[bigshot]

I think most of us listen to music. We use our equipment to play back music. Weird waveforms and ultra high frequency content are nice and all in theory. But what's the point of reproducing sound that our ears can't even reproduce? When it comes to reproducing the thing that really matters- music (even electronic music), 16/44.1 does a perfect job of it for human ears. That's all that matters for me.

 

[Dobrescu George]

triangle wave, really? where does that even come from?
this is just saying the same thing in a trick question. a triangle signal can be expressed as a series of sine waves, some of which will reach pretty high frequencies to perfectly represent the pointy part. if that frequency is bigger than 1/2 of the sample rate, Nyquist theorem is simply not followed. so you're really asking something like, can you encode 30khz with a 20khz band limiting? and of course the answer is no.
now will it make any difference in what you would hear? that's unlikely because the headphone and your ear will both low pass/band limit the signal anyway. which one will do it sooner depends on ears and gears. but as long as the file does band limiting at a higher frequency, we won't notice.

also you need to keep a clear view of what you're trying to achieve here. if you want perfect reproduction of any signals, then the highest resolution is obviously better. but as soon as you're talking human hearing, the limit is what we can hear. more is the same as a lot more and perfection was never invited.

You seem like a knowledgeable person.

A 30 kHz sound

A 25 kHz sound

A 40 kHz sound

A 50 kHz sound

Any of those can produce, depending on their loudness, any harmonics that would be audible to humans? I just want a straight answer.

I think most of us listen to music. We use our equipment to play back music. Weird waveforms and ultra high frequency content are nice and all in theory. But what's the point of reproducing sound that our ears can't even reproduce? When it comes to reproducing the thing that really matters- music (even electronic music), 16/44.1 does a perfect job of it for human ears. That's all that matters for me.

Now I am going to respectfully say that I want saw waves. For music. For Mindless Self Indulgence albums. I cannot guarantee whether you will like their music or not, but I do, and I want perfect sine waves because the synths they used to that and they have a vrey interesting texture. Seems that the better the top end is for a IEM / Headphone, the better those sound.,

I do not believe that bit depth or if music is Hi-Res will have any impact on that and I do believe that it still is within the range of what humans can hear, but having good frequency response above 10 kHz, so in 10 - 20 kHz really helps this music.

 

[pinnahertz]

You might not be aware of it, but there is no audio recorder that actually can accurately reproduce a 10 kHz square or triangular wave. No analog recorder, tape or vinyl, or digital system. Not even digital at 20/192!
Nor is there any need to record and accurately reproduce the theoretical test signal with infinite harmonics anywhere in the audio world.

 

[71 dB]

I once tried a simple multitrack project in Audacity. Tracking went well, mixdown...well I exported it to something else to preserve my hair.

I have never had problems with downmixing a multichannel project in Audacity.

 

[pinnahertz]

You seem like a knowledgeable person.

A 30 kHz sound

A 25 kHz sound

A 40 kHz sound

A 50 kHz sound

Any of those can produce, depending on their loudness, any harmonics that would be audible to humans? I just want a straight answer.

No. But when those sounds occur in nonlinear systems, intermodulation products occur that are audible. Those are defects though. Is that straight enough?

I do not believe that bit depth or if music is Hi-Res will have any impact on that and I do believe that it still is within the range of what humans can hear, but having good frequency response above 10 kHz, so in 10 - 20 kHz really helps this music.

Of course having response above 10 kHz to 20 kHz helps music, but it does not help you reproduce triangular or square waves.

 

[Dobrescu George]

You might not be aware of it, but there is no audio recorder that actually can accurately reproduce a 10 kHz square or triangular wave. No analog recorder, tape or vinyl, or digital system. Not even digital at 20/192!
Nor is there any need to record and accurately reproduce the theoretical test signal with infinite harmonics anywhere in the audio world.

I am not saying there is, but certain IEMs and headphones sound much better with Mindless Self Indulgence's music.

No. But when those sounds occur in nonlinear systems, intermodulation products occur that are audible. Those are defects though. Is that straight enough?

Of course having response above 10 kHz to 20 kHz helps music, but it does not help you reproduce triangular or square waves.

So... There are audible stuffs there.

As for triangular waves, I may not know what Mindless self indulgence employs, but it is synth stuff and I like it. and I know that it looks like it has a lot of really saw-y stuff if viewed under Audacity or an oscilloscope

 

[castleofargh]

Triangle waves: I must be spending too much time with synthesizers.

I didn't intend to be tricky. It seems I am misunderstanding what is meant by bandwidth limiting (hence my parallel question of what the "frequency" of a complex, or virtually non-repeating signal, really means); so the assumption is that such a limit is based on maximum sine-wave frequency (not maximum frequency of various wave shapes), presumably because sine waves are considered fundamental?

So then apparently a 20kHz sampler cannot reproduce a 10kHz square wave, triangle wave, saw wave, etc.

yup. because the sampling theorem is all about sine waves and they can be used to define the other signals. when we say 10khz square wave, it defines the repeating rate of the squared shape, not the frequency of all the sines that make up the square wave. a perfect square wave contains an infinite number of sines at increasing frequencies(and decreasing amplitude, probably one of the reasons why we don't miss them that much when some are missing). the lowest being indeed a 10khz sine(see Fourier for all the mathematical fun). so no band limited signal can reproduce it perfectly, simply because a strict limit doesn't go well with the concept of infinity.
and in real life aside from quantum theory, you'll be hard pressed to find anything that can go from one position to another one instantaneously.
so instead we tend to care for what's audible, and try to almost perfectly reproduce that part at least for music. when it's about internet signal transmission, we want our signal to change amplitude even faster to be able to send a lot more data in a given delay. for that, band limiting at 44khz would suck. different needs, different standards.

but to be clear, there is nothing wrong with increasing resolution and sample rate, as long at it doesn't lead to other issues. but we do have to keep in mind real life use starting with human limits. if our ear doesn't know what to do with the ultrasonic content of a signal, there is little benefit to making huge files just to have that extra content. just like there is little interest in having a TV that reproduces accurately ultraviolet frequencies or above. the eye is not going to notice anyway, and it could actually have some negative impact.

