H2 is non linear distortion, because it is caused by a nonlinearity in the transfer curve.
Sometimes people mention linear distortion which is a non-flat frequency and/or phase response for instance.
That can happen even with an amplifier or speaker with no nonlinearity in the transfer, hence the name linear distortion.
Harmonic distortion and intermodulation distortion are non-linear distortions, both caused by the same nonlinearity, just measured differently.
See for instance: https://www.nti-audio.com/en/support/know-how/lets-clear-up-some-things-about-distortion#:~:text=A linear distortion is defined,signal non-linear distorted signal :
"Linear vs. non-linear distortion
A linear distortion is defined as a change in amplitude or phase with no new frequencies added.
A non-linear distortion occurs when new frequency components are generated. Non-linear distortions are what is usually meant by "distortion"."
So if you see a 2nd harmonic in an output you immediately know that component has nonlinear distortion.
Edit: I see you changed that. Hope you don't mind the little essay, others might like it ..
Jan
AFAIK, Geddes never made an actual subjects study, not to talk about double blind tests. I don't blame him, it's extremely time consuming and it's expected to cause tons of backlash. It's a war you cannot win.
Yes, that will again fill 1k+ replies to a thread which will never ever be read again in full and never comes to an end or conclustion.
Well, in some HiFi magazines you can get a kind of rule of thumb ("K2 not so bad, K3 kinda sterile/analytic, K5 run for your life!") - which you have to take with a grain of salt. Or even a 1kg packet of salt. Yes, they kinda experienced it like that but there's a blurry line between experience and audiophool mumbo-jumbo. I do not mean to take their word for granted but try to take their description of the effect of the distortion and correllation to the sound as to WHERE to listen to (to recognize it) and don't follow strictly their 'cause' argumentation. Again, please don't rely on their causation theories but take it as a hint on what to listen to to get your own judgement.
THANK YOU! That's what I wanted to say in my former reply (which I butchered severely while re-re-re-writing my former post)
That's why I said; tried to make sense of it.
Meaning, making a start of this undefined and uncorrelated mess.
But when there is a lack of follow-up on that, it kinda falls flat, I agree.
However, that doesn't mean that (therefor) the entire concept or idea is untrue.
In fact, if you follow the literature and have knowledge of psycho-acoustics, it holds more true(ish) elements instead of untrue elements.
But I agree, to properly finish it scientifically off, practical experiments are a factual must.
Otherwise it's just a hypothesis and also nothing more.
I have tested at what level I can detect distortion harmonics up to 10th order.
The most important is 2nd and 3rd I I can detect those at 2nd: 0,63% and 3rd at 0,25%.
This tested with pure sine wave tones.
Listening to music is not the same. If you read Nelson Pass paper about distortion and understand that many separate sinus tones, like music is, you can get much much higher amounts of distortion. An amplifier with low distortion for a single sine wave is a completely different story with music.
To me, the character and individual levels of harmonics is much more important that the total.
Could write a book on the subject but just want to state that low distortion levels is not that important, the character of distortion is important for how distortion mix with the music.
The subject is very very complex, but just hunting the lowest possible distortion is a dead end.
That 0,00001% amplifier can have 10% of distortion at the input stage created by massive amount of negative feedback.
Sounds good?
NO, not at all.
The most important is 2nd and 3rd I I can detect those at 2nd: 0,63% and 3rd at 0,25%.
This tested with pure sine wave tones.
Listening to music is not the same. If you read Nelson Pass paper about distortion and understand that many separate sinus tones, like music is, you can get much much higher amounts of distortion. An amplifier with low distortion for a single sine wave is a completely different story with music.
To me, the character and individual levels of harmonics is much more important that the total.
Could write a book on the subject but just want to state that low distortion levels is not that important, the character of distortion is important for how distortion mix with the music.
The subject is very very complex, but just hunting the lowest possible distortion is a dead end.
That 0,00001% amplifier can have 10% of distortion at the input stage created by massive amount of negative feedback.
Sounds good?
NO, not at all.
