• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

What's Left In Speaker Design To Reduce Distortion/Increase Detail Retrieval?

Adi777

Addicted to Fun and Learning
Joined
Apr 14, 2022
Messages
690
Likes
461
Yes, you're right, yes, it's ASR - S for science, but I think it would also be nice to respect slightly less objective entries, listening reports, and so on. This is my opinion. Writing in every post, literally, about controlled blind testing is a bit weird - for me.
 

tuga

Major Contributor
Joined
Feb 5, 2020
Messages
3,984
Likes
4,285
Location
Oxford, England
No, every science minded person knows that if they disagree with something Toole or any other scientist has written, then a prerequisite is valid contrary scientific evidence, specifically in the form of a valid, controlled listen test, since that is what Toole writes primarily about. The conspicuous absence of a shred of any data from those who throw pot shots at Toole et al, firmly establishes their "science mindedness", usually without cognizance.

Have you any idea of how often the expression "seems to" or "seems like" pops up in Toole's work?
Your Black & White views and Absolute Certainties sound like dogma, not Science.
 

Purité Audio

Master Contributor
Industry Insider
Barrowmaster
Forum Donor
Joined
Feb 29, 2016
Messages
9,184
Likes
12,477
Location
London
Yes, a higher DI combined with low distortion and constant directity (if it's well designed) are attributes for more details.

Personally I think the dislike for horn has more to do with poor horn designs. Many, if not most, sound highly colored and not coherent.
That would be my experience one or the other often both.
Keith
 

FrankW

Senior Member
Joined
Mar 7, 2023
Messages
393
Likes
373
Have you any idea of how often the expression "seems to" or "seems like" pops up in Toole's work?
"Toole's work" is less that 1% of cited studies/papers in his writings. Of course one would have to read them to know this.
Your Black & White views and Absolute Certainties sound like dogma, not Science.
Ditto for your hand waving in lieu of a shred of contrary (or any) science evidence.
A sore spot, I know :)
 

fineMen

Major Contributor
Joined
Oct 31, 2021
Messages
1,504
Likes
680
Have you any idea of how often the expression "seems to" or "seems like" pops up in Toole's work?
Your Black & White views and Absolute Certainties sound like dogma, not Science.
The 'science' becomes dogma once decissions are based on it. Instead of keeping an attitude of uncertainty decision makers hate to be questioned.

We--the other 'we', invest tens of kilo-unities of our respective currency to enjoy music from a conserve. I question that, not being one of the we, the anti-social guy in this respect. The 'we' states a dogma of only accepting 'the best'. But the criteria are missing.

Toole serves with criteria, but not with an explanation why some anti-social guys on the internet see themselves entitled to ask for any 'best'. And only to remain in a 'critical listening' attitude as to assure there is not even fun in it.

Bad English again, guys? Or do some misuse science as an excuse for mindless exaggerations? Not the least because of the dismal level of technical understanding in the community, the question might be in order.
 
OP
MattHooper

MattHooper

Master Contributor
Forum Donor
Joined
Jan 27, 2019
Messages
7,332
Likes
12,292
(Ah well, since he's brought this strange act back to the thread for a moment...)

Wrong. Your title is a cause>effect. Begging the question fallacy. You're not asking IF "reduced distortion" (which you can't even quantify what type, IMD, THD, frequency, spaghetti) will "increase Detail Retrieval" (whatever that means), you are stating it does, then asking what aspect of speaker design accounts for this.

Now you've added the incorrect application of "Begging the question" to your bag of tricks.

The specific content of the question is posed in the body of my OP. In asking "what's left" to do in speaker design to lower audible distortion in the reproduction of the audio signal, a possible answer is "nothing." Which is why I asked in my OP "Are we done? Or is there more to achieve....?" And I have re-iterated the open nature of this question several times.

There is sonic information encoded in recordings. This includes everything from the most conspicuous sonic characteristics of melody/song structure, voices, instruments, musical performance, down to the timbre of the chosen instruments, even the difference in timbre between various drum cymbals, down even to the subtlest bits of reverb or acoustics added or captured in recordings. The question then is are the best speakers capable at this point of reproducing all the sonic information in any recording, accurately, or is there still any ways to go? And IF there is still enough happening in the best speakers to distort the recorded information, or there may be yet more detail left unresolved, where would speakers need to improve. It's a totally open question on all counts.

You thought you'd sniffed out a naive subjectivist but that turned out to be wrong as I've shown, yet you can't admit all the incorrect disparaging assumptions you made. Being driven apparently by the need to knock someone down a peg rather than fruitfully engage with the question in the thread, you continue emptying your quiver on the strawman that I take my sighted impressions as fact, that I'm not aware of the possibilities of sighted bias or alternative explanations for my anecdote, ignoring that I acknowledge the liabilities of sighted listening and have continually said those anecdotes can be ignored for the question at hand.

