[bigshot]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif Your understanding of harmonics is flawed. Only the 1st, 2nd, 4th, 8th, 16th, etc. harmonics are octaves of each other.

Yeah, I know that it isn't perfect octaves all the way. I was rounding off for simplicity's sake. What's the exact factor for volume dropoff for each harmonic level? Do you know that? That's the really important thing. It might actually be less than half.

The rule of thumb that I've always been told by audio engineers is that if you can get the first level harmonics perfect and the second pretty damn close, that's the battle right there. I don't know of any sound guys who ever sweated beyond two. Overtones would slide in the middle like the bologna in the sandwich!

See ya
Steve

 

[infinitesymphony]

Quote:

Originally Posted by bigshot /img/forum/go_quote.gif Yeah, I know that it isn't perfect octaves all the way. I was rounding off for simplicity's sake. What's the exact factor for volume dropoff for each harmonic level? Do you know that? That's the really important thing. It might actually be less than half.

Good point. I'm not sure about the levels, since they would vary from instrument to instrument, and even between different instruments in the same group (ex. Stradivarius vs. Guarneri violins [PDF]).

It's true that the musical information is located within the first few harmonics, but the extra harmonics and inharmonic overtones are what make instruments sound unique. Plus, with synthesis, it's possible to make sounds with overtones of any volume.

Relating it back to the thread... All of the harmonic distortion introduced by various pieces of gear could possibly add up to affect the sound quality, since the timbre of the instruments might be changed. One could argue that no recording perfectly captures an instrument, and it's a valid point, but let's not go there.

Edit: Err, wait, that's not what the thread's about... Really relating it back to the thread this time, I think that we should figure out what sort of information would convince most people one way or the other. For example:

1. Are we capable of measuring everything (or even the most important things) that might affect audio and its perception?
2. If so, and it's found that there are differences, are they audible?
3. If so, to whom are they audible, if anyone?

 

[Roam]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif Really relating it back to the thread this time, I think that we should figure out what sort of information would convince most people one way or the other. For example: 1. Are we capable of measuring everything (or even the most important things) that might affect audio and its perception? 2. If so, and it's found that there are differences, are they audible? 3. If so, to whom are they audible, if anyone?

1. We can measure most of the important things, though debate continues on what is & isn't important.

2. This will never be definitively answered as there's too much variation in hearing ability amongst humans. The ear-brain-perception system is a tricky thing to understand.

3. This can only be determined by actual testing.

 

[bigshot]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif Good point. I'm not sure about the levels, since they would vary from instrument to instrument, and even between different instruments in the same group

I'm pretty sure it's a specific factor. After the first harmonic, the volume drops a certain percent compared with the harmonic above it. I am pretty sure that by the third harmonic, it would not be audible above the fundamental and first two harmonics. By 4 or 5 levels, the frequency of just about any fundamental tone would likely be out of the range of human hearing anyway.

The timbre you are talking about is a result of the overtones, which certainly do vary from instrument to instrument.

See ya
Steve

 

[edstrelow]

There seems to be a basic point missing here. Do you buy based on your own perception, i.e. what you hear, even if mistakenly, or on what other authorities say is good , doesn't make a difference or whatever.

I will consider technical and scientific points about equipment, but in the end its my hearing that determines.

For all the blather in these forums about double-blind tests there is almost no good psychophysical testing (actually probably "no" good testing) on which to make decisions about audio issues.

By the way most sensitive tests of perception do not use randomized, constant stimulus techniques with groups of subjects because they are too crude. Rather you use within-subject comparisons with a small number of trained observers and long runs of testing. (Shades of golden ears!) My favorite of a cheesy looking, "unscientific" technique which works, is found in the Bekesy audiometer which uses an adjustable signal level, to find the threshold of hearing i.e. turn the volume down until you can't hear the signal anymore and then up again until you can. Interesting but no control condition. Didn't stop Bekesy getting a Nobel prize.

If it's my money I want it to sound good to me.

