[gregorio]

According to the laws of physics, at least in this universe, at 20 degrees Celsius, at sea level, at 50% relative humidity, the attenuation at 30 kHz is 0.94 dB / meter.
That means that, at six feet, the attenuation is just under 2 dB. Unless you were referring to the laws of physics in some other universe - that's what the ones in this universe say.

Many audiophiles make claims which are ridiculous because they not only simply ignore all the laws of physics but actually contradict them. You on the other hand are quoting one of the laws of physics, but it's still as effectively invalid because you are ignoring all the other pertinent laws of physics and then FALSELY asserting (twice) "that's what the ones in this universe say". Unless you're living in a different universe, that is absolutely NOT what the laws of physics in this universe say. So not only is your statement FALSE but as I've posted in response to you more than once what the relevant laws of physics are, and actually given you facts and figures, then you cannot claim ignorance or being "inadvertently" mistaken. This is the science forum and therefore pretty much the exact opposite of the "let's ignore and/or misrepresent the science (as suits our agenda)" forum!

For those interested in ALL the relevant facts/science:
We have to consider the practicalities/realities of recording, not just an individual law/rule of physics but the COMBINATION of variables, such as: What is being produced in the first place, what mics are capable of, how do they need to be employed in practice, what laws of physics apply and how do they apply. For example, a snare drum in a drumkit is virtually always very closely mic'ed, so if we're talking about recording from say 2-3m away (in the overhead mics) we not only have to take into account the inverse square law (which KeithEmo is TOTALLY ignoring) but also apply it appropriately, as the inverse square law deals with RELATIVE distance. @Arpiben and @Zapp_Fan - You might find this online calculator useful as it deals specifically with relative mic distance: http://www.sengpielaudio.com/calculator-distance.htm. If we enter 0.0254 (1 inch) into ref (mic) 1 and 2.5m into ref (mic) 2, we see that the SPL at 2.5m is almost 40dB (roughly 100 times) lower. Turning this around, if we had say 100dBSPL at 2.5m (which is not at all unreasonable for a snare drum hit), we would have 140dBSPL at the mic position of 1 inch. Now, we MUST COMBINE this fact with "what mics are capable of" and here we run into a problem because most studio condenser mics are only capable of a maximum level of around 130dBSPL, so we would get break-up distortion (and possibly physically damage the mic). For this reason, we would typically use a dynamic mic for the snare drum in a kit as they can cope with the high SPLs BUT, compared to condenser mics, dynamic mics have very poor high frequency response, typically with a roll-off starting around 9kHz-10kHz and listed with a response extending typically to around 15kHz (they may pick-up content above the freq response spec but at extremely reduced levels). The same is true of any instrument of course, not just the snare drum, although with other instruments we may have additional factors to consider (such as proximity effect for example). In ADDITION to the inverse square law, we have the physics rule/law of thermal noise which comes into play when mic'ing at a distance (even of 2.5m), which increases with frequency. Using the example above, to achieve the same output level from the mic at 2.5m (as at 1 inch) we would need to amplify the mic's output by 40dB, which would also increase the noise floor of the mic and mic-preamp by 40dB but in addition, at 20kHz we've got approximately 22dBm more thermal noise (power) than at say 100Hz. In addition again, we have the one law of physics KeithEmo did bother to quote (HF air absorption) and of course we also have to consider the amount of ultrasonic freq content we have in the first place, what the instrument is actually producing. In practice many instruments have no ultrasonic content but, as has been evidenced by the Boyk paper (which has been linked to several times), those that do typically only produce tiny fractions of a percent of their energy in the ultrasonic band and even only around 6% for the cymbals, which of all instruments probably produces the highest amount of ultrasonic content. So, we've got relatively little utrasonic content to start with, PLUS a higher noise floor.

If you "join the dots" of all this information: We pretty much always use condenser mics for the overheads when mic'ing a drumkit, because they have far lower SPLs to deal with than the close mics and they can more accurately capture the >9kHz content. Obviously though, the final sub-mix of the drumkit will not be comprised of only the overhead mics' output, it will also contain the close mics' output and commonly also the output of a room mic (at a much greater distance than 2.5m and therefore with significantly lower ultrasonic content). BTW, I've essentially already stated ALL of this in previous posts in this thread!

@gregorio's original assertion was essentially that:
"Even if there is a significant amount of 30 kHz being produced by the cymbal it won't matter because the air will absorb it all before it reaches the microphone."
He then went on to assert that the loss would be something like 50 dB at 20 meters. However, those numbers are not correct.

Do I really need to qualify my numbers with "in this universe"? I haven't quoted numbers for your alternate universe in which the only laws of physics that exist are the ones which support your agenda! In this universe those numbers ARE correct!

For everyone else (in this universe!): The numbers vary according to distance and other factors. For example, cymbals are not only used in popular music. In an orchestra for example the main mics will typically be around 10m from the cymbals so we'll get significantly more ultrasonic freq loss but we commonly we use a specific type of mic which is spec'ed to 14kHz and again, there will be multiple mic's (including room mics) all mixed with the main mics. However, the ideal audience position is at least 20m from the cymbals, probably closer to 30m. The ultrasonic freq loss at that distance plus the raised noise floor of the mic (if we were to record from that ideal position) would result in there not being ANY recordable ultrasonic content from the cymbals (or any other instrument).

