speaker distortion

[ThorstenL]

Konnichiwa,

for those of us who want to rely on science,

It is good to remember that the distinction between science and academica is that the scientist operates with and bases his understanding and theories etc on experiemntal observation while the academic bases his understanding and theories etc on Authority (eg. outdated and usually factually at least inaccurate if not totally wrong textbooks).

it is probably reasonable to say that speaker distortions are a few magnitude higher than amp distortion.

Only with reference to certain types of distortion. I would again recommed to consider Earl Geddes work on the subject.

and in our quest to lower distortion, amps aren't probably the bottleneck.

Lower "WHAT" distortion? Lower Total Harmonic Distortion (which illustrably has no correlation with subjectively percieved "undistorted sound")? Or lower percieved distortion?

So basically, do you want to be academic about it or scientific?

Sayonara

 

[5th element]

we knew that very well,




I would agree with that too: those high-end guys would then be really making something that is based on sound science, rather than voodoos and moon-rock cables, .

for those of us who want to rely on science, it is probably reasonable to say that speaker distortions are a few magnitude higher than amp distortion. and in our quest to lower distortion, amps aren't probably the bottleneck.

Yes but the original point I was getting at was a bit vauge.

What I meant was that speaker distortion is less benign then amp distortion such that 1%+ is clearly tollerable when we listen to speakers, not saying that is doesnt sound better with lower distortion. But a 1%+ speaker system can sound very clean. Whereas if we had an amplifier that had 1%+ distortion likely hood is it would sound terrible, I know SE valve amps asside, but generally speaking.

Also what is it exactly what dyn are doin in their drivers that other manufactures are not? and why are they not doing it. Looking at SEAS dis plots they seem to be fairly good and you could see that using the correct drivers you could make a low distortion system that would rival the dyn, however why do manufactures not employ low distortions systems to their drivers, do they cost massive amounts of money? Looking at the C4 they dont use particularly large drivers so excursion has to be higher then if they were larger.

 

[millwood]

What I meant was that speaker distortion is less benign then amp distortion such that 1%+ is clearly tollerable when we listen to speakers,

and when we listen to amps, that 1% thd suddenly sounds horrible?

let's me put this differently. Say you have a perfect amp and an imperfect speaker generating 1% thd (for argument sake, make that all in one harmonic). This is system 1.

say you have another system, system 2, that consists of an imperfect amp that generates the same 1% thd, and a perfect speaker.

Which of the two systems would sound better?

 

[phase_accurate]

Looking at SEAS dis plots they seem to be fairly good and you could see that using the correct drivers you could make a low distortion system that would rival the dyn, however why do manufactures not employ low distortions systems to their drivers, do they cost massive amounts of money?

IMO there are two things you can put into a driver to reduce it's deistortion. And this is a veeeeeery general way of looking at it:

1.) Knowledge

2.) Increased technical effort

Both cost not that much if spread ofer many thousand pieces manufactured.
But you also SAVE a lot if you don't use them, especially #2 if you manufacture large quantities !!!!

Regards

Charles

 

[ThorstenL]

Konnichiwa,

What I meant was that speaker distortion is less benign then amp distortion

Surely you mean MORE BENIGN (even if that point is still highly debatable)?

such that 1%+ is clearly tollerable when we listen to speakers, not saying that is doesnt sound better with lower distortion. But a 1%+ speaker system can sound very clean. Whereas if we had an amplifier that had 1%+ distortion likely hood is it would sound terrible, I know SE valve amps asside, but generally speaking.

Either SE Amp or generally speaking, saying "this DUT has 1% distortion" is saying absolutely nothing. The statement has ZERO informational content and even if start qualifying the power level and other items it still has ZERO informational content with regards to what is audible.

Please take the Geddes Test, if you wish to proove this to yourself. A cheap PC Speaker setup is sufficient for a quick test.....

Undistorted Wave file - Music Excerpt

Distorted Wave File, THD = 0.1%, IMD = 0.3%

Distorted Wave File, THD = 9.6%, IMD = 22.4%

More on the Geddes Lee Distortion Metric and Distortion perception

Also what is it exactly what dyn are doin in their drivers that other manufactures are not?

IIRC, copper sleeves, in some models aluminum voice coil carriers, carfull design of the mechanical and magnetic system for high linearity.

and why are they not doing it.

Cost? You will have to ask them....

Looking at SEAS dis plots they seem to be fairly good and you could see that using the correct drivers you could make a low distortion system that would rival the dyn,

There are number of "Skandinavian" Manufacturers who's driver engineering was influenced by the same self Engineer, Ragnar Lien, namely Dynaudio, Seas and Scan Speak. All these companies make comparably low distortion drivers, but you do pay for it. Other since have started to copy some of the features, but there is little consistence. ATC has their own methodes of making low distortion cone drivers.