Yes, I've already corrected myself and I'm sorry my post ended up in such a brain fart. I know a linear distortion does not result in additional frequency results, which any HD does. Linear distortion is often not as 'bad' as unlinear distortion, especially at low frequencies because there the ear is very insensitive to it - otherwise nobody would be able to bear with the phase shifts of any vented speaker.
Yes but as I said, you can't exactly interchange them.
No, I don't mind, just the opposite! You've posted a perfectly and easily understanable page which clears the issue in the easiest-to-consume-and-understand-as-possible way. Thank you for that. I've known that but made a mistake at my reply. I am sure there will also be others who read my former reply first and probably take that for a fact (because I can't edit it anymore), so your post will definitely help to clear up any confusion I've possibly caused (well, I at least hope it will clear that up).
Quoting that like newspapers do: That sounds a bit like distortion - kinda dirty! 😳 😉
No, really, there are so many variables and influences, many not even remotely measurable (psycho-acoustics, listener descriptions) and no rewards in sight but a lot of pushback, I don't blame anyone to not go down that rabbit hole.
Because they are not always from the same source, see my post before.
Saying that they are the same thing is just over-simplifying things.
Sure, you can create IMD from different HD components, but that doesn't tell us anything.
Particularly when there are elements that ONLY create HD distortion, while there are others that create IMD as well.
Le(x) and doppler non-linearities ONLY creates IMD and no HD distortion.
Don't know, but for the last 30 years at least, everything is measurable at least here on planet earth with the current state of science and engineering.
We are not talking about some extremely difficult science and physics here.
That includes stories people want or don't want to believe.
I personally think that the lack of going into the rabbit hole has more to do with that last part.
On top of that is the obvious commercial part.
Thank you very much for posting that table! It clearly shows that there is HD without IMD, again, thank you for confirming that!
It was already posted as link before 😉
Btw, it's not about "confirming" but knowing the literature as well as the relationships between all variables (nofi).
Compliance is a good example of that last part, because there is no relationship between the difference in pressure vs the compliance "interacting" with any other variable.
Meaning you can and will never have any IMD products from that part.
Or another way of saying this.
If we assume a black-box, measure its performance, we know for sure that the IMD we are measuring is not coming from the non-lineairity of the compliance.
Btw, it's not about "confirming" but knowing the literature as well as the relationships between all variables (nofi).
Compliance is a good example of that last part, because there is no relationship between the difference in pressure vs the compliance "interacting" with any other variable.
Meaning you can and will never have any IMD products from that part.
Or another way of saying this.
If we assume a black-box, measure its performance, we know for sure that the IMD we are measuring is not coming from the non-lineairity of the compliance.
Last edited:
Klippel (so far) concentrates on bass and cone drivers. Since it doesn't include any CD and horns (which I specifically mentioned), I'd really like to see a paper or article which explains where I would be wrong about that.
Where do you get that information from?
These basic elements count for any transducer.
Since they are made up from the same building blocks.
The only thing that is different from a tweeter vs compression driver, is the compression chamber.
Beside, this looks a lot like a compression driver to me;
I would highly recommend reading a bit more, sorry.
The topic was about "circuit" i.e. electronics as I understood it. If the intimate relation is between HD and IMD is the same for mechanics I have to admit, I'm not sure of but my guess would be that it is.
If there is a non linearity there is HD and if in this circuit, you inject 2 tones, you will have IMD. Thats indisputable.
//
No, that's not correct. An unlinear Bl(x) is one of the main causes for IMD, While almost all cone drivers are overhang VC constructions and therefore got a severe Bl(x) variance, every CD I know is an underhung motor driver, which does not result in an unlinear Bl (well, unless it's used way below the horn loading and excursion limit). That differs tweeters as CDs to the Klippel researched cone drivers. So maybe it's recommend to read up about driver motors and their magnet field principles?
Of course you did know that, reading the Klippel papers and not just taking a cone tweeter just as equal to a bass driver? Don't mind that, we're here in the forum to exchange and share informations, knowledge and resources. Sometimes claims in the forum seem to be lack of knowledge but sometimes you don't have all the informations and if all you've got is a hammer, everything looks like a nail. It doesn't matter who's right or who's wrong, as long as the forum gets a conclusion on a correct answer.