And, at one point, when I suggested that we might try recording cymbals with a microphone whose response extends to 30 kHz...
[1] You started by replying that most studio mikes don't go that high. But later, when I offered some microphones that do, and some other folks also listed a few...
[2] You changed your story, and claimed that "of course you'd already tried this and found no benefit to doing it"...
[3] Note that I'd be perfectly willing to accept your conclusions if you'd recorded those frequencies, played them back for some listeners, and found that nobody could hear the difference, or just plain didn't like it.
[4] However, arguing about the laws of physics seems to be sort of a waste to me.

1. I stated at that time almost no studio mics went that high. Today, largely because of false marketing, there is more of a requirement to demonstrate the presence of content >20kHz and therefore more mics that cater to that requirement.
2. No, that is a blatant lie "inadvertent" error, which doesn't even make any sense according to your own quote. If "most studio mics don't go that high" then there must be a few which do (only one or two at the time), we used those which did, duh! Also, because even those that did weren't much good above 20kHz we also tried with other laboratory/test mics.
3. We did and they couldn't! Furthermore, as I've already made CLEAR, we are not talking about one test that I did, we're talking about numerous tests that thousands of independent sound engineers carried out all over the world and repeated various/numerous times.
4. What "arguing about the laws of physics"? There are NO rational arguments about the laws of physics, only irrational arguments by audiophiles who don't actually even know the laws of physics or snake oil salesmen trying to misrepresent them! Now that we've demonstrated that you were misrepresenting the laws of physics, all of a sudden it all "seems to be a sort of a waste of time to you" ... of course it does!

This whole discussion is really a sort of detour associated with @gregorio's assertion that we would have difficulty creating the required test signal.
(Or something like that.)

Or in fact nothing like that whatsoever. OK then: This whole discussion is really a sort of detour associated with @KeithEmo asserting that unicorns exist and pigs can fly (or something like that).

And, let's be a little bit more adversarial...
We seem to agree that many modern recordings sound bad...
And we seem to agree that there are reasons for this...
However, since you're the recording engineer, isn't it YOUR job to find those problems and fix them?

No, it's not! If consumers are demanding to pay only a fraction of a cent (or nothing at all) for their recordings and therefore labels and artists are only spending a fraction of the time and money on making recordings, how's that the job of a recording engineer to fix?

Therefore, from the seller's point of view, if they want to stay in business, they'd better either:
a) find some way to improve a mousetrap that their customers and competitors didn't think of (and convince them it's really better)
b) make a new model mousetrap, with fancier decorations, and a few fancy blinking lights, that doesn't actually do any better at catching mice, and simply convince their customers that it's better

If a mousetrap is already effectively perfect then all we're left with, as you say, is to make another mousetrap and "simply convince their customers that it's better", IE. Marketing! Falsely convince customers that their perfect mousetrap isn't in fact perfect by lying about what "perfect" means, lying about the laws of physics, lying about the realities and practicalities of mousetraps, etc.

[1] And one of the indicators of relative distance from the source to the microphone/listener is the attenuation of ultrasonics - the depth of a soundstage with bandwidth limited recording and/or reproduction will NEVER sound natural.
[2] And I had a trumpet ( albeit not muted - muted trumpet is specially rich in ultrasonics, beyond 80 khz ) and sax on less than 3 metre .
[3] As the gig has been partially amplified ( electric guitar, double bass, most - but not all - of the time trumpet )
[4] I chose to record only in DSD64 - which can be problematic regarding inherent DSD ultraosnic noise. Still, a pretty good picture of what is going on up to at least 48 kHz

1. Clearly that's an obvious contradiction. If one of the indicators of relative distance from the source is the attenuation of ultrasonics, then attenuating ultrasonics MUST sound natural!
2. A muted trumpet produces about 0.03% of it's energy above 20kHz, beyond 80kHz obviously far less and that's at just 1m, so how exactly is that "specially rich in ultrasonics beyond 80kHz"? And you stated that the trumpet wasn't muted, so your "specially rich" is actually nothing whatsoever!
3. And that amplification extended to 80kHz did it? What super-tweeters were they using at this live gig?
4. A pretty good picture of what exactly up to 48kHz? A pretty good picture of what an un-muted trumpet and an amplification system isn't producing?

We could solve so many of these utterly ridiculous assertions and arguments by simply having an "Alternate Universe Sound Science" forum!!

G

 

[Zapp_Fan]

@gregorio thanks for the detailed response. I think we all acknowledge the inverse square law applies to all sound (or non-sound like ultrasonics), I guess that should go without saying. But your point about the higher noise floor above 20khz is a good one. When I tried the "filter, then pitch down" trick to see what ultrasonics were there in those recordings posted previously, I have to say, they sounded like cymbals to be sure, but noisy as hell. I think the "it's mostly noise" problem would also give us headaches on playback, i.e. I would expect most listening equipment to have high noise floors in the ultrasonic band also, making the problem that much worse. At this point, even if I were a space alien with good hearing up to 40khz, I have to wonder if I would even bother with these noisy recordings and pieces of gear.