 

[infinitesymphony]

Quote:

Originally Posted by bigshot I'm pretty sure it's a specific factor. After the first harmonic, the volume drops a certain percent compared with the harmonic above it. I am pretty sure that by the third harmonic, it would not be audible above the fundamental and first two harmonics. By 4 or 5 levels, the frequency of just about any fundamental tone would likely be out of the range of human hearing anyway.

Overtones, partials, harmonics... When discussing audio, they all refer to the same things. Partials and harmonics are equivalent, and the first overtone occurs at the second harmonic. Overtones can be harmonic or inharmonic. So, I don't understand how there could be a hard and fast rule about the levels of the harmonics of every instrument. For example, bells are known to have an especially loud overtone series relative to the fundamental.

When you say harmonic, I think that you mean integer-multiple harmonics, but I still don't think that there's a rule about halving the volume even in that case. Besides, the other overtones are just as important. Consider that the 32nd harmonic of a high E string on a guitar occurs at roughly 10.55 kHz, well within the range of most people's hearing.

I made a little MP3 for fun. I used a tone generator to make sine waves using the harmonic series, starting with the fundamental (E-4: 326.63 Hz) at 100% and halving the volume with each subsequent harmonic until I had a total of 8 harmonics (7 overtones plus the fundamental). The beginning of the MP3 starts with all harmonics sounding, then I mute the harmonics starting with the highest one. Once all harmonics are muted, I leave out the fundamental and unmute the harmonic series in ascending order, then put the fundamental back in. Can you hear all eight harmonics once the fundamental is unmuted? If so, then at least eight harmonics are important. If I had chosen to halve the volume with each new octave (integer-multiple harmonics), then the harmonics in between would be even louder.

This example probably doesn't approximate a real guitar very well, and I apologize for the timing of the unmutes--I did them by hand with a huge amount of sound card delay, so some mutes/unmutes may hold for a bit longer than others.

Quick and dirty harmonic series example.

I also apologize for this post, which is only tangentially relevant to the topic... But I hope someone enjoys it, anyway.

 

[Leporello]

Quote:

Originally Posted by edstrelow /img/forum/go_quote.gif There seems to be a basic point missing here. Do you buy based on your own perception, i.e. what you hear, even if mistakenly, or on what other authorities say is good , doesn't make a difference or whatever.

Quote:

Originally Posted by edstrelow /img/forum/go_quote.gif If it's my money I want it to sound good to me.

These are good points. Everyone is entitled to choose his/her equipment by his/her own criteria without justifying the choices to others.

However, if I want to make factual claims about the sonic properties of the equipment in public, I must accept the fact that I will be asked for evidence backing my claims.


Regards,

L.

 

[Riboge]

While the exchange between infinitesymphony and bigshot that is going on is not directly on topic, I want to express my appreciation for the way they are discussing harmonics and for how interesting and informative it is. It is on topic by way of demonstration.

I wish infinitesymphony would elaborate on the questions he felt were on topic. For instance, I don't quite get how 2) relates to 1) or what it means exactly.

 

[bigshot]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif When you say harmonic, I think that you mean integer-multiple harmonics... Consider that the 32nd harmonic of a high E string on a guitar occurs at roughly 10.55 kHz, well within the range of most people's hearing.

Yes, I'm talking about integer multiple harmonics. (The way I use the terms is harmonics for the major levels and overtones for the stuff inbetween.) Can you break down the frequencies of the harmonics on the e string of the guitar from fundamental on up for me? I'm confused. You're talking about integer multiple harmonics there too, right?

Thanks for the rest of the post. I'll come back and pore over it when I get some time to spare.

Thanks
Steve

 

[infinitesymphony]

Quote:

Originally Posted by Riboge /img/forum/go_quote.gif Quote: Originally Posted by infinitesymphony 1. Are we capable of measuring everything (or even the most important things) that might affect audio and its perception? 2. If so, and it's found that there are differences, are they audible? 3. If so, to whom are they audible, if anyone? I wish infinitesymphony would elaborate on the questions he felt were on topic. For instance, I don't quite get how 2) relates to 1) or what it means exactly.

