Khz(especially mics and tweeters). If there really is so much "harmonic" content in a signal then isn't it all getting lost ?? Is this what we have been made to believe as "High Fidelity" ??
Question # 2 -
I had just used a couple of "wind" instruments that were at my disposal for this little "experiment". It was a harmonica and a harmonium. I didn't get a chance to record the flute yet and I dont have access to a piano, violin or a guitar, yet, but I certianly intend to! In the meantime, I am very curious to know if such high harmonic content is characteristic of wind instruments only ? or string and impact instruments also - like violin, guitar and piano ? Has anyone every observed harmonic content of these musical instruments ?
Question # 3 -
So now how do you define the upper limit of human hearing ? I am not able to hear anything above 16Khz. But bear in mind that is a pure 16Khz fundamental with no overtones or harmonics whatsoever. Now consider a 1Khz signal that has several harmonics all the way through 30Khz. You can hear that just perfect, right ? So are you "hearing" the 30Khz harmonic ? I dont think you can look at it that way - it is not the 30Khz you are hearing, it is still 1Khz with lots of harmonics.
Another way of explaining this is -
Imagine both the "waves" as if you were looking at them on an oscilloscope (wish I could explain with a picture) - i.e. the 16Khz wave and the 1Khz wave consisting a lot of harmonics. On the o'scope you will see the 1Khz+harmonics wave in the same number of time divisions as a pure 1Khz wave, except that our 1Khz+hramonics signal is bent out of shape (lots of harmonics), but it is still oscillating at one thousand times per second!
Sorry for the long winded post. But I am eagerly trying to make sense of all of this. To me it looks like its time things like "Hi-Fi"delity and human hearing need to be redefined.
TiA!
--
Only because I was recording at 96Khz I could capture the spectrum all the way up to 48Khz, and hence the revelation.
Question # 1 -
Almost any recording or reproducing equipment microphone, pre-amps, amps, speakers, recording media(except digital), roll off heavily past 20Khz(especially mics and tweeters). If there really is so much "harmonic" content in a signal then isn't it all getting lost ?? Is this what we have been made to believe as "High Fidelity" ??
Question # 2 -
I had just used a couple of "wind" instruments that were at my disposal for this little "experiment". It was a harmonica and a harmonium. I didn't get a chance to record the flute yet and I dont have access to a piano, violin or a guitar, yet, but I certianly intend to! In the meantime, I am very curious to know if such high harmonic content is characteristic of wind instruments only ? or string and impact instruments also - like violin, guitar and piano ? Has anyone every observed harmonic content of these musical instruments ?
Question # 3 -
So now how do you define the upper limit of human hearing ? I am not able to hear anything above 16Khz. But bear in mind that is a pure 16Khz fundamental with no overtones or harmonics whatsoever. Now consider a 1Khz signal that has several harmonics all the way through 30Khz. You can hear that just perfect, right ? So are you "hearing" the 30Khz harmonic ? I dont think you can look at it that way - it is not the 30Khz you are hearing, it is still 1Khz with lots of harmonics.
Another way of explaining this is -
Imagine both the "waves" as if you were looking at them on an oscilloscope (wish I could explain with a picture) - i.e. the 16Khz wave and the 1Khz wave consisting a lot of harmonics. On the o'scope you will see the 1Khz+harmonics wave in the same number of time divisions as a pure 1Khz wave, except that our 1Khz+hramonics signal is bent out of shape (lots of harmonics), but it is still oscillating at one thousand times per second!
Sorry for the long winded post. But I am eagerly trying to make sense of all of this. To me it looks like its time things like "Hi-Fi"delity and human hearing need to be redefined.
TiA!
--
thanks for asking - it reminded me to mention a point about this -
This measurement was done using the ECM 8000. Yes it has fast rolloff after 20Khz. Point is that if the ecm 8000 could capture the harmonics, it indicates that those are indeed significant harmonics. With a mc having reasonable response up to that range, the spectrum would really be a whole different "picture".
The pre-amp (UB802) and soundcard (ESI Juli@) have good response over 40Khz+.
This measurement was done using the ECM 8000. Yes it has fast rolloff after 20Khz. Point is that if the ecm 8000 could capture the harmonics, it indicates that those are indeed significant harmonics. With a mc having reasonable response up to that range, the spectrum would really be a whole different "picture".
The pre-amp (UB802) and soundcard (ESI Juli@) have good response over 40Khz+.
Is that a statement, an assumption or a question?
If your hearing stops at 20kHz - what is the difference between the fundamental and the harmonics? Do you believe because they are instrument "generated" harmonics they are more audible than instrument generated fundamentals? I have a hard time understanding your confusion. If you can't hear it - you can't hear it.
That whole thing ain't new at all - somebody made astudy of the fundamentals and harmonics of non amplified acoustic instruments years ago, http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
100kHz ?
I came across an article about us hearing almost all the way up to 100Khz or so. Sounds around 20Khz and above not being 'audible' in the normal sense of 'hearing' but that the higher frequencies 'affect' our perception of what we 'hear'.
I downloaded that paper but can't seem to find it.
I've come across this topic several times but it seems to appear and disappear pretty fast ! I think it's a VERY interesting topic . How about others with more info putting up their opinions and any useful links to more information that they might have come across.