 

[KeithEmo]

I'm sorry... but I do see what's going on here...

I seem to be talking about scientific measurements....
While you seem to be trying to bring in a bunch of stuff that, while it may pertain to recording in many situations, is not directly related.

If I take two sources, in a totally open space, or an anechoic chamber (to eliminate outside factors)...
All frequencies will drop off EQUALLY due to the inverse square...
IN ADDITION TO THAT, at two meters, 30 kHz will lose an additional 1.8 dB due to air absorption.... (at 20 degrees C, at sea level, at 50% RH)
While low frequencies, for example 200 Hz, will lose virtually nothing due to air absorption....
(Therefore, my recording at two meters will be "down 2 dB at 30 kHz" compared to flat.)

And, IF YOU CAN FIND ENOUGH LEVEL TO RECORD AUDIBLE FREQUENCIES AT 20 METERS.....
The level at all frequencies will be significantly lower - based on the inverse square of the distance.
But, in addition to that loss in level, the level at 30 kHz will be an additional 18 dB lower than the level at 200 Hz.
(So, regardless of the overall level, it will then be "down 18 dB at 30 kHz relative to flat".)
Now, remember that we're talking about direct incident sound, so no room reflections, and no "acoustic contributions".
(I would definitely expect to see a much greater loss at 30 kHz if we take our measurements in a real room - and use a long enough window to include reflected sound and other "room contributions".)

Likewise, when we're talking science, we tend to look at things a bit differently.
For example, when I talk about measuring frequencies, I don't have to concern myself with the limitations of the microphone you would choose to use.
I have no reason to doubt it when you say that you would never mic a drum kit with a condenser microphone.
And I have no reason to doubt that the dynamic microphone you would choose, based on several considerations, like good sound quality, won't pick up 30 kHz.

HOWEVER, here's a cute little microphone: http://www.pcb.com/products.aspx?m=378A14
It has a rated frequency response from 4 Hz to 70 kHz (+/- 2 dB)
It also has an inherent noise of 50 dB, and a dynamic range of 173 dB.
From those numbers, my guess is that, with a proper preamp, it would have no trouble recording a snare drum, or a cymbal, from an inch away (or from two meters away).
And, at that range, it will record everything up past 70 kHz.
(So doing so is not only theoretically possible; it's actually possible with currently available equipment.)

It's not my problem if you don't happen to have one of these in your studio to take your measurements with...
And, as the recording engineer, it's your choice to choose not to use it if you don't think it sounds good...
HOWEVER, neither of those facts is relevant if we're discussing what's possible...

So, for the sake of SCIENTIFIC DISCUSSION.....
If I record that cymbal from two inches away... using that microphone... I will get a very accurate recording (+/- 2 dB from 4 Hz to 70 kHz).....
(We'll assume that I'm doing it in a free field, or an anechoic chamber, to exclude room acoustics and other outside influences.)
And, if I then do it again from two meters away, I will find that the level at 30 kHz is down by about 1.8 dB (compared to the level at 200 Hz)....
Therefore, I have shown that is is possible to do so, and, beyond that, that the equipment required to do so is currently available....
That satisfies the requirement of "proving that it's possible".... which is all that I stated to begin with.

And, while I've seen a wide variety of estimates, everyone seems to agree that cymbals produce sound well above 25 kHz.
(And breaking glass, which has been featured on more than a few albums, has measurable frequency components extending past 50 kHz.)

I am perfectly willing to concede to you things like....
- nobody ever does it that way
- no recording studio you know currently has the equipment required to do it
- it would probably sound bad anyway (or not sound especially good)
- it wouldn't be what somebody sitting in the audience would hear

I would again remind anybody coming in late that what we're discussing here is the possibility of recording cymbals and including spectral content at ultrasonic frequencies.
We are NOT discussing whether that would be audible or even whether it would be desirable... simply that it would be possible.
(We are discussing SCIENCE and NOT good recording technique.....)

I should also point out, just to make @bigshot happy, that, if we're talking about "results of test performed by thousands of recording engineers all over the world",
then it seems like we should have plenty of properly documented test data to look at. (Otherwise that would all just be "a lot of anecdotal evidence".)

Many audiophiles make claims which are ridiculous because they not only simply ignore all the laws of physics but actually contradict them. You on the other hand are quoting one of the laws of physics, but it's still as effectively invalid because you are ignoring all the other pertinent laws of physics and then FALSELY asserting (twice) "that's what the ones in this universe say". Unless you're living in a different universe, that is absolutely NOT what the laws of physics in this universe say. So not only is your statement FALSE but as I've posted in response to you more than once what the relevant laws of physics are, and actually given you facts and figures, then you cannot claim ignorance or being "inadvertently" mistaken. This is the science forum and therefore pretty much the exact opposite of the "let's ignore and/or misrepresent the science (as suits our agenda)" forum!