Number one asks, can we objectively and accurately measure all or most of the qualities that might affect audio reproduction and signal transmission? There are many possibilities...

Number two asks, even if there are measurable differences that are shown to affect the final resulting audio reproduction, are these differences of a magnitude perceptible to the human ear? This is a very difficult step to measure due to the physical variation between individuals' hearing systems (i.e. the physical dimensions of an ear's components can have huge variations, meaning that not everyone will hear the same way), their abilities to listen analytically, and due to possible nonlinearities in the ear's design and frequency response (ex. aural harmonics).

Quote:

Originally Posted by bigshot /img/forum/go_quote.gif Yes, I'm talking about integer multiple harmonics. (The way I use the terms is harmonics for the major levels and overtones for the stuff inbetween.) Can you break down the frequencies of the harmonics on the e string of the guitar from fundamental on up for me? I'm confused. You're talking about integer multiple harmonics there too, right?

No, for the example I used the audio/non-scientific definition of harmonics. This harmonic series calculator shows the frequencies that I used (fundamental freq. at E-4). The eighth harmonic was at 2637.02 Hz. As I noted in my post, if I had chosen to halve the volume with every octave (2nd, 4th, 8th harmonics), then the harmonics in between would have been louder than in the example.

Of course, the flaw with the example is that I'm not sure how loud each individual harmonic would be when played on a real guitar. The point is that changing the magnitudes of these middle- to higher-ordered harmonics could change the timbre of the sound. Thus, harmonic distortion, even in small amounts, might audibly affect the sound.

I understand that with your definition (integer-multiple harmonics), the eighth 'harmonic' of E-4 would be seven octaves higher than the fundamental frequency: 41,808.64 Hz. This is probably well above the range of hearing or even perception (if it's possible that humans can perceive ultra-high frequencies), so of course harmonics at this range would not matter a great deal, and with a real pitched instrument, the volume of this frequency probably would be approaching zero. This is why I think that it would be more beneficial to use the other definition of harmonics, so that we can examine what is happening to the sound in the audible range.

 

[Riboge]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif Number one asks, can we objectively and accurately measure all or most of the qualities that might affect audio reproduction and signal transmission? There are many possibilities...

Can we agree on what "all or most of qualities that might affect audio..." are since we cannot seem even to agree on what all or most of the qualities of audio reproduction are or if they are? Clearly your view is that what is measured/measurable is given reality and what we hear is some partial, faltering take on that, right? Isn't that the heart of what we are at odds about? Is the "reality" we are starting from music as we hear it produced live in a concert hall or studio or even a master tape of performance or synthesized 'performance' or what is reproduced via our audiophile equipment including cables; or is reality what non-human equipment measures and 'perceives' in the ways it is designed to do? Following from that, are we trying to find explanations for what we hear or to hear what we accept as the reality of the measurements taken? That as far as I can understand is what you go on to in:
Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif Number two asks, even if there are measurable differences that are shown to affect the final resulting audio reproduction, are these differences of a magnitude perceptible to the human ear? This is a very difficult step to measure due to the physical variation between individuals' hearing systems (i.e. the physical dimensions of an ear's components can have huge variations, meaning that not everyone will hear the same way), their abilities to listen analytically, and due to possible nonlinearities in the ear's design and frequency response (ex. aural harmonics).

It seems to me there are many things you could measure, and whether we or someone can hear them or not may not have much to do with what audiophiles are interested in as listeners or to what cables do or don't do.
It seems so much more fruitful to give those that hear differences in cables the benefit of the doubt to the extent that you look for what might be determinable/measurable to correlate with these differences--in contrast to starting with the huge number of things that might be measured and only eventually find or not find whether any of these measurements have to do or don't with the cable issue.

