A Broad Discussion of Speakers with Major Audio Luminaries

[Holmz]

Fast transient response is the same thing as high frequency response. The steep leading edge of the signal rise is the same either way.

So, the concept of a fast woofer makes no sense to me. If it’s top frequency is, say, 800 Hz, then a higher transient response to accommodate the steeper rise of a higher frequency will never be explored—the crossover will have filtered it out.

The sharp rise of a transient is shaped by the tweeters, and any subsequent ring or decay might be conveyed by all the drivers. There are correlations in the time domain, of course.

That’s how declicking works—it filters out transients that are unrealistically steep or that are clipped, which I think amounts to the same thing.

It seems to me that a speaker that muffles the thump of a kick bass drum that is close to the maximum signal amplitude doesn’t suffer from being “too slow.” It suffers from lack of dynamic range causing amplitude compression. That fault seems to me a lot more common than a lot of people realize if they use their systems at high listening levels.

Rick “small speakers seem to compress or ring bass signals more easily than large speakers, port notwithstanding” Denney

Or maybe group delay can be part of it?

 

[levimax]

Fast transient response is the same thing as high frequency response. The steep leading edge of the signal rise is the same either way.

So, the concept of a fast woofer makes no sense to me. If it’s top frequency is, say, 800 Hz, then a higher transient response to accommodate the steeper rise of a higher frequency will never be explored—the crossover will have filtered it out.

The sharp rise of a transient is shaped by the tweeters, and any subsequent ring or decay might be conveyed by all the drivers. There are correlations in the time domain, of course.

That’s how declicking works—it filters out transients that are unrealistically steep or that are clipped, which I think amounts to the same thing.

It seems to me that a speaker that muffles the thump of a kick bass drum that is close to the maximum signal amplitude doesn’t suffer from being “too slow.” It suffers from lack of dynamic range causing amplitude compression. That fault seems to me a lot more common than a lot of people realize if they use their systems at high listening levels.

Rick “small speakers seem to compress or ring bass signals more easily than large speakers, port notwithstanding” Denney

One of the biggest contributors to time domain issues with transient response are filters. Filters (electronic and mechanical) ring and delay the signal in complex ways which leads to the energy in a transient being delivered over a longer period of time than was in the original signal. The audibility of this is does not seem to be well defined but I think there is more to it than just "transient response = high frequency response" for an entire speaker system.

 

[j_j]

I think a lot of what people are talking about when they say 'faster' in reference to a loudspeaker is good transient response. The ability to respond instantaneously to changing input. The better a speaker does this the lower the distortion the speaker contributes to what a listener hears. Especially when we start talking about ported woofers and subwoofers does this become important. Not saying the overhang and phase shift that occurs in a ported enclosure is any more or less important than other types of distortion introduced by speakers, like ringing, cone break-up or resonances within the motor structure of the speaker in question, but distortions introduced by the speaker's interaction with its enclosure (including, but not limited to, the presence of a port) all contribute the inaccuracies we hear when listening.

Thank God speaker design has evolved and it is possible to get a modern designed speaker that has very low distortion in comparison to what was mainstream 50 years ago. Between advances in crossover design, improvements in cone and suspension materials, and improved magnet structures, the SOTA is leaps and bounds beyond what it used to be. I think it is fair to say that today its quite possible to find speakers that have well under 1% distortion from a number of different manufacturers, where 50 years ago, a speaker with 5% distortion was considered good or even excellent.

/pedant
Can we please remember to separate linear from nonlinear issues? Nonlinear issues, what is usually called distortion, can create new frequencies, things like harmonics, IM products, etc. Linear changes can change frequency response, phase response, and the like.

As to 1% distortion from a speaker, how are you measuring that, please? Start with a definition of what you mean by "distortion", and at what point in what space you measure the speaker.

 

[Jflijohn]

Perhaps you meant Jon Batiste as John is a military general?

Only an esteemed Moderator would know of the great general and concert pianist General John Batiste

 

[RickS]

Only an esteemed Moderator would know of the great general and concert pianist General John Batiste

Or google lol! With your permission will fix the misspelling in your post.

 

[Jiraya369]

Fast transient response is the same thing as high frequency response. The steep leading edge of the signal rise is the same either way.

So, the concept of a fast woofer makes no sense to me. If it’s top frequency is, say, 800 Hz, then a higher transient response to accommodate the steeper rise of a higher frequency will never be explored—the crossover will have filtered it out.

The sharp rise of a transient is shaped by the tweeters, and any subsequent ring or decay might be conveyed by all the drivers. There are correlations in the time domain, of course.

That’s how declicking works—it filters out transients that are unrealistically steep or that are clipped, which I think amounts to the same thing.

It seems to me that a speaker that muffles the thump of a kick bass drum that is close to the maximum signal amplitude doesn’t suffer from being “too slow.” It suffers from lack of dynamic range causing amplitude compression. That fault seems to me a lot more common than a lot of people realize if they use their systems at high listening levels.

Rick “small speakers seem to compress or ring bass signals more easily than large speakers, port notwithstanding” Denney

does transient response matter? how would i check if a speaker has fast transient response or something?

 

[PristineSound]

does transient response matter? how would i check if a speaker has fast transient response or something?

Glad you brought this up. Supposedly no, it does not, and driver material is immaterial. Frequency response and directivity is king.

If that is the case, I would love to know why Revel use Be and Ti for their drivers on their high end models. I have a feeling I know what the answer is, but I want to hear it directly.