For those interested in ALL the relevant facts/science:
We have to consider the practicalities/realities of recording, not just an individual law/rule of physics but the COMBINATION of variables, such as: What is being produced in the first place, what mics are capable of, how do they need to be employed in practice, what laws of physics apply and how do they apply. For example, a snare drum in a drumkit is virtually always very closely mic'ed, so if we're talking about recording from say 2-3m away (in the overhead mics) we not only have to take into account the inverse square law (which KeithEmo is TOTALLY ignoring) but also apply it appropriately, as the inverse square law deals with RELATIVE distance. @Arpiben and @Zapp_Fan - You might find this online calculator useful as it deals specifically with relative mic distance: http://www.sengpielaudio.com/calculator-distance.htm. If we enter 0.0254 (1 inch) into ref (mic) 1 and 2.5m into ref (mic) 2, we see that the SPL at 2.5m is almost 40dB (roughly 100 times) lower. Turning this around, if we had say 100dBSPL at 2.5m (which is not at all unreasonable for a snare drum hit), we would have 140dBSPL at the mic position of 1 inch. Now, we MUST COMBINE this fact with "what mics are capable of" and here we run into a problem because most studio condenser mics are only capable of a maximum level of around 130dBSPL, so we would get break-up distortion (and possibly physically damage the mic). For this reason, we would typically use a dynamic mic for the snare drum in a kit as they can cope with the high SPLs BUT, compared to condenser mics, dynamic mics have very poor high frequency response, typically with a roll-off starting around 9kHz-10kHz and listed with a response extending typically to around 15kHz (they may pick-up content above the freq response spec but at extremely reduced levels). The same is true of any instrument of course, not just the snare drum, although with other instruments we may have additional factors to consider (such as proximity effect for example). In ADDITION to the inverse square law, we have the physics rule/law of thermal noise which comes into play when mic'ing at a distance (even of 2.5m), which increases with frequency. Using the example above, to achieve the same output level from the mic at 2.5m (as at 1 inch) we would need to amplify the mic's output by 40dB, which would also increase the noise floor of the mic and mic-preamp by 40dB but in addition, at 20kHz we've got approximately 22dBm more thermal noise (power) than at say 100Hz. In addition again, we have the one law of physics KeithEmo did bother to quote (HF air absorption) and of course we also have to consider the amount of ultrasonic freq content we have in the first place, what the instrument is actually producing. In practice many instruments have no ultrasonic content but, as has been evidenced by the Boyk paper (which has been linked to several times), those that do typically only produce tiny fractions of a percent of their energy in the ultrasonic band and even only around 6% for the cymbals, which of all instruments probably produces the highest amount of ultrasonic content. So, we've got relatively little utrasonic content to start with, PLUS a higher noise floor.

If you "join the dots" of all this information: We pretty much always use condenser mics for the overheads when mic'ing a drumkit, because they have far lower SPLs to deal with than the close mics and they can more accurately capture the >9kHz content. Obviously though, the final sub-mix of the drumkit will not be comprised of only the overhead mics' output, it will also contain the close mics' output and commonly also the output of a room mic (at a much greater distance than 2.5m and therefore with significantly lower ultrasonic content). BTW, I've essentially already stated ALL of this in previous posts in this thread!



Do I really need to qualify my numbers with "in this universe"? I haven't quoted numbers for your alternate universe in which the only laws of physics that exist are the ones which support your agenda! In this universe those numbers ARE correct!

For everyone else (in this universe!): The numbers vary according to distance and other factors. For example, cymbals are not only used in popular music. In an orchestra for example the main mics will typically be around 10m from the cymbals so we'll get significantly more ultrasonic freq loss but we commonly we use a specific type of mic which is spec'ed to 14kHz and again, there will be multiple mic's (including room mics) all mixed with the main mics. However, the ideal audience position is at least 20m from the cymbals, probably closer to 30m. The ultrasonic freq loss at that distance plus the raised noise floor of the mic (if we were to record from that ideal position) would result in there not being ANY recordable ultrasonic content from the cymbals (or any other instrument).



1. I stated at that time almost no studio mics went that high. Today, largely because of false marketing, there is more of a requirement to demonstrate the presence of content >20kHz and therefore more mics that cater to that requirement.
2. No, that is a blatant lie "inadvertent" error, which doesn't even make any sense according to your own quote. If "most studio mics don't go that high" then there must be a few which do (only one or two at the time), we used those which did, duh! Also, because even those that did weren't much good above 20kHz we also tried with other laboratory/test mics.
3. We did and they couldn't! Furthermore, as I've already made CLEAR, we are not talking about one test that I did, we're talking about numerous tests that thousands of independent sound engineers carried out all over the world and repeated various/numerous times.
4. What "arguing about the laws of physics"? There are NO rational arguments about the laws of physics, only irrational arguments by audiophiles who don't actually even know the laws of physics or snake oil salesmen trying to misrepresent them! Now that we've demonstrated that you were misrepresenting the laws of physics, all of a sudden it all "seems to be a sort of a waste of time to you" ... of course it does!