Another approach might be to study what can be measured simultaneously about what passes thru the cables, what difference are heard and not heard from identified hearers of difference and what can be measured in relevant brain areas in order to see where the correlations are.

 

[Riboge]

Big Brother=the moderator has seen fit to remove several posts from today's activity. While I can see why that was done and don't for the most part disagree, I do take issue with there being no mention of it by him. Ironically they were more on point--though over the top--than most of the posts in this thread of which I am the OP.

Who is the moderator of this forum anyway? Why is that not clearly stated in a place that is readily come upon?

I have wished that the moderator would be more active in keeping people to the topic and within restrictions such as no DBT, etc. I realize it is easier in a mindless way to settle for picking off nasty name-calling and the like but past grade school that hardly seems the most important kind of guidance or concern.

 

[bigshot]

Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif I understand that with your definition (integer-multiple harmonics), the eighth 'harmonic' of E-4 would be seven octaves higher than the fundamental frequency: 41,808.64 Hz.

Okie doke. We're on the same page now. How many levels of your definition of harmonics would fit between two of mine?

Thanks
Steve

 

[infinitesymphony]

Quote:

Originally Posted by Riboge Clearly your view is that what is measured/measurable is given reality and what we hear is some partial, faltering take on that, right? Isn't that the heart of what we are at odds about?

Not exactly... I think that the human experience of hearing is much more important than its measurable degree of accuracy, but that if we are going to draw conclusions about what effects sound reproduction, we should try to remove human error from the equation as much as possible.

Quote:

Originally Posted by Riboge /img/forum/go_quote.gif Is the "reality" we are starting from music as we hear it produced live in a concert hall or studio or even a master tape of performance or synthesized 'performance' or what is reproduced via our audiophile equipment including cables; or is reality what non-human equipment measures and 'perceives' in the ways it is designed to do? Following from that, are we trying to find explanations for what we hear or to hear what we accept as the reality of the measurements taken?

Those are good questions... The human ear isn't perfect, but neither is measuring equipment. There may be characteristics of the human ear that currently cannot be measured, or are not possible to measure because they are constantly changing--we're living organisms, after all. But still, there are areas where measuring devices can exceed human hearing, for example in the measurements of ultrasonic tones.

If we could measure everything and know that the measurements were accurate, then there would be no debate. The problem is that we can't measure everything and don't always know that the measurements are accurate.

Quote:

Originally Posted by Riboge It seems so much more fruitful to give those that hear differences in cables the benefit of the doubt to the extent that you look for what might be determinable/measurable to correlate with these differences--in contrast to starting with the huge number of things that might be measured and only eventually find or not find whether any of these measurements have to do or don't with the cable issue.

Yes, that makes sense. I was just speaking in general terms, but it might be better from a practical standpoint to examine one issue at a time with the same questions. Is there a difference, if so, why, and who is able to hear it? In other words, is it valid, what is the cause, and of what degree of importance is it?

Quote:

Originally Posted by Riboge Another approach might be to study what can be measured simultaneously about what passes thru the cables, what difference are heard and not heard from identified hearers of difference and what can be measured in relevant brain areas in order to see where the correlations are.

That would be pretty interesting. Hard to measure, but then you would have a group of known-good ears to use for testing. Really, the only way to prove whether a difference is audible is with a loooong string of double-blind tests with a decent-sized group of people. Some would argue that it's still not definitive proof, but at least it would be something. Objective measurements could fill in the details of what does or does not result in audible differences.

 

[infinitesymphony]

Quote:

Originally Posted by bigshot /img/forum/go_quote.gif Okie doke. We're on the same page now. How many levels of your definition of harmonics would fit between two of mine?

It depends on the two specific harmonics, since there are twice as many harmonics between every octave. I think I was wrong when I said "integer-multiple harmonics" before, since those are the harmonics from the audio definition. I'm not sure where you'd find octave-only harmonics.

Between 329.63 Hz and 41,808.64 Hz, there would be 126 possible harmonics.