• 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!

Are speakers the Achilles heal in sound reproduction?

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
We all love the -120db SINAD DACs, but it seems that loudspeakers are the limiting factor in terms of sound reproduction.

I was once told that your speakers were the most important component that would determine what your ears hear (coloration, distortion, etc)

With that said, I am glad to see these speaker measurements in ASR. I would really like and wish audio science can concentrate on how loudspeakers reproduce sound. Frequency response is one thing. But distortion and ability to handle large swings in voltage / dynamic range are another. For example, how do loudspeakers behave when you are in your listening position, not the arbitrary measurement scale of 1meter? What is the THD? How do they behave when you feed 50 or 100watts instead of 1watt?

Who cares if your dac has -115db SINAD when the best your speakers can do is -40db or -50db SINAD.
 
Last edited:
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
I also have a question to all the engineers here. How is the passive crossover inside the speaker box altering sound? Does it introduce any distortion? Can you measure distortion that these passive crossovers create?

My guess is that they do introduce lots of distortion, but since the transducer distortions hover way above that level, we don’t care.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,985
Location
Riverview FL
Are speakers the Achilles heal in sound reproduction?

No.

They make the sound.

Imperfectly, but often remarkably well, from my viewpoint.

But distortion and ability to handle large swings in voltage / dynamic range are another. For example, how do loudspeakers behave when you are in your listening position, not the arbitrary measurement scale of 1meter?

My big ones do fine.

The little JBL 308s run out of steam at higher levels.


What is the THD?

Low on the big ones, high on the little ones.


How do they behave when you feed 50 or 100watts instead of 1watt?

I have 700W available for the mains, not yet used, to my knowledge.

300 or probably more at peaks still sounds clean but is way too loud overall.

The JBL internal amps are rated 56W.


How is the passive crossover inside the speaker box altering sound?

Changes amplitude with frequency for the individual drivers, and to some degree, phase.


My guess is that they do introduce lots of distortion

I don't think so, but can't quote a reference.

Crossovers are coils and capacitors and resistors.

The same devices show up in the very low distortion parts of active circuits, so I don't think they particularly present a problem with distortion.


but since the transducer distortions hover way above that level, we don’t care.

The transducer distortions increase as volume increases, but usually at a greater rate.

With pure tones, distortion can be heard rather easily.

Music is full of what might look like distortion to an analyzer, and small problems in that area are hidden from being noticed until quite severe


A musical signal... Single bass guitar note. Compare to a test tone with "no distortion".


1584780745329.png

.

Test tone via speakers. Same level at listening position, two different speaker designs:

The tone from the first speakers is audibly distorted, the third harmonic is noticed, the second speaker does not audibly distort the tone.


1584780988725.png


1584781017076.png
 
Last edited:
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
Appreciate the informative response!

When you say THD is low, is it at levels where we obsess about DACs?

When I mention dynamic range, I mean the usual quoted 12-20db above average listening level. G
Let’s say the amplifier can handle the transients, can the loudspeaker handle it without compression or overly high distortion?

Would like to see how loudspeakers behave when fed high voltages in addition to all the rest.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,124
Likes
2,785
I also have a question to all the engineers here. How is the passive crossover inside the speaker box altering sound? Does it introduce any distortion? Can you measure distortion that these passive crossovers create?

My guess is that they do introduce lots of distortion, but since the transducer distortions hover way above that level, we don’t care.

Passive crossover doesn't introduce distortion in any practical sense. Yes, you can measure it, but as it consists of simple passive RLC elements it would measure somewhere at the bottom of the measurement range of the instrument (say -150dB or lower). It is not something you should be concerned about.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,124
Likes
2,785
When you say THD is low, is it at levels where we obsess about DACs?

Look at the small rectangular window with distortion figures for harmonic components, it says THD is 0.165%. That is a figure only good speakers can achieve yet it is not even remotely comparable to DACs THD figures. ;)
 
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
I trust you guys about passive crossovers not introducing distortion. But would still like to see some actual data, especially with complicated 3 to 4 way crossovers some high end speakers have.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,985
Location
Riverview FL
Who cares if your dac has -115db SINAD when the best your speakers can do is -40db or -50db SINAD.

Feeding a poor signal to the speakers won't make things better.

When you say THD is low, is it at levels where we obsess about DACs?

No.

Above see .166% in the third graphic.

Let’s say the amplifier can handle the transients, can the loudspeaker handle it without compression or overly high distortion

They can until they can't, depending on their capabilities.

