Talking about lying, then changing to hypocrisy, appears to be ongoing avoidance of is the issue of why keep pointing out your age, when you do know better. Since you are obviously very smart, you know all this.
OK, HumbleDeer could be a 50 year old male. It does not really matter. Let's go on please.
HumbleDeer, please think of starting a thread on specific subjects to get good answers. Piling up questions in a general thread is not the optimal way to do stuff. At least not for most species of the male persuasion 🙂
HumbleDeer, please think of starting a thread on specific subjects to get good answers. Piling up questions in a general thread is not the optimal way to do stuff. At least not for most species of the male persuasion 🙂
A few comments with regard to distortion in audio systems:
The amount of THD is an almost meaningless specification. The cynical might feel that is why manufacturers are so enamoured with publishing it ... but I digress.
What is important is the spectrum of distortion ... the 2nd, 3rd, 4th, 5th, etc harmonics. We tolerate rather large amounts of 2nd and 3rd Harmonic Distortion ... the 2nd, 3rd (and 4th) are common components of sound found in nature, and in musical instruments. The ear canal itself has a relatively significant amount of low order harmonic distortion. Thus our brains routinely dismiss this as an aberration.
Variations in the harmonics are what makes a violin and a piano sound different when playing the same fundamental ... A 440 (Hz), for example.
Distortions ... variations in the frequencies found in nature ... can be measured as a % but if these change the character of an instrument ... say, enough to make a Steinway sound more like a Yamaha Piano but not enough to make a Steinway sound like a Stradivarius Violin ... they are not going to be obviously noticeable to most people.
The higher order harmonics, and in particular the odd higher order harmonics, however, are a different story. Our brains are quite sensitive to these "extra" treats added by our sound systems, and we can detect them in percentages much lower than what we would find objectionable in a 2nd harmonic distortion.
Music is nothing if it's not mathematics and mathematical relationships. Many of us claim to be poor at math, but the truth is our brains are very, very good at math and don't require our conscious approval to do complex calculations in real time. However our brains use energy, and our bodies are designed to always minimize energy costs. In other words, exams are hard because they really are hard, they really do use more energy than "vegging out", they really do tax the human system.
So our brains are always looking for ways to minimize the energy cost, and this becomes important when we discuss music, audio and distortion. Which, let's not forget, is supposed to relax us, or at least not irritate us.
"THD" becomes a jumble of some important, and some moderately irrelevant distortion levels; we can't glean much about the sound quality of a component from that specification.
Which brings us to another aspect. High Order Harmonic distortions tend to add information that was not present in the original ... in other words Harmonic Distortion ... and of course these show up high in the audible spectrum. If we look at perhaps the 8Khz to 20 Khz area, we will find some of these higher order distortions; perhaps the 7th, or 9th harmonic of an instrument's fundamental note, for example.
They tend to add a bright tonal character to an audio system ... no surprise there; it's added "sound" in the higher registers. In auditory tests (which I won't cite as I don't personally need to revisit them, but they are available to someone with reasonable search skills to discover) ordinary listeners tend to prefer the sound of this higher order distortion to a "cleaner" system initially.
In other words, people prefer the distorted music to the undistorted music, to a certain degree (we are talking moderate measurable high order harmonic distortion; very high distortions don't sound good to anyone).
However, after about an hour of listening, the same people start to dislike the distorted version and prefer the undistorted version. If given the option, people will tend to want to turn off the music from the distorted system while tending to want to continue listening to the undistorted system after the hour and beyond, for many hours.
This is because our brains are constantly "doing the math" with regard to processing audio and are processing these high order harmonics that are not present in nature but are somehow present in this sound system we are listening to. Initially the brain enjoys the job of processing music, sorting out the sounds, and putting meaning to the lyrics. The higher order distortions are just a part of that, a novel part, and somewhat interesting. But after a while, the novelty wears off. And the brain gets tired of all that work, to put it simply.
