Is There a Measurement for “Dynamics”

[mtemur]

What is the speed of a driver?
All the motors in the form of a voice-coil, and the cones attached to them move at the same rate for the same frequency.

Now I think maybe “Group Delay”, or “Transient Response” might be what you are referring to, and those are also measurements to describe the system.

The speed of a driver is similar to acceleration and braking of a car. Just as cars with the same motors can perform very differently from one another, the same is true for drivers with identical voice coils and magnets. Moreover there are very different voice coils and magnet assemblies used in drivers. Drivers differ in performance not only due to voice coil and magnet but also shape and material of the cone. The key element is cone, spider and surrounds in my opinion. A fast driver's cone must be extremely light and also extremely stiff to keep it's shape during long excursions. For this purpose diamond tweeters, carbon fiber and ceramic mids and kevlar and carbon woofers come to mind.

I don't know how speed of drivers are tested or is there a way to test. IMHO you're right "group delay" and "transient response" might be useful along with impulse response.

 

[Alrainbow]

try 'Georgia On My Mind' from Ben Webster's 'Live at the Renaissance', about 90 seconds in there is a drum whack, compare the best digital you can find to the best vinyl. using the AP 45 rpm pressing, i can get 170-200 watt peaks from the vinyl; around 100-120 watt peaks from the digital. experientially it's like a different recording. both are tape sourced. i have the AP Gold CD Doug Sax mastered from the original tapes with tube electronics. he also did the 45.

the digital is very nice, but dynamically they are in different worlds.

What you described is very true , but why ? What is correct ?
My guess is this eq playback curve . Digital does not have one I think ?
To play a digital track there is lot of background software taking place.
It always confuses me how a tape souce transfered to vinyl can end up better , define better is a question that's part of it.
Riaa is EQ but eq is also like a gain at certain freq and this moves the impact from flat .
Even tape has eq playback but the effect I think is not the same.
Now in vinyl playback there are many components but even if we vary then but set up well that twack is still there.

 

[Alrainbow]

The speed of a driver is similar to acceleration and braking of a car. Just as cars with the same motors can perform very differently from one another, the same is true for drivers with identical voice coils and magnets. Moreover there are very different voice coils and magnet assemblies used in drivers. Drivers differ in performance not only due to voice coil and magnet but also shape and material of the cone. The key element is cone, spider and surrounds in my opinion. A fast driver's cone must be extremely light and also extremely stiff to keep it's shape during long excursions. For this purpose diamond tweeters, carbon fiber and ceramic mids and kevlar and carbon woofers come to mind.

I don't know how speed of drivers are tested or is there a way to test. IMHO you're right "group delay" and "transient response" might be useful along with impulse response.

I like my infinites but it's obvious they have faults too. Need to play louder to hear low level sounds and this I think is a function of the drivers used. How fast a driver starts and stops. A lighter driver does both well but may suffer on impact or dynamics .

 

[Alrainbow]

Dynamics are like Porn. I may not know how to describe it in words but I know it when I see/hear it

I know this your wadax goebel Rivera has it in spades , very fast yet great decay

 

[mtemur]

I like my infinites but it's obvious they have faults too. Need to play louder to hear low level sounds and this I think is a function of the drivers used. How fast a driver starts and stops. A lighter driver does both well but may suffer on impact or dynamics .

Exactly. You're very right about lack of impact associated with light drivers. Proper impact is as important as dynamics itself. A fast driver should also have the weight when needed -especially on bass. IMHO conventional cones are better reproducing the weight compared to panels. Maybe it's better to stick paper cones on woofers.

 

[Holmz]

The speed of a driver is similar to acceleration and braking of a car. Just as cars with the same motors can perform very differently from one another, the same is true for drivers with identical voice coils and magnets. Moreover there are very different voice coils and magnet assemblies used in drivers. Drivers differ in performance not only due to voice coil and magnet but also shape and material of the cone. The key element is cone, spider and surrounds in my opinion. A fast driver's cone must be extremely light and also extremely stiff to keep it's shape during long excursions. For this purpose diamond tweeters, carbon fiber and ceramic mids and kevlar and carbon woofers come to mind.