As for "begging the question,"...hardly.

Of course I have assumed for the sake of my topic question that there is sonic information contained in recordings, and that speakers are capable of distorting that content.
This assumption is not a "question" that is begged: it's a proposition already understood as true, obvious and well established, by any informed person here. And therefore the basis on which to launch discussions like the one I've raised.

Take for example a discussion thread in a cartography forum about how to produce an accurate global map of the earth. This assumes the earth is a globe. It doesn't "beg the question" because nobody in such a forum is questioning the well established fact of the earth's spherical nature. It's only "begging the question" to Flat Earth nutters. But of course one doesn't expect to be addressing Flat Earthers in a cartography forum, and a thread discussion about accurate globe mapping is not for them. It's for people who already know the earth is a sphere.

Likewise, nobody here who knows anything about the nature of speakers, much less sound reproduction in general, questions whether there is sonic information in recordings that can be distorted to a greater or lesser degree by loudspeakers. And that much of the business of designing speakers (especially accurate speakers) involves seeking to reproduce recorded signals in ways that reduce the distortions that can be introduced in various ways by loudspeakers. And on this basis it's reasonable to discuss the current state of speaker design, with respect to accurate reproduction of recorded detail.

But here you come, suggesting that it's "begging the question" and that, first of all, that one can't just assume speakers can be distorting recordings - that this first must first be demonstrated to you. Like the flat earther saying to cartographers "Wait, you've just ASSUMED the earth is a sphere here - you need to first demonstrate it to me!" If you were really that ignorant on the fundamental assumption for this conversation, this thread isn't for you.

But you, being an AES member - for reasons best known to yourself - are just playing ignorant, right? It's why you refused to answer my questions about the rational in designing speakers here:




Because as soon as you actually directly address those questions - the gig would be up. All your handwaving about me relying on un-demonstrated assumptions would be answered by...yourself.

Now...go ahead if you want. You can accept the general principle that speakers can distort the recorded signal, and then go on to fruitful engage the question of whether the best current speakers are sufficient to accurately render all the recorded sonic information. If so, why? If not, what areas are there left to improve upon.

OR...you can continue instead your project of trying to one-up people in the thread, pretend the question might be moot because the very principle that speakers can distort recorded detail is dubious or un-demonstrated, and keep railing at strawmen.

Ball is in your court.
 
Last edited:

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,469
Likes
2,466
Location
Sweden
The 'science' becomes dogma once decissions are based on it. Instead of keeping an attitude of uncertainty decision makers hate to be questioned.

We--the other 'we', invest tens of kilo-unities of our respective currency to enjoy music from a conserve. I question that, not being one of the we, the anti-social guy in this respect. The 'we' states a dogma of only accepting 'the best'. But the criteria are missing.

Toole serves with criteria, but not with an explanation why some anti-social guys on the internet see themselves entitled to ask for any 'best'. And only to remain in a 'critical listening' attitude as to assure there is not even fun in it.

Bad English again, guys? Or do some misuse science as an excuse for mindless exaggerations? Not the least because of the dismal level of technical understanding in the community, the question might be in order.
I can only guess what :rolleyes: means, really.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,771
Likes
37,635
These distortion graphs for the Revel F328Be look pretty good to me.


index.php





Revel F208 isn't too bad above 100 hz.
index.php






KEF R3 which is not terribly expensive looks pretty good above 200 hz.


index.php






Then a counter-example. The JBL LSR308 which has good directivity and response, but not looking so great on distortion.


index.php




So larger speakers would seem to have distortion under control other than the bass. So less bass distortion, and I'd think what is left is further resonance reduction. These larger speakers show less of that as well in the better designs.

Other than that I'm pretty sure you'll find speakers that seem to retrieve detail better do so by some minor FR emphasis of this or that region. Or a result of some narrowed directivity in certain ranges.
 

Waxx

Major Contributor
Joined
Dec 12, 2021
Messages
1,984
Likes
7,884
Location
Wodecq, Hainaut, Belgium
I don't think that distortion is low enough actually. If we demand that amps are low distortion, speakers also should meet the same standards. Revel has low distortion to today's standards, but nog low enough to be perfect.

And on Toole, i question him all the time, I intuitive also have some things i don't believe, but no scientific proof (yet), so i am not right entitled to claim something against his science. He is right untill scientificly proven wrong!!!
 