I am not trying to advocate for anything, just offering some interesting things to think about with regard to High Fidelity and what people like and dislike when it comes to Harmonic Distortion and the spectra of exactly what harmonics dominate the spectra. And maybe, if you are building a car sound system, why it's important how long your commute is ;-)
* A note regarding frequencies and music. The highest fundamental tone of almost every musical instrument we normally listen to is almost always less than about 1200 Hz. A very few instruments can play a fundamental higher than that, but note the "very" in "very few". Not your guitar, not your piano, for example. So pretty much everything above 1 Khz is harmonics.
** Oh, and yes ... I am "really interested in Hi-Fi".
The amount of THD is an almost meaningless specification. The cynical might feel that is why manufacturers are so enamoured with publishing it ... but I digress.
What is important is the spectrum of distortion ... the 2nd, 3rd, 4th, 5th, etc harmonics. We tolerate rather large amounts of 2nd and 3rd Harmonic Distortion ... the 2nd, 3rd (and 4th) are common components of sound found in nature, and in musical instruments. The ear canal itself has a relatively significant amount of low order harmonic distortion. Thus our brains routinely dismiss this as an aberration.
Variations in the harmonics are what makes a violin and a piano sound different when playing the same fundamental ... A 440 (Hz), for example.
Distortions ... variations in the frequencies found in nature ... can be measured as a % but if these change the character of an instrument ... say, enough to make a Steinway sound more like a Yamaha Piano but not enough to make a Steinway sound like a Stradivarius Violin ... they are not going to be obviously noticeable to most people.
The higher order harmonics, and in particular the odd higher order harmonics, however, are a different story. Our brains are quite sensitive to these "extra" treats added by our sound systems, and we can detect them in percentages much lower than what we would find objectionable in a 2nd harmonic distortion.
Music is nothing if it's not mathematics and mathematical relationships. Many of us claim to be poor at math, but the truth is our brains are very, very good at math and don't require our conscious approval to do complex calculations in real time. However our brains use energy, and our bodies are designed to always minimize energy costs. In other words, exams are hard because they really are hard, they really do use more energy than "vegging out", they really do tax the human system.
So our brains are always looking for ways to minimize the energy cost, and this becomes important when we discuss music, audio and distortion. Which, let's not forget, is supposed to relax us, or at least not irritate us.
"THD" becomes a jumble of some important, and some moderately irrelevant distortion levels; we can't glean much about the sound quality of a component from that specification.
Which brings us to another aspect. High Order Harmonic distortions tend to add information that was not present in the original ... in other words Harmonic Distortion ... and of course these show up high in the audible spectrum. If we look at perhaps the 8Khz to 20 Khz area, we will find some of these higher order distortions; perhaps the 7th, or 9th harmonic of an instrument's fundamental note, for example.
They tend to add a bright tonal character to an audio system ... no surprise there; it's added "sound" in the higher registers. In auditory tests (which I won't cite as I don't personally need to revisit them, but they are available to someone with reasonable search skills to discover) ordinary listeners tend to prefer the sound of this higher order distortion to a "cleaner" system initially.
In other words, people prefer the distorted music to the undistorted music, to a certain degree (we are talking moderate measurable high order harmonic distortion; very high distortions don't sound good to anyone).
However, after about an hour of listening, the same people start to dislike the distorted version and prefer the undistorted version. If given the option, people will tend to want to turn off the music from the distorted system while tending to want to continue listening to the undistorted system after the hour and beyond, for many hours.
This is because our brains are constantly "doing the math" with regard to processing audio and are processing these high order harmonics that are not present in nature but are somehow present in this sound system we are listening to. Initially the brain enjoys the job of processing music, sorting out the sounds, and putting meaning to the lyrics. The higher order distortions are just a part of that, a novel part, and somewhat interesting. But after a while, the novelty wears off. And the brain gets tired of all that work, to put it simply.