I came across an article about us hearing almost all the way up to 100Khz or so. Sounds around 20Khz and above not being 'audible' in the normal sense of 'hearing' but that the higher frequencies 'affect' our perception of what we 'hear'.
I downloaded that paper but can't seem to find it.
I've come across this topic several times but it seems to appear and disappear pretty fast ! I think it's a VERY interesting topic . How about others with more info putting up their opinions and any useful links to more information that they might have come across.
Yes, the hypothesis by some japanese researchers was that sound can be perceived - but non aurally - above 20 kHz. There methodology and the results were quite doubtful and to say the least inconclusive. The claim of up to 100kHz is ludicrous, and unproven by any researcher.
I question the importance of such perception for the listening to music.
".................I question the importance of such perception for the listening to music............"
I think we can say that only after we do have the means to reliably record live music with VERY wide band equipment ( up to 100KHz? ) . That would mean that every piece of equipment including microphones would have to be be capable of 100Khz operation. Tough ? , yes I think so , but I think not impossible though it might be quite expensive to implement. We have tweeters going up close to 100Khz so why not microphones !
The other option is to put on high frequency filters over the ears that would absorb / block everything above 20 Khz and see what difference we can find when listening to live music. But would that be a reliable test ? The nostrils are open ... the bones conduct ...
I think we can say that only after we do have the means to reliably record live music with VERY wide band equipment ( up to 100KHz? ) . That would mean that every piece of equipment including microphones would have to be be capable of 100Khz operation. Tough ? , yes I think so , but I think not impossible though it might be quite expensive to implement. We have tweeters going up close to 100Khz so why not microphones !
The other option is to put on high frequency filters over the ears that would absorb / block everything above 20 Khz and see what difference we can find when listening to live music. But would that be a reliable test ? The nostrils are open ... the bones conduct ...
People exposed to ultrasonics tend to have headaches and become irritable. My guess is we're far better off with a 20kHz cutoff. If your system could perfectly reproduce live music in your living room, it might be a lot less desirable than it seems on the surface. OTOH, maybe very high bandwidth mics and such would do a better job on initial transients, the area someone here was promoting as most important. IMO, the sustained ultrasonics are inaudible.
Ok, like any true, by evidence unfazed audiophile you put a statement in here, whose value is more than doubtful, evidence for which does not exist and demand equipment to capture sound that might be important for bats?
With unsupported statements like that - I feel it is time to leave you guys to your play in the sandbox...
IF you play with a signal generator and a tweeter with good HF extension, you will find that your ears exhibit a far more steep roll-off above 16 or 18Khz than any tweeter or microphone ever made...
You may also find that it's difficult to hear short little bursts of tones above 10Khz, which is the same situation that arises when you have one or two 30Khz pulses on top of each period of a 1Khz fundamental.
You may also find that it's difficult to hear short little bursts of tones above 10Khz, which is the same situation that arises when you have one or two 30Khz pulses on top of each period of a 1Khz fundamental.
audio-kraut said:
That research showed that hi-freq sounds do affect brain activity. So, depending on whether there is sound above 20kHz or not gives different brain activities. But it is not clear whether that also gives different concious perceptions of the sound.
It is interesting to read but does give the impression that their agenda is to validate 96 and 192kHz sample rates.
I put the article on my home page for anyone who is interested. For personal study only, of course.
Jan Didden
Unless you actually have source material with the necessary frequency range, the question is, of course, purely academic.
If you do have such material, as always, the best answer is to try it and see. A small amp feeding, for example, a peizo transducer would be quite a simple project. If you (or others) can hear or otherwise perceive the extended HF, it's worth looking into; if not, be thankful that your system doesn't need upgrading in that particular respect!
If you do have such material, as always, the best answer is to try it and see. A small amp feeding, for example, a peizo transducer would be quite a simple project. If you (or others) can hear or otherwise perceive the extended HF, it's worth looking into; if not, be thankful that your system doesn't need upgrading in that particular respect!
So it looks like there is no disagreement that musical instruments do have harmonic content well over 20Khz.
What remains now to be found is whether it would sound the same to the human ear if the harmonics over 20Khz were removed v/s all harmonics retained just like in the original signal.
Which I'd imagine shouldn't be too hard to experiment today which the availibity of extended range tweeters and amplifiers.
Here's what I think about all this -
In terms of human hearing capability, a higher order harmonic is not the same way as a fundamental of the same frequency. i.e. the 29th harmonic at 30Khz is not the same as a 30Khz fundamental. It is intuitive to think that since you cant hear 16Khz so obviously you can't hear 30Khz, but I doubt its as simple as that.
What remains now to be found is whether it would sound the same to the human ear if the harmonics over 20Khz were removed v/s all harmonics retained just like in the original signal.
Which I'd imagine shouldn't be too hard to experiment today which the availibity of extended range tweeters and amplifiers.
Here's what I think about all this -
In terms of human hearing capability, a higher order harmonic is not the same way as a fundamental of the same frequency. i.e. the 29th harmonic at 30Khz is not the same as a 30Khz fundamental. It is intuitive to think that since you cant hear 16Khz so obviously you can't hear 30Khz, but I doubt its as simple as that.
It does not sound the same. I have a disc "Resolution Project" with same music recorded simultaneously with resolution and sample rate up to 24bit/192kHz. Have a look here:
http://mixonline.com/mag/audio_resolution_project/
http://mixonline.com/mag/audio_resolution_project/
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.