Or in fact nothing like that whatsoever. OK then: This whole discussion is really a sort of detour associated with @KeithEmo asserting that unicorns exist and pigs can fly (or something like that).



No, it's not! If consumers are demanding to pay only a fraction of a cent (or nothing at all) for their recordings and therefore labels and artists are only spending a fraction of the time and money on making recordings, how's that the job of a recording engineer to fix?



If a mousetrap is already effectively perfect then all we're left with, as you say, is to make another mousetrap and "simply convince their customers that it's better", IE. Marketing! Falsely convince customers that their perfect mousetrap isn't in fact perfect by lying about what "perfect" means, lying about the laws of physics, lying about the realities and practicalities of mousetraps, etc.



1. Clearly that's an obvious contradiction. If one of the indicators of relative distance from the source is the attenuation of ultrasonics, then attenuating ultrasonics MUST sound natural!
2. A muted trumpet produces about 0.03% of it's energy above 20kHz, beyond 80kHz obviously far less and that's at just 1m, so how exactly is that "specially rich in ultrasonics beyond 80kHz"? And you stated that the trumpet wasn't muted, so your "specially rich" is actually nothing whatsoever!
3. And that amplification extended to 80kHz did it? What super-tweeters were they using at this live gig?
4. A pretty good picture of what exactly up to 48kHz? A pretty good picture of what an un-muted trumpet and an amplification system isn't producing?

We could solve so many of these utterly ridiculous assertions and arguments by simply having an "Alternate Universe Sound Science" forum!!

G

 

[analogsurviver]

Since I do not see any @gregorio post while being logged into my account, only reply from the memory - regarding the recent jazz rcording.

1. ) All the sound has been a mixture of live , unamplified sound of acoustic instruments ( drums - no mic in sight, trumpet(s), double bass, saxophone ) and amplified sound ( trumpet, electric guitar, double bass ).
2. ) All the musicians/instruments were < 5 metres from my ears/microphones - solid brick wall after that. just behind the heavy curtains ). Totally acoustic sax and 2nd trumpet < 3 m away. Drum kit approx 4 m away. No silencers on trumpet(s) used that night, though.
As you can see, NO ultrasound amplified - or required to mic/amplify for that matter. Therefore, remark regarding speakers is moot.
3. ) We can - VERY WELL - tell if the sound source is close or far away. And trumpet is perhaps the best known example. First, because the further it is away, there will be less ulktrasonics and highs within the 20 k limit range.
And second, the further away, the less assymetric the waveform will be; any brass instrument can, in general, produce only over pressure ( it can not suck air inside itself/person playing it ) Real close miking will reveal there is almost no underpressure at the brass instrument nozzle - the output pressure is almost all in the positive pressure region. That's WHY frequency response to DC is required - to preserve this offset also in the recording. I know it takes TWICE the amplitude in recording it would take if the lows below say 15 Hz or so would be rolled off ... but the realism would be lost.
Yes, some better players can play their brass also by sucking - but that is VERY low level, totally incomparable to SPL of normal playing. On the order of some 20 dB difference - if not more.

OK - worst case scenario attenuation of 30 and 40 kHz for 3/4/5 metres ? Brass instrument "exhaust" to mic distance , direct path, no obstacles whatsoever. How many dBs ?
Whatever the correct answer, it can not get low enough to be possible to neglect.

 

[KeithEmo]

My guess is that, someday in the future, we'll end up with the obvious end game for binaural.

Several "reference listeners" will be seated at various places in the audience - fitted with bionic implants.
The actual nerve impulses that they hear will be recorded.
And you or I will be able to plug directly into their actual experience.
("Would you prefer to be John sitting front and center or Mary nearer the back?")

I believe someone did a sci-fi movie about that..... but they included recording the entire experience.....
So the result was more like "remembering you were actually there - even though you weren't".

I'm not suggesting that we should worry about this now - because there's nothing much we could do about it.
But I'm old enough to remember when most people thought cassettes were good enough that "there was no point in bothering to improve them".
(I actually heard a recording once, transcribed from an old Edison cylinder, where the narrator exclaimed that "listening to the cylinder was just as good as being there".)
I guess that, luckily, not everyone believed him.

I agree with you that I may be a bit optimistic...
The current trend is more towards streaming, and earbuds, and listening on a smart phone.... so convenience over quality...
Honestly, if I had to predict what we'll have in 100 years...
I'm guessing that we'll have full video and sensory recordings - perhaps with that direct plug going to our brains...
And virtual audio frequency response from DC to infinity will just be "an incidental benefit of the technology".