^That^ is mostly incorrect.
The motors and cones are mostly all around the same mass for a given woofer, midrange and tweeter.
Sure a lighter weight tweeter might play to 40 or 50k versus 20k, but there are not accelerating faster if they are both doing 20kHz.
Then they are going at exactly 20kHz.

The choice of the cone material, may be better with a stiffer material. Like carbon fibre or a ceramic.
A cone that holds its shape would not necessarily play the 30 kHz any different at the voice coil, but if the cone break up was not happening, then it would be better sounding without the cone break up

I don't know how speed of drivers are tested or is there a way to test. IMHO you're right "group delay" and "transient response" might be useful along with impulse response.

There is no way to test it because the premise is not correct. There are no fast and slow speeds.
It is like saying that a 12” woofer needs to move faster than a 6” woofer playing the same note.
They are moving at the same rate, but the larger needs to move less distance.

I like my infinites but it's obvious they have faults too. Need to play louder to hear low level sounds and this I think is a function of the drivers used. How fast a driver starts and stops. A lighter driver does both well but may suffer on impact or dynamics .

The driver is not starting and stopping, other than changing its acceleration constantly.
All this focus on the driver sort misses the effect of the XO and the box.
One can push in some number of cycles and it takes a while for a box to resonate at the freqnecy.

Then pull the signal away, and it takes the box a while to ring down.

The higher the order of the box, and the higher the order of the XO, then the more energy is is stored in the electronics and in the box.
Any magic extant in the driver, can be quickly lost in the box.

 

[Alrainbow]

In my case it's panels are open baffel. The drivers are for mids a push pull diagram open both sides.
My Maggie's are also open baffel but I'm not sure where magnets are. What ever the maggies are they play better at low vol and seem to have better low level details.

Lastly on the topic of dynamics and mike L,
Digital verses vinyl I have albums made from digital . Vinyl seems to be more dynamic David benwah is a great example.
I point to riaa as not just an eq playback curve
The fact it's a curve I think has effects on dynamics as well eq playback sound

 

[Al M.]

Then on the other side of the coin, if we have speakers with drivers that have a lot of distortion, then as the motors get well away from their middle point the distortion goes up and harmonics can make it sound louder than the SPL that was requested from the signal.
So they would be “extra dynamic”, or “more dynamic (sounding)”.
Again with the auto stuff, it would be like “oversteer”, where the achieved rate exceeds the requested yaw rate..

[...]

I suspect that there is a spectrum of “extra dynamic”, and some poeple like a lot more than others.

You raise a very good point.

An experienced audiophile once told me decades ago, when I was still in the beginning stages of pursuing this hobby, that "people tend to mistake hardness for dynamics". How right he was, with hardness of course being a kind of distortion.

This is what I wrote in the review of my speakers:

PranaFidelity Dhyana: Two-way quasi line array speaker

The Speaker Retail price/pair: $ 19,500 Dimensions in mm: 1219 x 278 x 455 mm (H x W x D) Dimensions in feet/inch: 4.0 feet x 10.9 inch x 17.9 inch (H x W x D) weight: 45 kg (100 pounds) Below data are largely based on personal communication by Steven Norber, the designer. The Pranafidelity...

www.whatsbestforum.com

Here I should say something about dynamics on the Dhyana speakers in general. While they are consistently among the best I have heard, in some passages that I know well they can, on the surface, seem less impressive and “exciting” than with other speakers. The reason is that these speakers just do not shout, even at loud volume. The lack of “pushing out” dynamics by concomitant surges in hardness and other timbral distortions distinguishes between real dynamics and “fake” dynamics as heard elsewhere. Already with prior improvements to my system (such as changing the DAC and insertion of appropriate power cords), as well as to my room acoustics, I had reduced “fake” dynamics while real dynamics were kept intact, as a check of crucial passages quickly showed upon each improvement. This reduction of “fake” dynamics also brings the dynamic experience closer to that of actual (unamplified) live music in the concert hall. The insertion of the Dhyana speakers into the system has made a great further advance toward that end.

