• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Can You Trust Your Ears? By Tom Nousaine

Status
Not open for further replies.

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
Hi.

I claim that we cannot hear above 20 kHz, so we will not notice if everything above 20 kHz is trimmed off or not.

We will notice, however, if our system suffers from intermodulation distortion from ultrasonic components, producing unwanted artefacts that come down into the audible range. By removing the ultrasonic components from the music we remove the unwanted artefacts - which may be registered as an audible difference from our system.

In other words, I am happy with CD sample rate.
Ok, but suppose that the process of trimming itself introduces unavoidable audible artifacts in the audible band? Then what?
 

Wayne

Active Member
Forum Donor
Joined
May 26, 2017
Messages
172
Likes
46
Location
Los Angeles, CA
This is not impossible, but how would you go about introducing audible artifacts?

The following assumes a HiFi system capable of producing frequencies >20kHz and a music source containing musical frequencies >20kHz

I once asked (actually several times) how could recorded sound above 20kHz have an effect on music. I was told that ultra sonic frequencies (> 20kHz) can combine with other frequencies, mainly those on the lower end (<20kHz), and harmonics can be produced in the audible range. Apparently much like two tuning forks of different frequencies can produce a different set of harmonics including a "beat frequency" (at a lower frequency than either).

I have no practical experience with sound so I have no idea if the above is accurate. Even if it is accurate (it does have a ring of truth) it may be an effect that only happens rarely or only under very specific circumstances.

Any comments as to the accuracy of the above would be appreciated.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
The following assumes a HiFi system capable of producing frequencies >20kHz and a music source containing musical frequencies >20kHz

I once asked (actually several times) how could recorded sound above 20kHz have an effect on music. I was told that ultra sonic frequencies (> 20kHz) can combine with other frequencies, mainly those on the lower end (<20kHz), and harmonics can be produced in the audible range. Apparently much like two tuning forks of different frequencies can produce a different set of harmonics including a "beat frequency" (at a lower frequency than either).

I have no practical experience with sound so I have no idea if the above is accurate. Even if it is accurate (it does have a ring of truth) it may be an effect that only happens rarely or only under very specific circumstances.

Any comments as to the accuracy of the above would be appreciated.

I am not j_j by any means.

What you are describing is intermodulation. Higher frequencies can intermodulate and create lower frequencies.

Let us suppose your hearing is limited to 20 khz. If it is above that in a live music situation you'll not hear it.

Just as there is some chance intermodulation could create frequencies below 20 khz you could hear live, a good microphone with 20 khz bandwidth would also pick it up on the recording.

So you don't need ultra sonic response to record the results of any such intermodulation (which are usually going to be low in level in the first place).

On the other hand, let us suppose you use wide bandwidth microphones that reach to 40 khz and a 96 khz sample rate that also reaches 40 khz. It is possible feeding that into your playback gear that at the amp or speaker or elsewhere there will be intermodulation distortion to be generated by your playback system. Then you could hear during playback a difference due to this added distortion that might not have been there were you hearing it live. So wider bandwidth recordings might sound different, but the difference isn't due to high fidelity.
 

Wayne

Active Member
Forum Donor
Joined
May 26, 2017
Messages
172
Likes
46
Location
Los Angeles, CA
What you are describing is intermodulation. Higher frequencies can intermodulate and create lower frequencies.

Thanks, I have wondered about that (turns out it was intermodulation - but I didn't know) since CDs came out and the discussions (arguments?) ensued about the CD sampling rate. I was told a lot of things and that was one of them. Problem solved
 

Wayne

Active Member
Forum Donor
Joined
May 26, 2017
Messages
172
Likes
46
Location
Los Angeles, CA
On the other hand, let us suppose you use

Blumlein 88, If you will indulge me, I would like to get a better understanding by drilling down on your reply.

An engineer is recording music for future playback; under what circumstances would he select, or what advantage would be gained:

1. ... using a 40kHz microphone and a 96kHz sample rate compared to a 20kHz microphone/44.1 sample rate?

2. ... using a 40kHz microphone and a 44.1kHz sample rate compared to a 20kHz microphone/44.1 sample rate?

3. ... using a 20kHz microphone and a 96kHz sample rate compared to a 20kHz microphone/44.1 sample rate?

Thanks
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
I don't know any good reasons for any of the three.

#1. A philosophical position of maximum fidelity. If instruments make sound at 40 khz (and some do), and you wish to capture all that you can as an effort at maximum fidelity to the physical sound, you might do this.