so not only do we have to adjust to tech that doesn't yet exist with the idea that it will, we also have to expect human mutation and cyborgs just so that we can at long last justify having ludicrous ranges of recording and levels of fidelity, and be able to say "see I told you those were significant for humans!".
if we're just a little bit rational when thinking about effectively pushing the limits of fidelity and perception, I believe that it should be pretty obvious that what was recorded so far will be treated like old low quality wax cylinders are treated now. transferred to new techs only for the sake of preservation. the moment we get rid of microphones and start using whatever we will use instead. 3Dcameras at stupid speeds capturing the mechanical actions of the instruments that will have been previously scanned, and then simulating the exact sound instead of trying to record it and lose a lot in the process, maybe interferometry to collect the signals coming right out of an instrument or somebody's mouth... who knows? it will be the future by then. or maybe experiencing music would bypass our senses and be plugged straight into out brain in a completely different form. all of that is already probably possible with various degrees of accuracy, so why not have it one day with massive levels of accuracy? seems to me that it could be science and not just fiction someday.
but again, most people then would be using music or whatever experience made for and with modern techs. because they will be superior, and because they will follow the trends of the period.

the people who will still be interested in the old high res albums, will be the same who are now still passionate about vinyls and tube amps with 4%THD. fidelity by the time it comes will no longer define those products and will no longer be the reason why somebody cares to have them. that's how I imagine things at least. yes the future could bring many surprises(including those that would challenge how loud we can hear the moment before getting vaporized), and yes different people will have different interests. but it seems completely unrealistic to believe that someone is really going to bother trying to poorly reconstitute data from high res records that would conveniently happen to now have some important value in the "experience" of music. I just don't believe that many people would bother if it happened. in a world where you often can't even trace back the tracks used in a master, let alone have any relevant information about the gears used or the processed applied while mastering, the day we get better techs, we'll make use of them to the max and forget about most of the old stuff. the amount of content keeps increasing, we already don't have enough of a lifetime of binge watching to get around all the movies and TV shows, the amount of new music is, at least to me, quite overwhelming. everything I observe points toward losing interest for the old stuff, and toward calling things old a lot sooner than the previous generations did.

I think you are once again being unrealistically optimistic with all this.

 

[Zapp_Fan]

I guess if we want to assume the technology from Total Recall or The Matrix becomes real at some point, then we should definitely be doing ultrasonic recordings now, just in case.

 

[analogsurviver]

I guess if we want to assume the technology from Total Recall or The Matrix becomes real at some point, then we should definitely be doing ultrasonic recordings now, just in case.

From DC to 100 kHz ( or few Hz, < 5 Hz to at least 70 kHz ) is possible for at least three years.

 

[KeithEmo]

I agree entirely.

Personally, when I visit a very small venue, I don't enjoy sitting three feet from the band... and one reason is that the cymbals are unpleasantly loud and bright.
And I wouldn't even argue if someone were to suggest that rolling off the cymbals sounds better to most people.
Those are the kind of artistic decisions that the recording engineer is expected to make.

However, as a discussion of pure science, that isn't the point.

First off, we were discussing whether it was possible to record those 30 kHz overtones, and not whether it was desirable.
I'm perfectly happy that I didn't get a terrible sunburn from those bright scenes in The Martian.
However, from a technical perspective, my TV and the video would have more accurate if I had.
Maybe it would have been nice if my TV had a setting for "pleasant" and "LAB".

Of course, since the movie was really filmed in a studio, and not on Mars, those frequencies probably really weren't there anyway.
Therefore, a camera that recorded the scene on the set with perfect accuracy would still not be "accurate to reality"; it gets complicated.
I'm also pretty sure that different cymbals make wildly different amounts of sound at different frequencies.

However, I still recall the first time I heard Tommy... by The Who.
There's one scene where the Gypsy Queen smashes a mirror to startle Tommy - who is staring at his reflection.
I don't recall ever hearing a version where that smashing mirror sounded like real glass.
(So I guess there was room for a bit of improvement there. Could it be those missing ultrasonic overtones?)

However, from a scientific perspective, there is another issue.
I'm not convinced that it's reasonable to assume that "most people prefer the cymbals rolled off".
And I'm also not convinced that: "nobody EVER notices ANY difference, with ANY recording, of ANY sound, under ANY circumstances, if you roll it off at 20 kHz or at 30 kHz."
That covers an unreasonably large amount of ground...
And, if we want to find out whether people prefer a recording that's flat to 30 kHz to one that isn't, we need test samples of both for them to listen to.
It's not reasonable to simply say: "That's what we're giving them now and they seem happy with it."
People frequently find new things to be preferable after they try them.

And, by the way, HAS anyone ever actually tested whether the sound of breaking glass sounds more realistic when you include frequencies up to 50 kHz?
We could blindfold some people, alternate some recordings of glasses breaking, and break some real glasses, and see if they can tell the difference.
(And see if they can tell the difference more often when we apply a sharp 20 kHz low-pass filter.)

A lot of what I keep hearing on this forum lately boils down to:
"We already know that people can't hear or be affected by ultrasonic frequencies, ever, under any normal listening conditions, so there's no point in doing a test to find out if they can or not."

From a scientific point of view that's a real stretch.

