[analogsurviver]

  OP: You split off another thread and ended up getting the same unsubstantiated ramblings in both threads. It's double now. Counter productive.

Agreed.
 
The only thing I am not going to tolerate is the insistance that beyond 20 kHz does not matter.
 
It just is not true. I know it will take quite some time to be accepted - but eventually it will happen.

 

[cjl]

It would happen pretty quickly if the people claiming it would just provide some evidence. A couple of well documented double blind tests demonstrating significant audibility of a 20kHz low pass filter in music with significant high frequency content would definitely change a lot of minds. Strangely enough though, all of the anecdotes seem to dry up when you remove the visual indications...

 

[analogsurviver]

  It would happen pretty quickly if the people claiming it would just provide some evidence. A couple of well documented double blind tests demonstrating significant audibility of a 20kHz low pass filter in music with significant high frequency content would definitely change a lot of minds. Strangely enough though, all of the anecdotes seem to dry up when you remove the visual indications...

I can and will provide some evidence - my own recordings. But, it takes to have the equipment that supports it - particularly headphones and speakers with response above 20 kHz - if at least 88.2/16 capable  DAC is self explanatory to be mandatory for such test(s).
 
I have just installed the voxengo FFT ( it still hiccups a bit, but it should do well tomorrow ) - and will go trough some 5 years worth of my DSD recordings. There is one harpsichord recording that clearly shows harmonics to at least 40 kHz - if not higher. Music - not stationary constant noise as seen in 192/24 Linn sourced recordings some 2-3 weeks ago in the other thread. But others I still have to check for ultrasonic contents. It may take a while (as DSD has to be converted to 192/24 for the analysis) - but contenders are other stuff with harpsichord, percussion, symphonic music ( brass, percussion, strings ) , etc. Most of my recordings are vocal music, that is to say without ultrasonic content. 
 
It will not be prior to 2 weeks or so from now on. I have to put together as versatile and as demanding demo that can fit within say 10 tracks each one minute long for the forthcoming visit to the Munchen High End 2015 show in exactly one week - and my PC has just today again started to recognize the RAID with all of my recordings...
 
I would like to listen to some stuff I am interested in - with recordings I am really familiar with.

 

[castleofargh]

 

  OP: You split off another thread and ended up getting the same unsubstantiated ramblings in both threads. It's double now. Counter productive.

Agreed.
 
The only thing I am not going to tolerate is the insistance that beyond 20 kHz does not matter.
 
It just is not true. I know it will take quite some time to be accepted - but eventually it will happen.

and it is the lead reason for us all arguying. vinyls, DSD, filters.... it always leads to you wishing to get all your ultrasounds.
and us saying you're wrong, always originates from the fact that the differences would be in the ultrasound and thus not audible.
 
we're not going to change our opinion because for most of us, what we tried with ultrasounds didn't change a thing. again I have a very hard time indentifying a loud 17khz nowadays, so I just cant figure out how some attenuated ultrasounds in a record would ever sound like something to me?
it's a very simple thought process, and I don't get where my logic is flawed enough to accept that you're right. I'm certainly willing to, because I want to learn much more than I want to be right, and I know for a fact that I'm often wrong. but still I can't help but see your reasoning about ultrasounds as flawed. how could something that I can't hear be of importance to my listening?

 

[analogsurviver]

 
and it is the lead reason for us all arguying. vinyls, DSD, filters.... it always leads to you wishing to get all your ultrasounds.
and us saying you're wrong, always originates from the fact that the differences would be in the ultrasound and thus not audible.
 
we're not going to change our opinion because for most of us, what we tried with ultrasounds didn't change a thing. again I have a very hard time indentifying a loud 17khz nowadays, so I just cant figure out how some attenuated ultrasounds in a record would ever sound like something to me?
it's a very simple thought process, and I don't get where my logic is flawed enough to accept that you're right. I'm certainly willing to, because I want to learn much more than I want to be right, and I know for a fact that I'm often wrong. but still I can't help but see your reasoning about ultrasounds as flawed. how could something that I can't hear be of importance to my listening?

Simple answer. Imagine a x-th harmonic of say trumpet being at 80 kHz, (x+1)-th harmonic being at 83 kHz. We could go to the real and exact harmonics of a trumpet from the link proving response of the muted trumpet extends past 100 kHz. 
 
Now, trough intermodulation - difference between the two frequencies - we arrive at (83 - 80) kHz = 3 kHz. And 3 kHz is audible - and these 3 kHz are missing in the recording that has gone trough 20 kHz low pass filter.
 
Clearer now ? 
 
