This thread has been very informative about the nature of audio drivers. I believe we are working with the wrong means to reproduce music if what some here are saying is true.
Please watch this video and you may see that working with the current state of drivers can never truly achieve the goal of accuracy some here strive for. For, when you put mass in motion it takes on a like of its own, let alone trying to get it to take on the life of another material in motion that is foreign to itself. I believe that the perfect transducer will have to be mass-less, yet able at the same time to move air, and only the air you want moved.
This is what happens to when mass is put into motion by a musical instrument, specifically the drums.
Mike Mangini on Discovery's Time Warp - YouTube
If the instrument distorts to achieve its sound? How can what is trying to reproduce it have to not be able to distort?
Please watch this video and you may see that working with the current state of drivers can never truly achieve the goal of accuracy some here strive for. For, when you put mass in motion it takes on a like of its own, let alone trying to get it to take on the life of another material in motion that is foreign to itself. I believe that the perfect transducer will have to be mass-less, yet able at the same time to move air, and only the air you want moved.
This is what happens to when mass is put into motion by a musical instrument, specifically the drums.
Mike Mangini on Discovery's Time Warp - YouTube
If the instrument distorts to achieve its sound? How can what is trying to reproduce it have to not be able to distort?
Maybe he is, but it begs the question of just how "bad" distortion is in the larger context. What if the instrument distorts, then the speaker distortion was able to reverse that distortion? It's a rhetorical question, so don't bother to answer. Just gets me thinking. And, of course, if the drum does distort, yet that's what we hear and enjoy in our live music, who cares?
The practical example for me is SET amps. They tend to have high distortion compared to other amps, yet they tend to sound (to many) beautiful. Perhaps that's because, as with violins, we tend to like certain kinds of harmonics even if they are a distortion of the original signal. Maybe faithful reproduction of the source would sound dry and awful. Oversimplifications, of course, but it's a rhetorical thought exercise for me anyway.
Although I don't think it's the whole story, I think there is some truth to what you suggest, having fallen prey to this effect myself. The ear is extremely sensitive (and fussy) in the presence region - which for the sake of argument I'll label 2Khz to about 5Khz, possibly a bit wider than some peoples definition.
The response balance between the presence region and the the rest of the spectrum (particularly mid and lower mid, but also the high end treble) seems to be one of the major determining factors in whether a speaker has a "you are there" hyper realism (elevated presence region) or the typical lifeless "hi-fi" sound. (a depressed presence region, like the BBC dip or similar)
The former is often a hallmark signature of the raw response of a full range driver (with at least one major resonance falling between 2-5Khz, sometimes two) and the latter being quite common on multi-way designs where the design puts the crossover frequency in the presence region, but fails to accurately integrate the two drivers. (Or chooses a poor driver combination)
Sometimes in multi-way designs I think a dip is actually put in the response on purpose to make the speaker more "palatable" to the general speaker buying public, especially on poor recordings or to attempt to hide uncontrolled cone breakup of a poor driver in that critical region.
The result is that un-offensive but lifeless sound that we've all heard on consumer level commercial designs. (In extreme cases, "boom and tizz") A speaker of this type never has that "you are there" realism that some of us look for, but also generally doesn't annoy.
On the other hand a full range driver often has an elevated presence region and/or discrete resonances in the presence region that can add spikes as much as +6dB or more, as well as the attendant time domain ringing.
On some recordings this can certainly sound exciting or even exhilarating, and what I would refer to as "hyper realism", (and which many people might refer to as "detailed") however I find it quickly becomes fatiguing and whilst it can make some recordings sound more visceral and "real", on many other recordings it can make the recording un-listenable to the point of being painful or obnoxious.
Basically I think the ear/brain is easily "excited" by the presence region, (in much the same way your taste buds are by sugar and fat) but like many things, too much of a good thing is a bad thing.
A shelved elevation in the presence region increases the overall stimulation within that frequency band, but a resonance there also increases the time duration of stimulation so that the frequencies around the resonance are constantly being overstimulated, and I think this is where the presence related listener fatigue comes from.
(I suspect too much continuous stimulation at the same frequency at high frequencies can cause irritation, and a high Q resonance will tend to stretch out the content of the recording in that frequency region in time long enough that it's more or less continuously present, whereas in the original recording there would have been gaps in time)
While playing with full range drivers that have a peak in their response at 2Khz and 4Khz, the effect is very noticeable. After adding both mechanical damping and EQ to correct the response so that the peaks in the presence region are gone, and that it's flat to better than +/-1 dB from 2Khz to 6Khz, the result is that although it's now far more listenable, neutral, and not fatiguing, a lot of that sense of "you are there" is lost.
However now lift the entire presence region from 2-5Khz flat by only 0.5dB or so and it's possible to have the "you are there" realism, but without the fatiguing characteristics of the resonances. To my ear this new flat but (possibly) slightly elevated presence region sounds cleaner and more "detailed" than the former peaky presence response, because there is no ringing from resonances.
It seems that a very small increase in the level in the presence region over a broad frequency range can compensate for the "lost" sense of realism that can sometimes occur after eliminating resonances in that region that were giving a false sense of "detail".
Anybody modifying drivers to reduce or eliminate resonances should always keep in mind that a slight modification to the overall broadband frequency response in that frequency region may be necessary to "compensate" for the lost response of the eliminated resonances. (Like a small broad increase in response over a couple of octaves to compensate for eliminated narrow high Q resonances)
Another example would be bracing cabinets to reduce resonances in the lower midrange - at first sight it may appear that bracing the cabinet causes the overall response (which sounded balanced initially) to sound a bit thin and lacking in "warmth", from which one might conclude that the bracing was the wrong thing to do. In matter of fact all you're doing is reducing or eliminating the peak in response due to a resonance, and a small, much broader band increase in drive level around that frequency range to "compensate" and restore overall balance may be needed.