That's about the lot in "HiFi", in Pro Audio (Sound Reinforcement and Studio) low distortion (and compression) drivers are much more common, but you do have to leave the "low cost commodity" ranges of the better manufacturers and avoid some manufacturers (like most Far Eastern ones and general Eminence Drivers).

however why do manufactures not employ low distortions systems to their drivers, do they cost massive amounts of money?

Generally - yes.

Fitting a copper sleeve (or a ferrite sleeve) in the airgap of the magnet system widens the airgap, to keep a given fieldstrength you need a stronger (more expensive) magnet.

The parts required to make a pretty linear magent system are usually not made in quantity in the Far east, neither are the comodity magnet system parts (magnet, polepieces and plates) readily available in suitable material quality. This means comparably small production runs of special items, invariably more costly dur to lack of economics of scale.

To give you an idea, a well reviewed UK made Speaker with 4pcs of 5" driver per speaker selling for > £ 500 (nearly $ 900) per pair needed the drivers in one speaker replaced. At spare part pricing the drivers where 5 pound each. In quanteties of 1,000 from Audax/Harman they where < £ 1 each. No margin for £ 10 Magnet system in that, is there....

Sayonara

 

[millwood]

Both cost not that much if spread ofer many thousand pieces manufactured.
But you also SAVE a lot if you don't use them, especially #2 if you manufacture large quantities !!!!

Regards

Charles

for the same reason, machining the tip of your car antenna down to 0.01mil isn't going to be that difficult: the knowledge is common, the equipment is sunk, etc. But nobody does it. Why?

because no one cares about how precisely machined the tip of your antenna is.

Can any of us hear, in a scientific test, down to 1% THD? or 0.1%? or 0.01%? you name it.

I suspect few of us can hear 10% thd in loud music passages. If that is indeed true, what value is there to incur costs to lower speaker distortion down to levels that are not audible?

the majority of the population will not pay for that. And for those who will, they do pay through their nose for it. I can easily find drivers custom made for low distortion at prices over $1K each. Will you pony up the money?

If you don't, why should companies invest in such a product?

 

[5th element]

Well I can deffo hear the greater distortion in both of those musical clips. The 0.1% is noticeable when compared to the 0% but its not a bad sound like the 9.6% is. Its more a very very very slight haze is on the sound. Note I would not be able to hear this if my hifi wasnt as good as it is. Now im not making assumptions that my hifi is super duper, but as far as things go its very vice free and distortion less.

One thing is clear however lower distortion is better.

 

[millwood]

Well I can deffo hear the greater distortion in both of those musical clips.

how many of you can hear if the same songs are unlabeled?

 

[tschrama]

Hi Kuei Yang Wang,

thx for the graphs but ...


next time you want to show-off your god-like insight, please read my posts a little better.


I'm getting of you pleople trying to win some argument and not trying to stick to a question, but showing-off with strong language and pseudo knowlegde. I seemd to have become impossible to have a interesting discussion on this forum, without someone racing right through the discussion, without having even a tiny bit of modesty (?) ...


Have fun all,


Sayonara

 

[sam9]

"What I meant was that speaker distortion is less benign then amp distortion such that 1%+ is clearly tollerable when we listen to speakers, not saying that is doesnt sound better with lower distortion. But a 1%+ speaker system can sound very clean."

There may also be a question of what we have come to expect. There are reports of the early listeners to Edison cylinders that claimed what they hear was indistinguishable from an actual human voice. In present times, I recently read a piece by Siegfried Linkwitz where he talks about using headphones/earphones to establish a baseline for loudspeaker evaluation -- head/earphone (good ones anyway) having much lower distortion than any speaker. This inspred me to do some recent listening with headphones in comparrison to speakers. I tried to ignore the spatial properties of the phones and just concentrate on the basic sound. I reached one firm comclusion - speakers (even the very best) have lots and lots of room for improvement. I also reached some very tentative impressions: ignoring spatial and imaging considerations, planar dipoles (ESLs, Magnepans, etc) are the least different from good headphones and traditional enclosed box speakers are usually farther away. I'm interprecting this for now as a rough indication relative distortion levels (what kind of distortion is an other question).

Anyway I can't think of any amplifiers that exhibtssuch disparities. So I'm inclined to agree with the view that the type of speaker and all that goes wth it (room effects, EQ, etc) tend to have a greater impact on what we hear than amplifiers.

 

[ThorstenL]

Konnichiwa,

Well I can deffo hear the greater distortion in both of those musical clips. The 0.1% is noticeable when compared to the 0% but its not a bad sound like the 9.6% is.