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,903
Likes
16,917
"If the only tool you have is a hammer, you tend to see every problem as a nail"

In the end humans tend to observe and prioritise most what they are currently working at, independently if they are researchers, engineers or hobbyists.

Loudspeakers and their understanding and engineering have gone a slow but continuous progress in the last 6 decades, personally I guess we will see the biggest audible differences in the immersive multichannel solutions like for example the new Sonos Era 300, than our audiophile bubble hunt here for even lower distortions which already are close to the borders of inaudibility.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,771
Likes
37,635
I don't think that distortion is low enough actually. If we demand that amps are low distortion, speakers also should meet the same standards. Revel has low distortion to today's standards, but nog low enough to be perfect.

And on Toole, i question him all the time, I intuitive also have some things i don't believe, but no scientific proof (yet), so i am not right entitled to claim something against his science. He is right untill scientificly proven wrong!!!
I've seen measurements of the Quad ESL 63 showing .03% at 96 db SPL @ 1 meter for everything above 100 hz. I don't think it breached 1 % until 40 hz.
 

Axo1989

Major Contributor
Joined
Jan 9, 2022
Messages
2,902
Likes
2,954
Location
Sydney
I've seen measurements of the Quad ESL 63 showing .03% at 96 db SPL @ 1 meter for everything above 100 hz. I don't think it breached 1 % until 40 hz.

Very impressive really. Just don't listen to one of them in mono. :)
 

pablolie

Major Contributor
Forum Donor
Joined
Jul 8, 2021
Messages
2,097
Likes
3,547
Location
bay area, ca
I don't think that distortion is low enough actually. If we demand that amps are low distortion, speakers also should meet the same standards. Revel has low distortion to today's standards, but nog low enough to be perfect.

And on Toole, i question him all the time, I intuitive also have some things i don't believe, but no scientific proof (yet), so i am not right entitled to claim something against his science. He is right untill scientificly proven wrong!!!

But... what kind of distortion? It is known that several form of distortion are actually very benign - bass distortion being a widely-accepted example for that. Hence I find it funny when a bookshelf-style speaker is slammed for bass extension or distortion in what seems to be a very objective discussion of measurements.... until you realize that trying to offload bass frequencies to a sub was never measured/considered - even when that could easily provide a very simple solution to that corner case...

Don't get me wrong, sure being distortion-free sounds like a very desirable goal for many of us - but at what point do such distortion measurements become a placebo?

I think sometimes the wish for perfect measurements is just as irrational as flowery platitudes in subjective hearing "tests". Everything we use in our lives is ultimately a compromise. Sure it'd be impressive to own a car that breaks light speed and is powered by fission or such, and yet we seem happy with the imperfect cars we own. :) Those who talk about science and engineering should always be aware we seldom build anything to perfection. We build stuff to "good enough" standards, even when those standards are high. And I am aware we come full circle with this - since the original question is about what can/should/must be improved, it seems (not sure which one of the three, it does make a difference :-D).
 
Last edited:

Aerith Gainsborough

Addicted to Fun and Learning
Joined
May 4, 2020
Messages
853
Likes
1,280
I'm sure you can push technical specs further. Lower distortion at ever higher volumes while maintaining an even anechonic frequency response and all that jazz.
Though the question is: would it be audible to the average person at all? At what point do "better specs" become a point just for bragging rights? Certainly we are long past that point as far as electronics are concerned.

On a more "down to earth" note: If we factor in human hearing and the available rooms the majority of listeners have at their disposal, I'd say we're pretty much there.
Better tech would just be negated by real life constraints.

Once I go past 100dB Z-peak, stuff sounds shrill and unpleasant to me. You could say "Your AVR is clipping" or "Your lil' Focals are at their limit" and you'd probably be right on both accounts but then I put on headphones and have the same experience.
Certainly one could not say that the RME is clipping at a measly 100dB peak, nor would any headphone worth it's salt be anywhere near it's limit.

What's really limiting in that instance is my actual hearing. Nothing that can be done about that, I'm afraid.

I think, DSP and its ability to mitigate bad rooms is probably the most interesting avenue of further improvement for the common listener. If that process could somehow be further streamlined and automated, it could become more compatible for the masses.
 

kyle_neuron

Active Member
Joined
Jun 18, 2021
Messages
149
Likes
254
These distortion graphs for the Revel F328Be look pretty good to me.


index.php





Revel F208 isn't too bad above 100 hz.
index.php






KEF R3 which is not terribly expensive looks pretty good above 200 hz.


index.php






Then a counter-example. The JBL LSR308 which has good directivity and response, but not looking so great on distortion.


index.php




So larger speakers would seem to have distortion under control other than the bass. So less bass distortion, and I'd think what is left is further resonance reduction. These larger speakers show less of that as well in the better designs.