***

The phenomenon of mistaking distortion for dynamics is also the reason why I do not automatically trust reports by others about this or that speaker or component sounding more dynamic, or sounding less dynamic.

 

[Elliot G.]

I know this your wadax goebel Rivera has it in spades , very fast yet great decay

Thank you
One of the great things about Oliver’s speakers is their ability to do micro and macro dynamics. People are wowed by the big stuff however the low level changes are what in my mind makes music sound real. Great musicians and singers know how to use changes in level for emphasis and emotion. The Gobel speakers do this really well so they sound good at every level

 

[Holmz]

…

The phenomenon of mistaking distortion for dynamics is also the reason why I do not automatically trust reports by others about this or that speaker or component sounding more dynamic, or sounding less dynamic.

I trust that their words mean something… (Just I am not sure what it is.)
But if we have data of distortion at various SPL levels, and data on compression, then we can at least start with that.

It is my gold standard “thumb up” when a speaker/system sounds quiet.
Others may like that dynamic sound though… one cannot deny that they must.

Another “thumb up” is when one can listen to a system for hours and not be fatiqued, and a “thumb down” if I can only stand it for ten or twenty minutes.

Really it is probably difficult for manufactures as people buy all sorts of stuff and sometimes having an exciting and dynamic speaker makes them walk themselves out of the door, under a customers arm, more easily.

“Hardness” seems like a good word for it.

 

[microstrip]

I trust that their words mean something… (Just I am not sure what it is.)
But if we have data of distortion at various SPL levels, and data on compression, then we can at least start with that.

As far as I see it we do not have meaningful data on compression and the very little data on distortion is a THD figure, sometimes dominated by harmonics that are not relevant for such effects.

To make thing even less clear we should remember that cables also change our feeling of dynamics and there is not any audio frequency measurement that shows it.

 

[Robh3606]

As far as I see it we do not have meaningful data on compression and the very little data on distortion is a THD figure, sometimes dominated by harmonics that are not relevant for such effects.

Why would you say that? Here you have testing over a 26dB SPL window looking specifically for changes at each measured point. It is more than obvious which system suffers from less compression. Clear cut to me? What else are you looking for?

WRT distortion? Easily measured lots of data for the most part meaningless.

Rob

 

[microstrip]

Why would you say that? Here you have testing over a 26dB SPL window looking specifically for changes at each measured point. It is more than obvious which system suffers from less compression. Clear cut to me? What else are you looking for?

WRT distortion? Easily measured lots of data for the most part meaningless.

Rob

Thanks. Could you supply links to this data so we can appreciate it and know the exact measuring conditions? It seems I am looking for something more than you ...

 

[Robh3606]

Thanks. Could you supply links to this data so we can appreciate it and know the exact measuring conditions? It seems I am looking for something more than you ...

Dynamic Range (Instantaneous Compression Test)​

The below graphic indicates just how much SPL is lost (compression) or gained (enhancement; usually due to distortion) when the speaker is played at higher output volumes instantly via a 2.7 second logarithmic sine sweep referenced to 76dB at 1 meter. The signals are played consecutively without any additional stimulus applied. Then normalized against the 76dB result.

The tests are conducted in this fashion:

1. 76dB at 1 meter (baseline; black)
2. 86dB at 1 meter (red)
3. 96dB at 1 meter (blue)
4. 102dB at 1 meter (purple)

The purpose of this test is to illustrate how much (if at all) the output changes as a speaker’s components temperature increases (i.e., voice coils, crossover components) instantaneously.

You can look for whatever you want but I don't think there is much else needed. Distortion measurements as well go through the speakers testing.