In the past, some digital processing worked better without artifacts or with lesser artifacts when done at 96 khz then released at 44 khz. That day has passed I do believe.

#2 is easy. Unlike digital responses, few analog devices, especially transducers just die immediately at some frequency. They roll off or have resonances of various kinds. So a wide bandwidth microphone might maintain much more linear accurate response if it works to 40 khz even for 20 khz recordings.

#3 I can't think of a good reason. Maybe a given 20 khz microphone has a great color one wishes to use. Like ribbons for instance. Not going to see too many ultrasonic capable ribbon recording mikes. On some voices or some instruments they have a sound, one that is pleasing, and if using other microphones on other instruments, using those for a 96 khz recording might make some kind of sense. Of course that is if the 96 khz recording made sense (which generally it may not). The other case is for people who worry about the audibility of filters. I think it an overblown non-issue for the most part exceeded only by the blather about jitter. If you think that filtering at 44 khz has audible consequences, you might record at 96 khz to push those filters well beyond audibility while knowing you only need a pretty good 20 khz response from your microphones.
 

j_j

Major Contributor
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
2,267
Likes
4,759
Location
My kitchen or my listening room.
I don't know any good reasons for any of the three.
In the past, some digital processing worked better without artifacts or with lesser artifacts when done at 96 khz then released at 44 khz. That day has passed I do believe.

Actually, there is often reason to upsample after capture at 48 or 44.1 if you are going to do anything expressly nonlinear to the signal. You must do any nonlinear processing at an oversampling rate above the highest power polynomial expansion of your nonlinearity if you are determined not to allow aliasing of your processing artifacts back down into the passband.

There's a section in Bob Katz' book about this, and somewhere I have a talk about the problems of nonlinearities. Clipping is a classic one here. Clipping of tone-like signals will introduce 3, 5, 7 harmonics, which if done to higher frequencies can alias right down into midrange (or bass if you're particularly unlucky). This is one of the reasons you just do not ever clip tonal signals digitally. The clipping is much harder, and does terrible things with the components that alias down.
#2 is easy. Unlike digital responses, few analog devices, especially transducers just die immediately at some frequency. They roll off or have resonances of various kinds. So a wide bandwidth microphone might maintain much more linear accurate response if it works to 40 khz even for 20 khz recordings.
What people call "analog" (i.e. a continuous time, continuous level analog of the original) systems most often have resonances and filter performance that works on an 'octave' or "decade" frequency scale, i.e. 6dB/octave/order for filters, and the like. This kind of response does not generally have fast rolloff at 20kHz. Among other things, this is why it is common to use digital (i.e. discrete time, discrete level analogs) filters for the antialiasing filter, after sampling at a higher rate that allows for a simple analog antialiasing filter.

Digital filters (both IIR and FIR) as generally constructed have rolloff characteristics on a linear frequency scale, as opposed to the log frequency scale of "analog" systems.

Basically, the native frequency mapping of a digital filter (z domain) is linear frequency, and that of an analog second-order section (s domain) is log frequency space. It is possible to use 'z' domain filters in the analog world (FIR's built on a chip using Surface Acoustic Waves and the like) but they are much less common. Needless to say, one can, within the bandwidth, create a filter in 'z' domain that kind of mimics the analog performance, but with the caveat that you can't go above fs/2.
#3 I can't think of a good reason. Maybe a given 20 khz microphone has a great color one wishes to use. Like ribbons for instance. Not going to see too many ultrasonic capable ribbon recording mikes. On some voices or some instruments they have a sound, one that is pleasing, and if using other microphones on other instruments, using those for a 96 khz recording might make some kind of sense. Of course that is if the 96 khz recording made sense (which generally it may not). The other case is for people who worry about the audibility of filters. I think it an overblown non-issue for the most part exceeded only by the blather about jitter. If you think that filtering at 44 khz has audible consequences, you might record at 96 khz to push those filters well beyond audibility while knowing you only need a pretty good 20 khz response from your microphones.

The choice of microphones depends both on the pattern (cardioid, etc) the mic provides, and the evenness of the pattern it is built to have. Ribbons, for instance, often have tighter patterns at higher frequencies, and provide a "mellow" or "smooth" sound due to the fact they reject more high frequencies off-axis. (they don't have to do this, many do) Part of miking an instrument is understanding what sound you want. This is part of the creative process as well as part of the documentary process, and I am not going to write that book this night.
 

danadam

Addicted to Fun and Learning
Joined
Jan 20, 2017
Messages
957
Likes
1,497
Just as there is some chance intermodulation could create frequencies below 20 khz you could hear live, a good microphone with 20 khz bandwidth would also pick it up on the recording.