 

[Jiraya369]

Glad you brought this up. Supposedly no, it does not, and driver material is immaterial. Frequency response and directivity is king.

If that is the case, I would love to know why Revel use Be and Ti for their drivers on their high end models. I have a feeling I know what the answer is, but I want to hear it directly.

hmm ok thanks

ig they use those materials because they know it quite well so they have mastered the directivity of those drivers. And they're also able to push these to extreme levels without distortion/compression.

Seems to me that each manufacturer uses what it knows best instead of having a secret sauce like "THE DIAMOND TWEETER WILL MAKE YOU HEAR THINGS YOU'VE NEVER HEARD BEFORE" lol

 

[Holmz]

One of the biggest contributors to time domain issues with transient response are filters. Filters (electronic and mechanical) ring and delay the signal in complex ways which leads to the energy in a transient being delivered over a longer period of time than was in the original signal. The audibility of this is does not seem to be well defined but I think there is more to it than just "transient response = high frequency response" for an entire speaker system.

^100%^
The lows can “hang”, with the highs being immediate.

does transient response matter? how would i check if a speaker has fast transient response or something?

Does athe Impulse response measurement/graph help?

 

[Jiraya369]

^100%^
The lows can “hang”, with the highs being immediate.




Does athe Impulse response measurement/graph help?

i checked impulse/step response and group delay and i must say i dont understand anything about em. I look at some and go "huh looks like shit" and amir and erin go "CLEAN!" or "no issues here". Like bruh im just so lost

 

[Sokel]

i checked impulse/step response and group delay and i must say i dont understand anything about em. I look at some and go "huh looks like shit" and amir and erin go "CLEAN!" or "no issues here". Like bruh im just so lost

Maybe seeing the (short of) ideal ones would help:

(that's only electrical, to be clear about it)

 

[Jiraya369]

Maybe seeing the (short of) ideal ones would help:

View attachment 497450
(that's only electrical, to be clear about it)

thanks for trying to help but im still quite lost. I looked at genelec, neumann, ascilab, kef and a few other stuff and i still dont find a common thing between em that makes it "good"

 

[Sokel]

thanks for trying to help but im still quite lost. I looked at genelec, neumann, ascilab, kef and a few other stuff and i still dont find a common thing between em that makes it "good"

Look at the green (short of) ideal one I posted above.
And now look the Gen 8361A one:



The difference is a little pre - ringing (due to DSP probably) and the negative spike before the impulse.
As Amir notes, really close to ideal.

 

[j_j]

I would urge those who are trying to understand the contributions of different parts of a loudspeaker to the impulse response to get some kind of plotting software, and try it for themselves. Plot DC to 20 Hz, 20+ to 400, 400-4000, and 4000-nyquest.

Then add them all together, and surprise!

 

[Jiraya369]

Look at the green (short of) ideal one I posted above.
And now look the Gen 8361A one:

View attachment 497454

The difference is a little preprinting (due to DSP probably) and the negative spike before the impulse.
As Amir notes, really close to ideal.

thanks but it looks quite messy to me, and then the ascilab f6b for example is called "nice and smooth" yet it loosk RADICALLY different. i'll look up some things on them later on.

 

[j_j]

thanks but it looks quite messy to me, and then the ascilab f6b for example is called "nice and smooth" yet it loosk RADICALLY different. i'll look up some things on them later on.

Your "messy" one shows a bandlimited impulse that's been modified only by a bit of time delay. It's symmetric, so there is no relative phase shift, and it looks like a classical sync function, so it's close to a rectangular frequency response shape.

I have no idea why you'd call that "messy", it's about as good as is mathematically possible.

 

[Jiraya369]

Your "messy" one shows a bandlimited impulse that's been modified only by a bit of time delay. It's symmetric, so there is no relative phase shift, and it looks like a classical sync function, so it's close to a rectangular frequency response shape.

I have no idea why you'd call that "messy", it's about as good as is mathematically possible.

like, i have no idea what clean is, the ups and downs on the response looked messy to me, that's all, again, i probably need to research this to have a proper idea.

 

[markstein]

I think the impulse or step response is the least thing to review when evaluating the sound quality of a loudspeaker. Obviously many people do not even understand the theory behind it. I for myself focus on other - more relevant - performance metrics.

 

[JaapD]

A loudspeaker cone has mass and as a result we'll need to deal with mass inertia. This directly relates to i.e. settling time which easily can be an in-band property*.

Also, impulse response is not directly 1 to 1 related to bandwidth (yes, I have knowledge about Fourier transformations). Electronic engineers among us know this when comparing specs of Opamps: slew rate vs. settling time vs. bandwidth. With loudspeakers, especially woofers, we need to think of a dynamic system, containing Proportional , Integral, and Differential properties.

* simple test: apply i.e. 10 high power sinewaves of 250 Hz to a large woofer. After the last sinewave ends one might expect the woofer to immediately end as well, but you'll get an over / undershoot.

 

[Keith_W]

i checked impulse/step response and group delay and i must say i dont understand anything about em. I look at some and go "huh looks like shit" and amir and erin go "CLEAN!" or "no issues here". Like bruh im just so lost

See the section on Impulse response, Section 4.4, page 46 of the REW eBook. Black = all drivers summed. On the right, individual components.

And that's the step response, with the same colour scheme.

I'm glad I wrote that. It has saved me so much time answering the same questions over and over