Other than that I'm pretty sure you'll find speakers that seem to retrieve detail better do so by some minor FR emphasis of this or that region. Or a result of some narrowed directivity in certain ranges.
As amazing as the Klippel NFS is, the plots only show harmonic distortion (THD), which listeners don't tend to object to so much. In my experience, people tend to find other types of distortion more objectionable - two of the main measures being intermodulation distortion (IMD) and non-linear distortions (Gm as defined by Earl Geddes & Lidia Lee in http://www.gedlee.com/Papers/Distortion_AES_I.pdf and tested against listener subjectivity ratings in https://www.dastereo.ru/uploads/short-url/8GkFsBlMf76Y4wR71QZUGlrRrge.pdf)

Anselm Goertz measures distortion in Production Partner magazine by using sine bursts as proposed by Keele and a weighted multitone stimulus that better represents music. Both of these measurement methods include the contributions of IMD, and non-linearity is monitored by increasing the drive level in 1 dB steps and plotting the results normalised against a low-level signal capture.

The results are plotted as a maximum SPL achievable for a given distortion threshold, which befits the use case:
V8-MAX.jpeg

V8-MLT-580x432.jpeg

This is taken from a recent example measurement of a small column loudspeaker. The site and magazine is in German, but it lends itself well to automated translation tools in your browser:

It would be nice to add a variation on this measurement type to the standard Klippel reporting here, as I think it can be quite revealing.
 

fineMen

Major Contributor
Joined
Oct 31, 2021
Messages
1,504
Likes
680
I'm sure you can push technical specs further. Lower distortion at ever higher volumes while maintaining an even anechonic frequency response and all that jazz.
Though the question is: would it be audible to the average person at all? At what point do "better specs" become a point just for bragging rights? Certainly we are long past that point as far as electronics are concerned.

On a more "down to earth" note: If we factor in human hearing and the available rooms the majority of listeners have at their disposal, I'd say we're pretty much there.
Better tech would just be negated by real life constraints. ...

As many do I could pretend to not understand your English. But I won't, and confirm the message. What our dear audiophiles don't get is the mere fact that people use their stereos for a purpose other than showing off in having a stereo. It is a tool to them to listen to preferred (sic!) music; reiterated not listening to preferred speakers. The notorious preference score is a non issue to nearly all consumers, latest when they've got the before mentioned tool for before mentioned purposes.

That said, what are the criteria for a 'good' speaker at all?

Of course all audiophiles who enliven this board / forum rely on Toole's work, which is considered 'scientific' hence reliably ascendent in backing up claims regarding the best choice they made for their own use.

O/k, next criteria. What about, again, predicted preference?

Answer: what the heck is YOUR preference?!

What again and reiterated over and over do YOU feel when listening to music using speakers?

Logic: for what quirky reason do YOU look at others when evaluating YOUR preference?

In post #195 I asked a very humble question which could help to explore one clearly relevant topic, distortion namely (#195). Reaction was active ignorance. Despite the reference to the well reknown Siegfried Linkwitz.

What do you think I should think of you taking best effort to keep a fully ignorant stance?

In case you feel the English is too bad, please visit deepl to translate to your mother tongue, even if it is English. It works. (https://www.deepl.com/translator)
 
Last edited:

fineMen

Major Contributor
Joined
Oct 31, 2021
Messages
1,504
Likes
680
Anselm Goertz measures distortion in Production Partner magazine by using sine bursts as proposed by Keele and a weighted multitone stimulus that better represents music.

It doesn't. At least not for the most common use case of listening to electro pop. The bass content of 'multitone' is way (by 20..40db!!) too low. In other words, the problem of IM is extremely more important than anticipated. But it is actively with full effort ignored by the regular audiophile. You'll see this in the progression of this thread again.
 

Galliardist

Major Contributor
Joined
Jun 26, 2021
Messages
2,558
Likes
3,277
Location
Sydney. NSW, Australia
Preposterous, I asked you all to take the test according to https://purifi-audio.com/blog/tech-notes-1/doppler-distortion-vs-imd-7 with once headphones and second loudspeakers in-room.

It was an attempt to confirm a (comparatively to here local chit chat) qualified conjecture of S. Linkwitz (r/i/p) regarding transformation of Doppler aka phase distortion (ha ha, the other phase) to amplitude distortion.

Guess what, nothing, actually nothing. What the heck are you after?!
Guess what, nothing, I looked to do that test and the audio files aren't there.
 
Top Bottom