The whole myriad of naturally occurring intermodulation products we hear in live music is missing in the CD version. 
 
I can (now...54 years old)  hear sine wave to approx 14 kHz - yet frequency response  limited music did not sound right to me from the day one. Philips prototype CD player at our electronics fair in 1979 or 1980.
 
Still doesn't.

 

[Steve Eddy]

Simple answer. Imagine a x-th harmonic of say trumpet being at 80 kHz, (x+1)-th harmonic being at 83 kHz. We could go to the real and exact harmonics of a trumpet from the link proving response of the muted trumpet extends past 100 kHz. 

Now, trough intermodulation - difference between the two frequencies - we arrive at (83 - 80) kHz = 3 kHz. And 3 kHz is audible - and these 3 kHz are missing in the recording that has gone trough 20 kHz low pass filter.

Clearer now ? 

The whole myriad of naturally occurring intermodulation products we hear in live music is missing in the CD version. 

What's clear is that you don't have the foggiest idea what you're talking about.

What is the non-linearity that would cause there to be an intermodulation product at 3kH?

If a trumpet is producing harmonics at 80kHz and 83kHz, only some sort of non-linearity, some distortion mechanism, would cause those two frequencies to fold down to 3kHz. Sure, some poor electronics in the recording chain could cause that to happen, but that's something you DONT WANT to happen. You simply don't understand what you're talking about.

se

 

[analogsurviver]

What's clear is that you don't have the foggiest idea what you're talking about.

What is the non-linearity that would cause there to be an intermodulation product at 3kH?

If a trumpet is producing harmonics at 80kHz and 83kHz, only some sort of non-linearity, some distortion mechanism, would cause those two frequencies to fold down to 3kHz. Sure, some poor electronics in the recording chain could cause that to happen, but that's something you DONT WANT to happen. You simply don't understand what you're talking about.

se

I specifically did say it is NOT intermodulation in the equipment. This would be something WE DO NOT WANT.
 
Human ear is not a linear device - FAR from it. And it COULD function something like I have described. We need to do more research in this direction - but how on earth it can be possible to encourage such research if everybody is claiming above 20 kHz does not matter ? There must be some place to start - one or another.
 
The easiest way out is to say that it does not matter.

 

[maverickronin]

  OP: You split off another thread and ended up getting the same unsubstantiated ramblings in both threads. It's double now. Counter productive.

 
*sigh*
 
Operation failed.....

 

[Steve Eddy]

I specifically did say it is NOT intermodulation in the equipment. This would be something WE DO NOT WANT.

Ok...

Human ear is not a linear device - FAR from it. And it COULD function something like I have described.

Ok. So you're hypothesizing that it's the non-learity of the ear that's producing that 3kHz intermod product, yes?

We need to do more research in this direction - but how on earth it can be possible to encourage such research if everybody is claiming above 20 kHz does not matter ? There must be some place to start - one or another.

Ok. So let's say we're going to explore this hypothesis. First, what microphone do you plan to use that has a flat, very linear response out to 80+ kHz? Are you aware of one? Because you'll need to find one before this can be put to the test. And once you've found the microphone, then you'll need to find some transducer that has an equally flat and very linear response out to 80+ kHz.

So which microphone and which transducer will we use to put this to the test?

The easiest way out is to say that it does not matter.

Yeah, it's pretty easy to simply say that it doesn't matter.

But we can also take what we do know, think it through, and get a pretty good feel for whether it would matter or not.

Let's say we have the perfect microphone and perfect transducer available to us. Record your trumpet and play it back.

First off, what do you expect the levels of those circa 80kHz harmonics to be? They'll certainly be far far far below the level of the fundamental and its near harmonics, wouldn't you agree?

And that means that any intermod products would be well below even those circa 80kHz harmonics.

Now given that the instantaneous dynamic range of human hearing is only about 30-40 dB, what are the chances that those intermod products would be of such staggeringly high levels to bring them within 30-40 dB of the fundamental in order for us to perceive them?

Pretty much zero if you ask me.

se

 

[analogsurviver]

Ok...
Ok. So you're hypothesizing that it's the non-learity of the ear that's producing that 3kHz intermod product, yes?
Ok. So let's say we're going to explore this hypothesis. First, what microphone do you plan to use that has a flat, very linear response out to 80+ kHz? Are you aware of one? Because you'll need to find one before this can be put to the test. And once you've found the microphone, then you'll need to find some transducer that has an equally flat and very linear response out to 80+ kHz.

So which microphone and which transducer will we use to put this to the test?
Yeah, it's pretty easy to simply say that it doesn't matter.