So you don't need ultra sonic response to record the results of any such intermodulation (which are usually going to be low in level in the first place).
What happens in the following situation:
  • there is a source of frequencies A and B
  • they intermodulate and create frequency X
  • you record it with a microphone which is able to capture all of A, B and X
  • you play it back on a system which is able to produce all of A, B and X
Obviously the system will produce what was recorded, but will A and B intermodulate again and add to X?
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
83
Hi.

I claim that we cannot hear above 20 kHz, so we will not notice if everything above 20 kHz is trimmed off or not.

Not to pick nits, but hearing an audible effect of a tone or musical sound above 20 KHz is dependent on a number of things other than it presence or its high frequency.

One, you mention right away is IM, and that is of course right on. For example I can play a 20 KHz tone and not hear it, but if there is IM I can play a 19+20 KHz tone with the same level (ether RMS or peak) and potentially hear technical artifacts.

Another is masking. A sound whether simple or complex at a significantly (e.g. next critical band down) lower frequency is prone to mask sounds at higher frequencies.

In general but not exactly all of the time, the signals in musical sounds decrease in amplitude with frequency.

The potential maskers may be significantly higher in amplitude than the higher frequency sounds that they can mask.

The higher frequency sounds may be simply below the threshold of hearing at the higher frequency.

Furthermore, hearing is rolling off above 5 KHz, so even equal sized maskers at lower frequencies will tend to mask the components of the music at higher frequencies.

We will notice, however, if our system suffers from intermodulation distortion from ultrasonic components, producing unwanted artefacts that come down into the audible range. By removing the ultrasonic components from the music we remove the unwanted artefacts - which may be registered as an audible difference from our system.

High frequency IM is more common than one might think. My quick testing tool is a set of files with 19 and 20 KHz tones at different levels such as FS -1 dB, FS -3 dB, FS -6 dB and FS -10 dB. Easy enough to generate with many audio editors.

With the USB audio chip headphone amp plugged into this PC based on a modern ESS DAC, I obtain clearly audible white noise @ FS -1 dB and FS -3 dB. I can mitigate this by turning down the gain on the PC - it is present with the PC gian at 100 and 90, but disappers wtih a gain of 71% and below. It also disappears for test signals @ FS - 6dB. With other combinations of levels I can also obtain the expected artifact @ 1 KHz which is of course clearly audible.
 

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
This is not impossible, but how would you go about introducing audible artifacts?
Is the reconstruction filtering at 22k Hz for RBCD always free from audible artifacts in the audible band both in the frequency and time domains? Good engineering can minimize these effects, but is this better than using a higher sampling rate and employing gentler filters at frequencies much further above audibility? Why then is it commonplace today to employ hirez sampling rates with ultrasonic response in recordings and a to d even for music intended for release on CD?

I do not disagree with the theoretical premise that ultrasonic signal or noise at ultra high sampling rates might cause intermodulation distortion products to "beat down" into the audible band. But, does it do so measurably and audibly in many/most systems? I have not seen conclusive proof that it is a significant problem. In the mean time, I prefer to listen to hirez recordings at hirez sampling rates. Actually, I find the sound of hirez downsampled to even 44k/24 bit to be consistently less preferable. It is not night/day and the music can still be enjoyed, but the difference is noticeable to me.
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
83
Is the reconstruction filtering at 22k Hz for RBCD always free from audible artifacts in the audible band both in the frequency and time domains?

Of course not! There are few things that someone hasn't figured out how to screw up. And in times past a few of these even made it into the production of mainstream audio gear. And as surely as we have SET amplfiiers, similar anti-engineering has ben applied to boutique digital gear. Even AES papers have been published based on making this kind of mistake and trying to trick people into presuming it was mainstream and common.

Good engineering can minimize these effects, but is this better than using a higher sampling rate and employing gentler filters at frequencies much further above audibility?

Yes. The reasons for this need not be belabored, right?

Why then is it commonplace today to employ hirez sampling rates with ultrasonic response in recordings and a to d even for music intended for release on CD?

Is it? I'd like to see an actual survey based on a significant number of real-world practitioners relating to this.

I do not disagree with the theoretical premise that ultrasonic signal or noise at ultra-high sampling rates might cause intermodulation distortion products to "beat down" into the audible band.

It is not just theoretical. It shows up in the professional literature. I've encountered situations when it happened with gear that I was using. I've seen where it happened to others or was likely to have.