@gregorio thanks for the detailed response. I think we all acknowledge the inverse square law applies to all sound (or non-sound like ultrasonics), I guess that should go without saying. But your point about the higher noise floor above 20khz is a good one. When I tried the "filter, then pitch down" trick to see what ultrasonics were there in those recordings posted previously, I have to say, they sounded like cymbals to be sure, but noisy as hell. I think the "it's mostly noise" problem would also give us headaches on playback, i.e. I would expect most listening equipment to have high noise floors in the ultrasonic band also, making the problem that much worse. At this point, even if I were a space alien with good hearing up to 40khz, I have to wonder if I would even bother with these noisy recordings and pieces of gear.

 

[bigshot]

This is the science forum and therefore pretty much the exact opposite of the "let's ignore and/or misrepresent the science (as suits our agenda)" forum!

And this is the Testing Audiophile Claims and Myths thread, not the "thinking up excuses why we don't have to bother to test anything and wait for someone else to do it for us" thread.

When I tried the "filter, then pitch down" trick to see what ultrasonics were there in those recordings posted previously, I have to say, they sounded like cymbals to be sure, but noisy as hell. I think the "it's mostly noise" problem would also give us headaches on playback,

At a super audible frequency and attenuated that much, I doubt there would be any effect on you at all. If you turned up the volume high enough that the sound pressure was detectable, you would get a headache, but you would have to be careful not to incur hearing damage while you do that.

In practice, inaudible is just plain inaudible. Nothing more, nothing less.

I never knew that "pure science" consisted of counting how many angels dance on the head of a pin.

 

[bfreedma]

I agree entirely.

Personally, when I visit a very small venue, I don't enjoy sitting three feet from the band... and one reason is that the cymbals are unpleasantly loud and bright.
And I wouldn't even argue if someone were to suggest that rolling off the cymbals sounds better to most people.
Those are the kind of artistic decisions that the recording engineer is expected to make.

However, as a discussion of pure science, that isn't the point.

First off, we were discussing whether it was possible to record those 30 kHz overtones, and not whether it was desirable.
I'm perfectly happy that I didn't get a terrible sunburn from those bright scenes in The Martian.
However, from a technical perspective, my TV and the video would have more accurate if I had.
Maybe it would have been nice if my TV had a setting for "pleasant" and "LAB".

Of course, since the movie was really filmed in a studio, and not on Mars, those frequencies probably really weren't there anyway.
Therefore, a camera that recorded the scene on the set with perfect accuracy would still not be "accurate to reality"; it gets complicated.
I'm also pretty sure that different cymbals make wildly different amounts of sound at different frequencies.

However, I still recall the first time I heard Tommy... by The Who.
There's one scene where the Gypsy Queen smashes a mirror to startle Tommy - who is staring at his reflection.
I don't recall ever hearing a version where that smashing mirror sounded like real glass.
(So I guess there was room for a bit of improvement there. Could it be those missing ultrasonic overtones?)

However, from a scientific perspective, there is another issue.
I'm not convinced that it's reasonable to assume that "most people prefer the cymbals rolled off".
And I'm also not convinced that: "nobody EVER notices ANY difference, with ANY recording, of ANY sound, under ANY circumstances, if you roll it off at 20 kHz or at 30 kHz."
That covers an unreasonably large amount of ground...
And, if we want to find out whether people prefer a recording that's flat to 30 kHz to one that isn't, we need test samples of both for them to listen to.
It's not reasonable to simply say: "That's what we're giving them now and they seem happy with it."
People frequently find new things to be preferable after they try them.

And, by the way, HAS anyone ever actually tested whether the sound of breaking glass sounds more realistic when you include frequencies up to 50 kHz?
We could blindfold some people, alternate some recordings of glasses breaking, and break some real glasses, and see if they can tell the difference.
(And see if they can tell the difference more often when we apply a sharp 20 kHz low-pass filter.)

A lot of what I keep hearing on this forum lately boils down to:
"We already know that people can't hear or be affected by ultrasonic frequencies, ever, under any normal listening conditions, so there's no point in doing a test to find out if they can or not."

From a scientific point of view that's a real stretch.

You work for a relatively large company in this industry. Why don’t you execute some testing? Are you unwilling, or is your company not equipped to perform these tests in order to identify audio elements you so strongly believe are important?

 

[Zapp_Fan]

You work for a relatively large company in this industry. Why don’t you execute some testing? Are you unwilling, or is your company not equipped to perform these tests in order to identify audio elements you so strongly believe are important?

Earlier ITT we discussed the fact that most audiophile-oriented companies don't have any logical incentive to publish results on these topics. If consumers already believe the products have high performance, then proving ultrasonics are important won't affect sales. Proving ultrasonics aren't important would nuke their sales. Most companies that could afford to run the tests fall into the category of "consumers believe they have high performance"... so you see why nobody actually does the work.

Also a really rigorous, put-the-final-nail-in-the-coffin test would be expensive. My guess is no less than $250K to do it truly properly. $250K is not chump change for most audio companies, particularly for a project that is only of academic interest.

Keith (or I, or some other MOTT) could spend a work day at his desk testing these things with... himself... probably, but there isn't much point in doing that, it wouldn't be considered authoritative by anyone.