But we can also take what we do know, think it through, and get a pretty good feel for whether it would matter or not.

Let's say we have the perfect microphone and perfect transducer available to us. Record your trumpet and play it back.

First off, what do you expect the levels of those circa 80kHz harmonics to be? They'll certainly be far far far below the level of the fundamental and its near harmonics, wouldn't you agree?

And that means that any intermod products would be well below even those circa 80kHz harmonics.

Now given that the instantaneous dynamic range of human hearing is only about 30-40 dB, what are the chances that those intermod products would be of such staggeringly high levels to bring them within 30-40 dB of the fundamental in order for us to perceive them?

Pretty much zero if you ask me.

se

 

Ok...
Ok. So you're hypothesizing that it's the non-learity of the ear that's producing that 3kHz intermod product, yes?
Ok. So let's say we're going to explore this hypothesis. First, what microphone do you plan to use that has a flat, very linear response out to 80+ kHz? Are you aware of one? Because you'll need to find one before this can be put to the test. And once you've found the microphone, then you'll need to find some transducer that has an equally flat and very linear response out to 80+ kHz.

So which microphone and which transducer will we use to put this to the test?
Yeah, it's pretty easy to simply say that it doesn't matter.

But we can also take what we do know, think it through, and get a pretty good feel for whether it would matter or not.

Let's say we have the perfect microphone and perfect transducer available to us. Record your trumpet and play it back.

First off, what do you expect the levels of those circa 80kHz harmonics to be? They'll certainly be far far far below the level of the fundamental and its near harmonics, wouldn't you agree?

And that means that any intermod products would be well below even those circa 80kHz harmonics.

Now given that the instantaneous dynamic range of human hearing is only about 30-40 dB, what are the chances that those intermod products would be of such staggeringly high levels to bring them within 30-40 dB of the fundamental in order for us to perceive them?

Pretty much zero if you ask me.

se

Will answer tomorrow today (03:39 AM here ) after I wake up.

 

[Steve Eddy]

Will answer tomorrow today (03:39 AM here ) after I wake up.

*****!

se

 

[bigshot]

  *sigh* Operation failed.....

 
This guy is just looking for attention. I suspect there's reasons for that, but I'm not going to speculate on it. It might help if you could delete the thread. If you can't do that, change the title to take his name out of it. Having his name in the headline just feeds his neediness.

 

[analogsurviver]

Ok...
Ok. So you're hypothesizing that it's the non-learity of the ear that's producing that 3kHz intermod product, yes?
Ok. So let's say we're going to explore this hypothesis. First, what microphone do you plan to use that has a flat, very linear response out to 80+ kHz? Are you aware of one? Because you'll need to find one before this can be put to the test. And once you've found the microphone, then you'll need to find some transducer that has an equally flat and very linear response out to 80+ kHz.

So which microphone and which transducer will we use to put this to the test?
Yeah, it's pretty easy to simply say that it doesn't matter.

But we can also take what we do know, think it through, and get a pretty good feel for whether it would matter or not.

Let's say we have the perfect microphone and perfect transducer available to us. Record your trumpet and play it back.

First off, what do you expect the levels of those circa 80kHz harmonics to be? They'll certainly be far far far below the level of the fundamental and its near harmonics, wouldn't you agree?

And that means that any intermod products would be well below even those circa 80kHz harmonics.

Now given that the instantaneous dynamic range of human hearing is only about 30-40 dB, what are the chances that those intermod products would be of such staggeringly high levels to bring them within 30-40 dB of the fundamental in order for us to perceive them?

Pretty much zero if you ask me.

se

Regarding the transducer - the microphone - - there are more options.
 
The first choice would be the only mic intended for music recording covering the 100 kHz response : 
 
https://www.sanken-mic.com:6015/en/product/product.cfm/3.1000400
 
https://www.sanken-mic.com:6015/en/product/freqpola.cfm/3.1000400
 
Trouble - I do not have it - and neither I do have access to it - yet.
 
As it happens that the representative for the Bruel & Kjaer in our country is a 10 minute walk from my home, I rang them up if they have any 100 kHz capable mics. Yes, there are even two available, meaning a stereo recording could be made. Problem? They are measuring microphones, requiring 200 V polarizing voltage, supplied of course by the matching B&K preamp, output of which is terminated by a proprietary B&K 7 pin plug. How to mate that to anything usual in audio is beyond me at the moment, but  I have arranged for a meeting coming Monday to discuss the possibility to work something out. I will also know exactly which type of B&K mics are available.
 