Here are some sweeps at increasing volumes, I don't see compression, and measured distortion goes down until that trend reverses at the very highest levels.

Ambient or measurement noise dominates the distortion reading up to 90dB or so. Which is quite loud for a pure tone.

You can see noise intrude on SPL in the lower left below:

I don 't see any compression.

1584782082206.png


And the distortion measures.

The lowest SPL (top trace) measures the highest distortion - there are a few missing due to the scrolling of the sources, but the lowest SPL trace at 944Hz claims 9.27% THD.

Then traces bunch up as the actual speaker distortion begins to rise out of the noise floor.

The highest SPL (bottom orange trace at 944Hz) gives the lowest distortion at 944Hz, 0.071%

1584782360959.png
 
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
Look at the small rectangular window with distortion figures for harmonic components, it says THD is 0.165%. That is a figure only good speakers can achieve yet it is not even remotely comparable to DACs THD figures. ;)

My exact point! Seems speakers ARE the limiting factor
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,985
Location
Riverview FL
I trust you guys about passive crossovers not introducing distortion. But would still like to see some actual data, especially with complicated 3 to 4 way crossovers some high end speakers have.

The crossover on the speaker above is at 180Hz. It is a 2-way15"x48"electrostatic panel and 12" sealed woofer.

I can provide no other data.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,124
Likes
2,785
I trust you guys about passive crossovers not introducing distortion. But would still like to see some actual data, especially with complicated 3 to 4 way crossovers some high end speakers have.

Well you can search for it but I doubt you will find many actual measurements as THD is simply not an issue with simple networks made of passive components.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,749
Location
Alfred, NY
I trust you guys about passive crossovers not introducing distortion. But would still like to see some actual data, especially with complicated 3 to 4 way crossovers some high end speakers have.

If the passive components don't individually distort, putting them together won't suddenly make them do so. Linear superposition.

The number one "distortion" is two channel stereo.

Then rooms and speakers.
 

QMuse

Major Contributor
Joined
Feb 20, 2020
Messages
3,124
Likes
2,785
My exact point! Seems speakers ARE the limiting factor

As a mechanical component that actually produces sound instead of processing analog electric signal of course they are. Don't expect Nobel price for your discovery. ;)
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,985
Location
Riverview FL
My exact point! Seems speakers ARE the limiting factor

I find recordings to be the limiting factor far more often.

Some poor, some good, some sublime.
 
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
Feeding a poor signal to the speakers won't make things better.



No.

Above see .166% in the third graphic.



They can until they can't, depending on their capabilities.

Here are some sweeps at increasing volumes, I don't see compression, and measured distortion goes down until that trend reverses at the very highest levels.

Ambient or measurement noise dominates the distortion reading up to 90dB or so. Which is quite loud for a pure tone.

You can see noise intrude on SPL in the lower left below:

I don 't see any compression.

View attachment 55249

And the distortion measures.

The lowest SPL (top trace) measures the highest distortion - there are a few missing due to the scrolling of the sources, but the lowest SPL trace at 944Hz claims 9.27% THD.

Then traces bunch up as the actual speaker distortion begins to rise out of the noise floor.

The highest SPL (bottom orange trace at 944Hz) gives the lowest distortion at 944Hz, 0.071%

View attachment 55250

:) Appreciate the explanation. I would still love to see how and where speakers behave normal and where and when they start to collapse as voltage increases
 
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
As a mechanical component that actually produces sound instead of processing analog electric signal of course they are. Don't expect Nobel price for your discovery. ;)

haha.. I am just trying to get some validation from you guys, nothing else.
 
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
Let me put it this way, we do tests with amps showing when they start clipping and collapsing. Can we have and show some data on when speakers start compressing and collapsing as the voltage increases? I don’t think ai am askin a very unreasonable question here.
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,985
Location
Riverview FL
I would still love to see how and where speakers behave normal and where and when they start to collapse as voltage increases

They "clip" as they reach the limits of excursion.

Basically, you'll see, when measuring a sine tone, more and more harmonics at higher and higher levels.

It will "look" very similar to what an amp does.
 
OP
G

gags11

Senior Member
Forum Donor
Joined
Feb 8, 2019
Messages
356
Likes
528
Thee are people here arguing that a good amp is judged by continuous max output vs distortion, etc. why not put speakers to similar tests?
 

RayDunzl

Grand Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
13,204
Likes
16,985
Location
Riverview FL
I don’t think ai am askin a very unreasonable question here.

I'm not gonna blow up my gear to satisfy your curiosity.

Maybe someone else will oblige.
 
Top Bottom