I am not trying to advocate for anything, just offering some interesting things to think about with regard to High Fidelity and what people like and dislike when it comes to Harmonic Distortion and the spectra of exactly what harmonics dominate the spectra. And maybe, if you are building a car sound system, why it's important how long your commute is ;-)
* A note regarding frequencies and music. The highest fundamental tone of almost every musical instrument we normally listen to is almost always less than about 1200 Hz. A very few instruments can play a fundamental higher than that, but note the "very" in "very few". Not your guitar, not your piano, for example. So pretty much everything above 1 Khz is harmonics.
** Oh, and yes ... I am "really interested in Hi-Fi".
Last edited:
Would you agree we probably tend to think in terms of THD or THD+N simply because it is easy to measure?
Personally I don't think that is the root of it. There is one thing that has to be said about it ... back in the day there was so much ... well, it's found in baby diapers ... with regard to amplifier specifications in the market, that the US Government felt compelled to step in and define amplifier power and how it must be measured.
The result was "The FTC Rule" with regard to amplifier power, enacted in 1973 (I believe; around there anyway). That specified how you could advertise amplifier power ... it did not, as some believe, legally define amplifier power specifications. What it did define is what you could advertise as amplifier power, and mandated that if you did mention it in an ad, it had to be a certain measurement and that the font used in the ad must be the largest font when specifying power according to this method, and that the measurement must be derived after following a specific measuring technique.
So, we had a stereo amplifier that could advertise power as:
"30 watts RMS per channel, both channels driven, from 20 to 20,000 Hz, into an 8 ohm impedance, with no more than 1% Total Harmonic Distortion".
All of the above were required; not one element is optional.
Now, you could simply not mention power output in your magazine ad, and still be compliant. But if you did mention power, it had to reflect the measurement technique (which also specified a pre-conditioning at ⅓ power for 1 hour, and at a specific AC voltage, eg 117 V AC, and so on) and the result in the form above.
If the amplifier mis-behaved or if the protection circuit was tripped, you could not use those conditions in a specification. So, for example, if the amp would overheat into a 4 ohm load during the pre-conditioning, you could not advertise ANY power output into 4 ohms. And so on.
Today we are back in the pre-1973 days, because the FTC has bowed to industry pressure, they reduced the pre-conditioning requirement to 1/6th power for 1 hour, and we have manufacturers specifying power at 1 Khz, using "peak power" instead of RMS, and so on.
They have also argued that the "FTC Rule" does not apply to Home Theatre receivers because they have more than two channels, and the FTC has not challenged this. [Some manufacturers, such as Anthem, still measure HT power as if the FTC rule applied, so their multichannel amps specify power with all channels driven, for example]. Generally, though, HT amplifier power is not comparable to a stereo amplifier's power specification (because the FTC rule still applies to the stereo amp).
It also never applied to "portable" equipment, which is why you can buy a blaster for $200 with 300 watts of power. Not "real" watts, but, hey. Not breaking the law, so there.
For a long time, manufacturers simply put a handle on gear that had no business being carried by anyone, called it portable, and lied through their teeth. TV manufacturers put a handle on a 150 pound TV that was 3 feet across, and made ridiculous claims on power output. But I digress.
Back to your question, I think that because the FTC rule "ruled" the specification sheet during the entire stereo years of '73 to perhaps 2000, the THD specification is an established part of amplifier specifications. So it has inertia.
The second is that the quality and resolution of audio test gear has improved significantly over the years since the FTC rule became part of the vernacular. Even a 1970's era $10,000 Distortion Analyzer, which even some manufacturers could not afford *, were weak in that regard. So being able to measure the higher order harmonics is a relatively recent ability, and that has to be part of it.
But, now that we can measure it, there is certainly no reason to ignore it, and there is merit in trying, as always, to correlate what we hear with what we can measure, to see if there is a relationship we can comfortably work with.