But, does it do so measurably and audibly in many/most systems?

It definitely happens in some real world systems that well-meaning people assemble for their own use. Try the test tones from my post of 10:03 am EDT today. Note that they aren't even truly ultrasonic. I suspect my "keys jangling" sample can activat this problem, if and when it exists.

I have not seen conclusive proof that it is a significant problem.

What does significant mean? Especially in a global sense? It happens, people who are familiar with it recognize it and correct it.

In the mean time, I prefer to listen to hirez recordings at hirez sampling rates. Actually, I find the sound of hirez downsampled to even 44k/24 bit to be consistently less preferable. It is not night/day and the music can still be enjoyed, but the difference is noticeable to me.

Those are choices you get to make. No matter how you choose no innocent people or animals are molested or killed. ;-) These days, everybody with gear with any pretenses of quality probably thus already has gear that can handle unnecessarily high sample rates. It just comes that way! I don't see any purpose to trying to keep people from living on the wild side if that is what they wish to do. AFAIK most of the really good evidence against needlessly high sample rates is well known. You makes your choices. SACD and DVD-A already died as mainsteam formats and the careers that they damaged or destroyed are already water over the dam of history.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,836
Likes
16,501
Location
Monument, CO
It may (or may not) be worth noting that modulation of ultrasonic energy to the audio band can occur in speakers, not just the electronics driving them...
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
83
It may (or may not) be worth noting that modulation of ultrasonic energy to the audio band can occur in speakers, not just the electronics driving them...

Excellent point. As is usual, speakers are typically more nonlinear than just about any other kind of component in audio. Some published testing of the effects of ultrasonic signals used drivers dedicated to the ultrasonic range to avoid this problem, but of course, that did not eliminate IM of signals that are all in the ultrasonic range.
 

Wayne

Active Member
Forum Donor
Joined
May 26, 2017
Messages
172
Likes
46
Location
Los Angeles, CA
It may (or may not) be worth noting that modulation of ultrasonic energy to the audio band can occur in speakers, not just the electronics driving them...

Would this be a (good?) reason for limiting/eliminating ultrasonic frequencies (>20kHz) to the speakers?
 

Arnold Krueger

Active Member
Joined
Oct 10, 2017
Messages
160
Likes
83
Would this be a (good?) reason for limiting/eliminating ultrasonic frequencies (>20kHz) to the speakers?


CD players generally do a pretty good job of that. If you stick to mainstream recordings in mainstream formats, you are pretty safe.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Actually, there is often reason to upsample after capture at 48 or 44.1 if you are going to do anything expressly nonlinear to the signal. You must do any nonlinear processing at an oversampling rate above the highest power polynomial expansion of your nonlinearity if you are determined not to allow aliasing of your processing artifacts back down into the passband.

There's a section in Bob Katz' book about this, and somewhere I have a talk about the problems of nonlinearities. Clipping is a classic one here. Clipping of tone-like signals will introduce 3, 5, 7 harmonics, which if done to higher frequencies can alias right down into midrange (or bass if you're particularly unlucky). This is one of the reasons you just do not ever clip tonal signals digitally. The clipping is much harder, and does terrible things with the components that alias down.

I was aware of this just keeping my answer simple. At one time some DAW software would apply processing without upsampling. To my knowledge, all modern software does apply upsampling for such non-linear effects just for the reason you describe. It upsamples does the processing and downsamples the result. Which is why I said that day has passed. You can usually reduce or eliminate the upsampling in the DAW if you want the effect of aliasing which some say they like for some sounds.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
What happens in the following situation:
  • there is a source of frequencies A and B
  • they intermodulate and create frequency X
  • you record it with a microphone which is able to capture all of A, B and X
  • you play it back on a system which is able to produce all of A, B and X
Obviously the system will produce what was recorded, but will A and B intermodulate again and add to X?
Dependent upon the playback system.
 

j_j

Major Contributor
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
2,267
Likes
4,759
Location
My kitchen or my listening room.
It may (or may not) be worth noting that modulation of ultrasonic energy to the audio band can occur in speakers, not just the electronics driving them...

Indeed many tweeters are startlingly nonlinear at higher frequencies when pushed to their limits. There are a variety of "interesting" physical issues involved in shaking a soft bit of material around at 20kHz.

These days, I tend to stick to decent ribbon tweeters. They are now inexpensive and well-behaved, at least the ones I've used lately.
 
Status
Not open for further replies.
Top Bottom