Just my 2c
Haven't there been articles on the non-linear damping characteristics of some cone materials, polypropylene in particular ? I think it's well known that polypropylene exhibits non-linear damping with signal amplitude due to the material's plasticity. Something which I don't think is an issue with paper.
Also, on any cone that is operating in cone breakup - such as a full range driver, the cone is bending back and forth with transverse bending waves. If the cone is more easily able to bend in one direction than the other, you're at least going to get some increase in 2nd harmonic distortion over a true piston.
The linearity of a material bending would seem to be an important factor to its distortion at high frequencies above cone breakup, although I haven't seen anything to suggest that it's a big issue, at least with paper.
That would be the ideal of course. I don't know how closely actual drivers approach this, and how closely different materials approach it though.
Last edited:
So because a SET amp has more of certain types of distortion than a solid state amplifier, yet they sound "beautiful", (to some, apparently) it must be because higher distortion sounds better, right ?The practical example for me is SET amps. They tend to have high distortion compared to other amps, yet they tend to sound (to many) beautiful. Perhaps that's because, as with violins, we tend to like certain kinds of harmonics even if they are a distortion of the original signal. Maybe faithful reproduction of the source would sound dry and awful. Oversimplifications, of course, but it's a rhetorical thought exercise for me anyway.
Correlation does not equal causation. There are a whole host of other characteristics that are different between the two amplifiers, other than 2nd harmonic distortion.
Also I'd like to strongly disagree that adding non-linear distortion to violins makes them sound anything other than worse. In my experience, any significant amounts of intermod distortion, whether by speakers or amplifiers, can quickly turn violins and other spectrally dense material like orchestras into a horrible congested mess.
Forget about simplistic discussions of harmonic distortion, (like the often claimed "a little bit of 2nd harmonic distortion can actually sound good") any time a speaker/amplifier with harmonic distortion is reproducing more than one frequency at once (and certainly more than one instrument at once) you are also producing intermod distortion, which is nearly always un-musical in nature since it's sum and difference frequencies.
Maybe a SET amplifier can sound good in some circumstances, but it won't be because it's distortion is higher it will be because it's slightly better in some other metric. (Some research is now starting to suggest that maybe it's the higher order harmonics in a high feedback design which are largely lacking in a SET amplifier that could be one of the causes...)
I am speaking in reference to how a speaker contorts as a mirror image of what is being contorted. I am not saying I am right. But, I am saying that there may be something in the future discovered that today was seen as wrong. It won't be the first time that has happened in any science.
Speakers are supposed to re-create what's on the recording and one usually hopes that it approximates the original sound. That's the whole purpose.
It does not do that.
Dave
Last edited:
Gee...I don't remember saying that. I did suggest that MAYBE some kinds of distortion are pleasing and may be the reason, or part of the reason, some amps sound subjectively better than others. But in general higher distortion sounds better? Please don't put words in my mouth.
Of course. Remember, it was a thought exercise, just playing with concepts (hence the spiral smilie!). But I am one who DOES believe, at least for now, that the 2nd order distortion is a big factor in the appeal of SETs. Not the only factor, just a big one. In short, I'm happy to challenge the notion that all distortion is bad. Since I've never heard source material without distortion (and never will) it remains a thought exercise.
Maybe, maybe not. Can you point me to some sound research that quantifies that it's NOT the distortion that's appealing, or that it IS something else (since proving a negative is always hard)? I'm still attracted to the idea of odd vs even order harmonic distortion as a factor, but I'm always open to suggestion.
Last edited:
Don't feel sad.
Yup... This is interesting.. Bowers & Wilkins - Kevlar
Yet you (and Tade) are still here?
I, for one, have appreciated your input in the past, John - it contained some of the only really well presented counter-arguments - but griping about the thread seems like a waste of everyone's time. I'd prefer your engaged involvement.
I can't speak for him, but I'd say he follows it partly because misrepresentations of his past involvement with testing resurfaces with what I have to say is intentional.
Dave
Hey, folks!
Again? Watch this! I believe they achieve a great deal of what EnABL seeks to do.
Bowers & Wilkins - Kevlar
Again? Watch this! I believe they achieve a great deal of what EnABL seeks to do.
Bowers & Wilkins - Kevlar
Hi GeneZ,
surely i had a reason, why i posted that B&W link, but please
bear in mind this beeing an - though instructional - advertizing spot.
Woven fibre is one among many possible solutions to achieve something
like "controlled breakup".
And there are still manufacturers using metal (and metal composites) e.g.
with good success.
Each (valid) strategy chosen will - in the end - be as good as its implementation.
___
Btw. What is it EnABL seeks to do ?
As Pano stated basically, it served - at least - as an introduction to discussing
vibrational behaviour of cones. Any kind of introduction may be valuable, but it
might not be the end or the result of the discussion.
Above the individual frequency range of pistonic motion every conventional speaker
acts as some kind of bending wave transducer, that's how i interpret things.
Regards
surely i had a reason, why i posted that B&W link, but please
bear in mind this beeing an - though instructional - advertizing spot.
Woven fibre is one among many possible solutions to achieve something
like "controlled breakup".
And there are still manufacturers using metal (and metal composites) e.g.
with good success.
Each (valid) strategy chosen will - in the end - be as good as its implementation.
___
Btw. What is it EnABL seeks to do ?
As Pano stated basically, it served - at least - as an introduction to discussing
vibrational behaviour of cones. Any kind of introduction may be valuable, but it
might not be the end or the result of the discussion.
Above the individual frequency range of pistonic motion every conventional speaker
acts as some kind of bending wave transducer, that's how i interpret things.
Regards
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.