The third reason is that sales staff at audio resellers, and this includes the sales information at an online reseller, want simple concepts they can present to consumers. THD has become that for audio gear. This isn't different than any other consumer commodity ... if you buy a camera they talk megapixels (when other aspects of the sensor determine picture quality) ... if you buy a car they talk horsepower, when how we feel power applied while driving is all about torque, but with a small motor you can just spin it faster and get a bigger HP number ... and so on.
* Today $10,000 sounds like the kind of money any reasonable enterprise could shell out for. But in 1973, it bought two or three new cars, or perhaps 4,000 hours (two years worth) of an Electrical Engineer's time.
Last edited:
Please let me raise some other related issues: If a chord is played on a musical instrument, recorded into a computer, and the spectrum analyzed, either there will be beat frequencies present or not. If so, some nonlinear mixing of the notes struck on the instrument would appear to have occurred. Wouldn't that amount to some sort of IMD in the instrument itself, and if so would that mean that we like to hear a little IMD too? Say in order to fully appreciate chord textures?
Many good and interesting points. At the same time though, if one looks to buy a distortion analyzer, many used units from a few years ago only have a digital readout to display THD+N. Whether or not that was because of skewed market demand traceable to government regulation, when that's all the instrument one has on hand can display, people don't usually focus attention on what is not there.
Even our most complex Intermodulation Distortion measurements are orders of magnitude simpler than reproducing a simple melody played by a soloist, let alone a band or orchestra.
Broadly speaking we see IMD values that are higher than THD values, so we know the typical audio device has issues with complex musical information compared to simple sine waves.
So we cannot say we are able to measure "the full measure" of distortion an amplifier (for example) generates.
I am not one to dismiss measurements as somehow inferior to listening evaluation. But I find it strange when people advocate measurements as the whole picture of performance.
We ask audio systems to perform a task and we build and design them to perform that task, yet some feel we should not use our ears as a measure of performance?
It is not necessary that everyone "hears the same" in order to evaluate audio by ear. It is not even necessary that audio systems perform to some arbitrary standard, such as being able to reproduce a live event sonically. I have no argument with someone who prefers a different sonic presentation than I do; his system was not built to please me.
Some people can discern very small differences in sonic presentation; most people cannot. The current literature seems to suggest it's a ratio of about 1:4, or 20%/80%. When it comes to similarities, there are even fewer people who can reliably and consistently identify them; perhaps as low a 1:9, or 10% to 90%. There is nothing wrong with being amongst the 80 or 90%. Buy what pleases you and enjoy.
Above all, enjoy.
Broadly speaking we see IMD values that are higher than THD values, so we know the typical audio device has issues with complex musical information compared to simple sine waves.
So we cannot say we are able to measure "the full measure" of distortion an amplifier (for example) generates.
I am not one to dismiss measurements as somehow inferior to listening evaluation. But I find it strange when people advocate measurements as the whole picture of performance.
We ask audio systems to perform a task and we build and design them to perform that task, yet some feel we should not use our ears as a measure of performance?
It is not necessary that everyone "hears the same" in order to evaluate audio by ear. It is not even necessary that audio systems perform to some arbitrary standard, such as being able to reproduce a live event sonically. I have no argument with someone who prefers a different sonic presentation than I do; his system was not built to please me.
Some people can discern very small differences in sonic presentation; most people cannot. The current literature seems to suggest it's a ratio of about 1:4, or 20%/80%. When it comes to similarities, there are even fewer people who can reliably and consistently identify them; perhaps as low a 1:9, or 10% to 90%. There is nothing wrong with being amongst the 80 or 90%. Buy what pleases you and enjoy.
Above all, enjoy.
Well, you can buy an Audio Precision Distortion Analyzer, the finest instrument of it's type you could find at any price in 1980, for under $1000 today.
It's interesting that you mention THD+N. It's something that has bothered me, or put another way, has me thinking about, for a while now.