Are you sure you listened to the right track?

If you downloaded to your Harddrive and played from there - Phantom9.wav is the track with 0.1% THD, Phantom8.wav is the one with 9.6% THD!

Everyone to whom I played both tracks considered Phantom9.wav to be very distorted while Phantom8.wav was considered by all as sounding in effect "clear" and no knowing it was the "distorted" track some preferred it to the undistorted track. I did this with 4 or 5 people at work, quite unscientifically, admittedly...

Maybe you want to check again you had the right track?

One thing is clear however lower distortion is better.

Let me play Sporting Life here, "It ain't neccesarily so...".

Sayonara

 

[ThorstenL]

Konnichiwa,

next time you want to show-off your god-like insight, please read my posts a little better.

moushiwake arimasen

I misread your position. Still, there are enough floating around who insist speakers are "low distortion". Another point I tried to bring out is that there are many types of distortion and the presence of high levels of HARMONIC distortion (and related Intermodulation) in amplifiers does not neccesarily mean they will mask small amounts of distortion of a different nature.

This whole "THD" game is completely meaningless anyway, just like about any other one-dim,ensional number bandies about in audio advertising. We are dealing with a very complex set of relationsships, simply trying to tie them to a single "number of merit" is daft, to say the least.

Clearly I did not say it right.....

Sayonara

 

[jeff mai]

1% thd suddenly sounds horrible?

generating 1% thd

same 1% thd

THD is 100% meaningless. 1% THD could be undetectable or horrible sounding depending on how it is made up.

 

[fdegrove]

Hi,

1% THD could be undetectable or horrible sounding depending on how it is made up.

Amen.

Cheers,

 

[5th element]

Aye im sure I got the correct tracks, to me its audible which is which. Zero is very clear 0.1 has a slight haze of fuzz, which can be perceived as more pleasent sounding as it would mask maybe inner detail and maybe a slight harshness present on harsh bits . The distortion added hear doesnt seem unpleasent as shown when you listen to the 9.6% it has a smoothing effect.

Perhaps its just what we all listen out for in music, I for one really like a clear as a bell like sound and the more clear the hifi the better in my eyes, so the more euphonic, smooth nature of the 0.1 although appealing, is not what I prefer.

 

[ThorstenL]

Konnichiwa,

Aye im sure I got the correct tracks, to me its audible which is which.

I ask again, have you checked? The 0.1% THD track is actually audibly SEVERELY distorted (burned to CD, played via Windoze Media Player and Winamp), no slight haze......

Here again a link to a track distorted using an algorythm resulting in 0.1% THD and 0.3% IMD to be measured using the usual methodes.....

Question to all, does that track linked above sound "has a slight haze of fuzz, which can be perceived as more pleasent sounding"?

Sayonara

 

[jeff mai]

Aye im sure I got the correct tracks, to me its audible which is which.

It is a possibility that you would have a different opinion in a blind selection of these tracks. The desire to be sensible can be as strong as the desire to be fashionable. Your use of the word "euphonic" immediately suggests bias.

Beauty is in the eye of the beholder: to some a low distortion figure on the page is as blinding / deafening as a machined faceplate and expensive components.

Again, I'm only stating that this is a possibility that must be considered, not that it is fact.

 

[phase_accurate]

For those who want to use a microphone to measure distortion quantities: What if anyone would not only check what his speakers does to sinusoids but also to REAL MUSIC (by feeding the input signal and the mic output to a dual trace scope and watch the difference; don't forget to take the delay caused by sound-travel into account) ?
I am convinced the quantitative deviation would be much larger in the latter case than just some THD added to a sinusoid !

Regards

Charles

 

[till]

For those who want to use a microphone to measure distortion quantities: What if anyone would not only check what his speakers does to sinusoids but also to REAL MUSIC (by feeding the input signal and the mic output to a dual trace scope and watch the difference; don't forget to take the delay caused by sound-travel into account) ?

i did.

you can´t see much, as the signal is not repetetive and too fast, and you can´t trigger on it. One would need to use a very short sample of music and repeat it, to have a stable curve one the CRT.

Even more interesting would be running a FFT one the 2 signals, autocorrelate them, and find some way to make an mathematical comparison on the complex music signal, not on sinewaves.

 

[phase_accurate]

you can´t see much, as the signal is not repetetive and too fast, and you can´t trigger on it. One would need to use a very short sample of music and repeat it, to have a stable curve one the CRT.

Ouch!

I will try it on my storage scope as soon as I find some time to do so.

But a signal that is a little more complex (but still not as complex as music) than a sinusoid is a rectangular. This will also show the "deviations" I am talking of.

Regards

Charles