As of recently, there is a privately owned Merging Technologies Hapi http://www.merging.com/products/hapi
in our country - to which I might be eventually able to get access. This is DSD256/PCM384 capable recorder - and should be "flat" to 100 kHz - and above - but I did not see yet any objective measurements confirming this.
 
Second best option would be to use my Korg recorders - DSD128 and PCM192 - not entirely covering 100 kHz bandwidth; DSD going past 100 kHz, but starts rolling off around 50 kHz with 6 dB (or so..) per octave above that frequency, PCM192 being flat to about 80-85 ( from mmemory, give or take a few kHz) kHz with the customary brickwall filtering above that. It drops off to nothing in extremely small bandwidth.
 
It is also a logistics problem - as far as I know, Hapi requires DAW and I am not sure if it can work as a stand-alone recorder. ( It is out of my reach financially, so I did not investigate as far as I would have in case it was *affordable* ). And the studio in which it is located is dedicated to the recordings of piano - meaning I would have to bring musicians ( trumpet, percussion, violin, etc ) there. 
 
As to the level of the harmonics extending beyond 20 kHz - please see http://www.cco.caltech.edu/~boyk/spectra/spectra.htm
 
My own experience with a SINGLE of my recordings so far tested with FFT (harpsichord) indicates a clear output to minimum 40 kHz - despite being only DSD64 recording using microphone that is "specified" to 20 kHz - which in practice obviously is exceeded. The levels are LOW - I will post a video on YT after I get my video software up and running - I am NOT photo and video guy, I treat both as the necessary evil. I will have to go trough 5 years worth of recording to find other recordings that may include ultrasonics (from percussion, brass , strings and the like) .
 
Regarding instantaneous dynamic range of human hearing being only 30-40 dB - I am not so convinced about that. Any studies confirming it is really that low ? But it is a good question worth investigating further if any of the effects of ultrasonics are not being masked by this.

   
This guy is just looking for attention. I suspect there's reasons for that, but I'm not going to speculate on it. It might help if you could delete the thread. If you can't do that, change the title to take his name out of it. Having his name in the headline just feeds his neediness.

I am in no way needy of attention to myself - or my name. I would have created such or similar thread if I wanted to - by myself, long ago - wouldn't I ? I even prefer NOT to have my name in the title of the thread.
 
All I want is to put the audibility of the effects of ultrasonics on the map - and I do not require any personal endorsement, credit etc for it. 

 

[Steve Eddy]

Regarding the transducer - the microphone - - there are more options.

Ok. But it's not enough to just have response out to 100kHz, it also has to be very linear. If not, the microphone would produce its own intermodulation products which could overwhelm any intermodulation products produced in the ear. And how would you go about separating the two?

As to the level of the harmonics extending beyond 20 kHz - please see http://www.cco.caltech.edu/~boyk/spectra/spectra.htm

Yes, I'm familiar with Boyk.

Those ultrasonic harmonics are 60dB+ below the fundamental. So any intermod products would be below even that.

Regarding instantaneous dynamic range of human hearing being only 30-40 dB - I am not so convinced about that. Any studies confirming it is really that low ? But it is a good question worth investigating further if any of the effects of ultrasonics are not being masked by this.

I'll try and get some specific cites. This comes from a friend of mine who has a very very good understanding of how our hearing works. He's formerly a researcher at AT&T Bell Labs and was behind the development of the MP3 codec, which of course required a very good understanding of how we hear.

All I want is to put the audibility of the effects of ultrasonics on the map - and I do not require any personal endorsement, credit etc for it. 

Unless you can find a microphone and transducer that's very linear at those ultrasonic frequencies, perhaps the best way to explore this would be with a real trumpet and some sort of absorptive material that starts to kick in around 20kHz, which could be placed over the bell. Then do some controlled, blind listening tests.

se

 

[analogsurviver]

Unless you can find a microphone and transducer that's very linear at those ultrasonic frequencies, perhaps the best way to explore this would be with a real trumpet and some sort of absorptive material that starts to kick in around 20kHz, which could be placed over the bell. Then do some controlled, blind listening tests.

se

Only regarding the microphone - I said the other is a Bruel A Kjaer Measuring microphone - presumably very linear to 100 kHz. Exact type will be known on Monday.
 
It is a good suggestion to use some absorptive material that kicks in at  select frequencies. Trouble is, it is not ABXable for others - one can not stash that into a computer for other members to test on their end of the globe. Which brings us to square one.
 
Never had the "pleasure" of sitting behind a similar, even if less abundant  "clad" woman at the concert

 ? Half the sound, not only ultrasonics - gone ...