Others have also pointed out that, according to any reasonable approach to audio among academics, recording formats are basically a solved problem. The importance of frequencies over 20khz is - if anything - very low. There are more pressing research problems in the world of human hearing, like figuring out how to treat tinnitus, or how to prevent hearing damage, things like that.

 

[Zapp_Fan]

At a super audible frequency and attenuated that much, I doubt there would be any effect on you at all. If you turned up the volume high enough that the sound pressure was detectable, you would get a headache, but you would have to be careful not to incur hearing damage while you do that.

I meant headaches in the sense of "problems if you're trying to actually reproduce ultrasonic recordings", not a literal headache. But yeah, if I'm getting a headache from an actual ultrasonic noise I imagine other things are going wrong too.

 

[bfreedma]

Earlier ITT we discussed the fact that most audiophile-oriented companies don't have any logical incentive to publish results on these topics. If consumers already believe the products have high performance, then proving ultrasonics are important won't affect sales. Proving ultrasonics aren't important would nuke their sales. Most companies that could afford to run the tests fall into the category of "consumers believe they have high performance"... so you see why nobody actually does the work.

Also a really rigorous, put-the-final-nail-in-the-coffin test would be expensive. My guess is no less than $250K to do it truly properly. $250K is not chump change for most audio companies, particularly for a project that is only of academic interest.

Keith (or I, or some other MOTT) could spend a work day at his desk testing these things with... himself... probably, but there isn't much point in doing that, it wouldn't be considered authoritative by anyone.

Others have also pointed out that, according to any reasonable approach to audio among academics, recording formats are basically a solved problem. The importance of frequencies over 20khz is - if anything - very low. There are more pressing research problems in the world of human hearing, like figuring out how to treat tinnitus, or how to prevent hearing damage, things like that.

I understand and agree. That’s why I find it so interesting that a MOT keeps discussing testing. The use of “all things may be possible in the future” as an argument seems to be pure marketing. It’s the confluence of these two approaches within the same post that I find problematic.

 

[KeithEmo]

Fair question... and I'm going to give you an honest answer.
The answer is that, outside of academic interest, it really has little relevance to the design of our products.
(And, while we are in fact a relatively large company, we aren't large enough to have a division dedicated to academic research.)

Many of our customers and potential customers request that our products support digital audio files up to 24/192k.
It's simply "what people expect from current products".
The DAC chips we use in all of our high-end units support 192k.
We don't pay extra for it; we don't charge our customers extra for it; and we would not reduce our costs by limiting our equipment to 44k.
(In fact, if we wanted to limit our gear to 44k, it would cost us more to make the modifications necessary to specifically block higher rates from working.)

So, to be very blunt, our next DAC will support a 192k or higher sample rate, and there is no possible test result that will convince us to change that.
What possible reason would we have to eliminate a feature that some people actually ask for, and that costs nothing to include, whether we believe it's important or not?
And, that being the case, what motivation would we have to expend the time and effort necessary to do a reasonably valid test, if we're not going to do anything regardless of what we may find out?

Based on your comment, I would also advise you to read a few of my posts more carefully...
I DO NOT "strongly believe that high-res files sound better"...
What I strongly believe is that there isn't enough evidence to make a legitimate claim either way...
And, AS AN INDIVIDUAL, it always bugs me when people make strong and far-reaching claims based on flimsy and inadequate evidence.
(I simply cannot understand how some people can be so dead set on talking me out of paying a few dollars more for a file that I think might sound better.)

Just to be clear, as a company, Emotiva doesn't specifically have "an official company line" about whether high resolution files are better or not.
Our official standpoint is more like:
"We think our equipment should be able to play anything you ask it to - and make it sound as good as possible".
"Since a lot of our customers seem to want to play high-resolution files, it's fine with us."

I will point out that we do listen to a lot of equipment...
And many of us are quite convinced that our equipment sounds a lot better than the equipment made by many of our competitors.
Many of us are also quite convinced that, for example, different of our own amplifier and DAC models sound a tiny bit different.

You work for a relatively large company in this industry. Why don’t you execute some testing? Are you unwilling, or is your company not equipped to perform these tests in order to identify audio elements you so strongly believe are important?

 

[Zapp_Fan]

I understand and agree. That’s why I find it so interesting that a MOT keeps discussing testing. The use of “all things may be possible in the future” as an argument seems to be pure marketing. It’s the confluence of these two approaches within the same post that I find problematic.

I think the real problem is a disconnect on this proposition.

Keith (and I personally) and a few others find ultrasonics fun to think/talk about, regardless of whether there's any practical application today.

This is mostly because they're the only thing you can discuss materially improving in audio without getting deep into like... materials science and mechanical engineering. Back here on planet earth, 2018, as far as most reasonable people are concerned, the biggest bottleneck in audio reproduction today is transducers, (and acoustics) and talking about how to improve transducers beyond today's state of the art is frankly above my pay grade. And acoustics is sort of boring and out of place on this forum because it's more of a woodworking project than "audio" most of the time.

Others find them annoying to think about or talk about because there's no practical application today. So those others argue against Keith as if he's advocating for something in particular, when really he's just (as I read it) more on a "wouldn't it be interesting if..." sort of trip.