We spend most of our time listening to amplifiers (for example) at 1 watt or less ... often much less ... output. Yet the measurements we are able to make at this part of the performance envelope are rudimentary at best ... the conventional wisdom is this is buried in noise, and we can't hear it because of that.
But can we? And if there were no noise, what would the distortion spectrum look like? Would it also rise like the THD+N curve does? Is an amp "cleaner" at 1 watt than at 1/10th watt?
Does it play a role in what some perceive as differing sound qualities? Or is it really as they say ... just noise?
-Regards
Starting EE in '67 was already $7500 in US and actually 4 AMC gremlins for $10,000.
Last edited:
Well, maybe. So only one EE, then, who did a bit of overtime.
Current starting salary for a green just-out-of-college EE here today is $50~101,000, apparently. Relative to the cost of a good piece of test equipment, that's still real money in anyone's book.
Last edited:
I didn't mean to offend you. I quoted my first boss' pay at Transitron because he shared it with me I was still in highschool in '67.
Sound is an analog phenomena, our ears work in the analog domain and no matter how you store your information describing a musical performance, it all has to be presented in the analog domain sooner or later.
As far as the current topic is concerned, I don't think it's relevant whether your storage media and the retrieval of that stored information is digital or analog.
It all has to be encoded somehow (whether that's by FM or AM or magnetic tape or PCM or cut by an lathe) and they all have their idiosyncrasies and their merits, but in the end they also all have to be decoded to analog, before we can listen to it.
Fear not, it takes a lot more than an anecdote to offend me. Your input is welcome. I was in primary school in '67, but I had begun my (first) audio career in January 77. And if you really want a laugh, get this ... I was hired to build cables. Should have stayed in that field. ;-)
If anything is "way below" audible levels, it's spectrum, it's colour, it's way with women even, is meaningless. Difficult to disagree there.
However it's well established that the spectrum is important with regard to what is the "audible levels (sic)". If by "inaudible" we mean that an ordinary listener in a Double Blind Test cannot say which they prefer when presented with, for example, a set of sound samples with 3% THD and 1% THD when both samples consist of only 2ndHD, then 3% 2nd Harmonic Distortion is inaudible to many people; if the 3% THD consists only of 5thHD, then that same 3% Harmonic Distortion is offensive to most people.
Last edited:
In part, I think it is a question of "upbringing". If you have gone through a system of learning to use and trust measurements, then you will do just that.
If you haven't you won't. Generalising, obviously.
Your post reminded me of this;
"When my hearing was last tested there was a noise floor calibration test done as part of it. But I asked the tech to continue with quieter and quieter signals, until they were much quieter than the noise floor. I could still pick out signals @-10db below the noise floor. Noise floors are complex and audio tones are not, so detecting the tones was easy for me. The tech herself tried to hear these tones that far below the noise floor and wasn't able to do it. I applaud Whitneyville because it sounds like he has preserved his hearing as well as possible (considering he is a shooting enthusiast). But since I have not ever heard that it is physically impossible for anyone in the world to ever use their ears to detect frequencies above 35KHz, I obviously need to go back to school on that subject."
From a long dialogue at Vinyl Engine posted by desktop, who used to design speaker systems for Disney. https://www.vinylengine.com/turntable_forum/viewtopic.php?t=39071
I expect my electronics to be a few orders of magnitude below your examples. Speakers however are a different story, they all have audible levels of distortion and then the distortion spectrum carries some interest.
The folks at Benchmark say they built a power amp to test their DACs with, since they could not buy an amp up to the task. The specs are here: https://benchmarkmedia.com/products/benchmark-ahb2-power-amplifier
Some, but not many here have the test equipment to measure amplifier performance to that level. Of course, you may be able to, but where does that leave the DIY enthusiast who wants to work toward more accurate reproduction?
- Status
- Not open for further replies.