Rob

Erin's Audio Corner

The Sound of Science

www.erinsaudiocorner.com

 

[microstrip]

Dynamic Range (Instantaneous Compression Test)​

The below graphic indicates just how much SPL is lost (compression) or gained (enhancement; usually due to distortion) when the speaker is played at higher output volumes instantly via a 2.7 second logarithmic sine sweep referenced to 76dB at 1 meter. The signals are played consecutively without any additional stimulus applied. Then normalized against the 76dB result.

The tests are conducted in this fashion:

1. 76dB at 1 meter (baseline; black)
2. 86dB at 1 meter (red)
3. 96dB at 1 meter (blue)
4. 102dB at 1 meter (purple)

The purpose of this test is to illustrate how much (if at all) the output changes as a speaker’s components temperature increases (i.e., voice coils, crossover components) instantaneously.

You can look for whatever you want but I don't think there is much else needed. Distortion measurements as well go through the speakers testing.

Rob

Erin's Audio Corner

The Sound of Science

www.erinsaudiocorner.com

I could see that the graph is carried with "2.7 second logarithmic sine sweep" - impossible to correlate with real music signals. A sweep signal is a permanent peak in a coil for some time. It is a good test signal for manufacturers, meaningless for audiophiles, except those who listen mostly to electronic music with high content.

Nice site, BTW. But measurements per se are not science!

 

[Robh3606]

I could see that the graph is carried with "2.7 second logarithmic sine sweep" - impossible to correlate with real music signals. A sweep signal is a permanent peak in a coil for some time. It is a good test signal for manufacturers, meaningless for audiophiles, except those who listen mostly to electronic music with high content.

Nice site, BTW. But measurements per se are not science!

Baloney! You use sine sweeps for everything from distortion to frequency response. When I use CLIO for speaker testing Sine and Impulse are the primary test signals used.

They are doing consecutive sweeps increasing the level to look for the effects of heating and a sine sweep is perfect for that.

It doesn't have to be an exact equivalent to music, the purpose is to show if heating occurs and where in the frequency spectrum. Which it shows very well. No standard test signal is a exact equivalent to music. This in no way makes them inappropriate for their purpose.

They are test signals designed to show where issues occur in an easily understandable format. They are not music and there is no requirement for them to be. It does not diminish their usefulness or accuracy.

Just because it doesn't fit your personal paradigm doesn't make it an invalid test.

Measurements are not science? Doesn't even warrant a response keep you head firmly placed in the sand.

Rob

 

[Holmz]

As far as I see it we do not have meaningful data on compression and the very little data on distortion is a THD figure, sometimes dominated by harmonics that are not relevant for such effects.

Klipple machines do give the ability to measure compression comparing say 86, 96 and 102dB referenced against 76dB.
The speaker shown is “Extra dynamic” between 600-800 Hz, and “under dynamic’’(compressed) below ~500 Hz.
They (Klipple) also supply the distortion versus frequency at multiple levels, so there are measurement methods to quantify and measure things that have been described subjectively.

To make thing even less clear we should remember that cables also change our feeling of dynamics and there is not any audio frequency measurement that shows it.

And ^right there^ is a lot of the problem.
And the problem is that there is so much reliance and favour for the frequency domain.

Most of what we talk about with speed and transient response need to be expressed in the time-domain.
Group delay is shown in the frequency domain though… but transfer functions and impulse response and in the time-domain.

 

[Robh3606]

Most of what we talk about with speed and transient response need to be expressed in the time-domain.
Group delay is shown in the frequency domain though… but transfer functions and impulse response and in the time-domain.

They are all related and phase, frequency response, group delay, CSD, STEP are all derived from the same impulse response. Flat frequency response will give you flat phase and group delay both of which are directly related to FR.

Rob

 

[Holmz]

They are all related and phase, frequency response, group delay, CSD, STEP are all derived from the same impulse response. Flat frequency response will give you flat phase and group delay both of which are directly related to FR.

Rob

(Well done)

Well technically yes, but we almost always look solely at the amplitude. And seldom the phase.
One can have flat amplitude in the FFT, and have wild phase.

I do not know how one can do a complex conjugate (in the middle) and the compute the impulse response.