 

[castleofargh]

You seem like a knowledgeable person.

A 30 kHz sound

A 25 kHz sound

A 40 kHz sound

A 50 kHz sound

Any of those can produce, depending on their loudness, any harmonics that would be audible to humans? I just want a straight answer.

then I'm becoming better at pretending to know stuff. ^_^
if loudness isn't limited, then we can physically feel ultrasounds. that has been established. but the levels were not really applicable to typical musical content and humanly reasonable listening levels.

not sure if I really get what you're pointing at. are we considering a perfect DAC, amp and headphone? or do we account for the distortions they could produce themselves in the audible range when fed with high amplitude ultrasounds? because that's another can of worms.

 

[Jonman503]

All my music in is 24 bit

 

[Dobrescu George]

then I'm becoming better at pretending to know stuff. ^_^
if loudness isn't limited, then we can physically feel ultrasounds. that has been established. but the levels were not really applicable to typical musical content and humanly reasonable listening levels.

not sure if I really get what you're pointing at. are we considering a perfect DAC, amp and headphone? or do we account for the distortions they could produce themselves in the audible range when fed with high amplitude ultrasounds? because that's another can of worms.

Okay....

Yes, I think I was referring much more to what the audible distortions would be if we fed the drivers and the DACs / AMPs with ultrasonics rather than just have them exist, because besides the typical harmonics, I was wondering if having Hi-Res music with what would probably be Hi-Res noise, would in fact lead to some effects on the drivers / DACs / AMPs, especially if the said drivers were not optimized to go beyond 20 kHz, or the DAC or the AMP (given capacitor input, how the roll-off filter impacts the sound and all).

I am asking not for me, but to understand why people hear Hi-Res better. My music is not in hi-res and I won't judge music I'm not accustomed to...

I don't think the reason some IEMs present Mindless Self Indulgence with better subjective sound is related to ultrasonics then

 

[Darren G]

Personally I look at transducers. They have mass. They can only be accelerated/decelerated so fast. Likewise I look at my own ears, mass, limits, etc. Personally I stopped fretting that there are limits on the frequencies I can hear, and just want the frequencies I can to sound good.

 

[RRod]

Triangle waves: I must be spending too much time with synthesizers.

I didn't intend to be tricky. It seems I am misunderstanding what is meant by bandwidth limiting (hence my parallel question of what the "frequency" of a complex, or virtually non-repeating signal, really means); so the assumption is that such a limit is based on maximum sine-wave frequency (not maximum frequency of various wave shapes), presumably because sine waves are considered fundamental?

So then apparently a 20kHz sampler cannot reproduce a 10kHz square wave, triangle wave, saw wave, etc.

Sine waves are the fundamental units of decomposition for the Fourier transform. Nyquist is a theorem about Fourier transforms. The question it answers is "When can the Fourier transform of a function (and thus the function itself) be exactly reproduced from the discrete-time Fourier transform (DTFT)?" The answer it gives is "When the Fourier transform has bounded support (i.e. the function is bandlimited) and the sampling frequency is at least twice that bound."

Triangle/square/saws etc. are *NOT* bandlimited functions, and thus they violate Nyquist and are not perfectly representable by the DTFT, not at 44.1kHz, not at 96kHz, not at 1e100kHz. But that doesn't matter because no microphone (or natural phenomenon) ever produced a perfect triange/square/saw wave! Then add on top of that the capabilities of the human ear, and you find much fewer problems with audio than some people want you to believe.

As far as "virtually non-repeating signals", note that a large class of functions allow Fourier transforms, again certainly anything you're getting out of a mic.

 

[RRod]

Okay....

Yes, I think I was referring much more to what the audible distortions would be if we fed the drivers and the DACs / AMPs with ultrasonics rather than just have them exist, because besides the typical harmonics, I was wondering if having Hi-Res music with what would probably be Hi-Res noise, would in fact lead to some effects on the drivers / DACs / AMPs, especially if the said drivers were not optimized to go beyond 20 kHz, or the DAC or the AMP (given capacitor input, how the roll-off filter impacts the sound and all).

I am asking not for me, but to understand why people hear Hi-Res better. My music is not in hi-res and I won't judge music I'm not accustomed to...

I don't think the reason some IEMs present Mindless Self Indulgence with better subjective sound is related to ultrasonics then

What's missing is the level of the ultrasonic content. If it's low, then the IMD products should also be low unless you have some kind of unbounded distortion. The IMD products would also have to not be audibly masked by OTHER content, distortion or otherwise. So yes, if your benchmark is "can I play high intensity ultrasonic content without lower frequency content and end up with something audible", then, SURE! That isn't exactly what people who hear hi-res as 'better' are talking about, though…

 

[pinnahertz]

I have never had problems with downmixing a multichannel project in Audacity.

You need to experience a real pro DAW to understand how limiting Audacity is.

 

[pinnahertz]

All my music in is 24 bit

So...you up-sampled everything then?