Importance of impulse response

[Multicore]

I have seen a lot of people here say that impulse response is not relevant, and that basically only frequency response matters.

I thought they were two different representations of the same information, and they are mathematically tautological w.r.t. each other.

 

[fpitas]

I thought they were two different representations of the same information, and they are mathematically tautological w.r.t. each other.

Somebody here may have once said that, maybe? But yes, they are illustrating the same thing.

 

[DonH56]

I thought they were two different representations of the same information, and they are mathematically tautological w.r.t. each other.

Problems occur because too often the information is incomplete, at least in presentation... Frequency magnitude response plots without phase information and/or insufficient bandwidth (at either end) will not allow accurate conversion to the time domain, and time domain response must cover a sufficient time period with enough resolution to be converted to an accurate frequency response. Devil's in the details...

 

[Multicore]

Also I always thought that it makes sense that a smaller driver would be faster and thus sound faster, due to it being lighter.

Well, maybe, all else being equal a lighter thing can be accelerated with less force and maybe therefore be more efficient. But if to make it lighter you have to make it smaller then it will move less air, so the efficiency is down again.

 

[fpitas]

Well, maybe, all else being equal a lighter thing can be accelerated with less force and maybe therefore be more efficient. But if to make it lighter you have to make it smaller then it will move less air, so the efficiency is down again.

And decent big drivers have enormous, powerful motors. Beware the internet myths.

 

[levimax]

I thought they were two different representations of the same information, and they are mathematically tautological w.r.t. each other.

You need both FR and phase to derive the impulse, FR alone is not enough to derive impulse.

 

[Multicore]

Somebody here may have once said that, maybe?

Probably. I haven't read everything on ASR.

But yes, they are illustrating the same thing.

Well, illustrate suggests the graphical presentation in a to a human viewer and I think only someone unusually skilled in reading those images will be able to the same information in both.

For example, I can't do it. I don't know what to make of impulse or step response graphs (illustrations). And I suspect this may relate to why some other people prefer to talk about them.

 

[fpitas]

Probably. I haven't read everything on ASR.


Well, illustrate suggests the graphical presentation in a to a human viewer and I think only someone unusually skilled in reading those images will be able to the same information in both.

For example, I can't do it. I don't know what to make of impulse or step response graphs (illustrations). And I suspect this may relate to why some other people prefer to talk about them.

Exactly, lots of posts here and who knows what someone said.

And yes, impulse response is the tough way to look at the problem. That's why we usually use FR graphs.

 

[Multicore]

You need both FR and phase to derive the impulse, FR alone is not enough to derive impulse.

Yes. Or the real and imaginary will also do. But if you throw away half the data set you can't get it back.

 

[fpitas]

I suppose if a speaker seriously deviates from minimum phase, that has to be taken into account.

 

[Multicore]

Devil's in the details...

It usually is. Or, answers to most genuinely interesting questions should probably start with, "It depends..."

 

[Multicore]

I suppose if a speaker seriously deviates from minimum phase, that has to be taken into account.

Amir did a video about phase recently and it made sense to me as far as it went but I still have some questions about its significance and how to interpret it in measurements. For example, in multi-microphone recording, phase is a big deal up to a point, and then it isn't. Or something like that. Good recording engineers appear to know what they are doing somehow know to avoid the danger zone. I'd like to understand that better. Fair warning: I was reprimanded for being off topic once already today.

 

[fineMen]

This is slightly technical audio forum so I suggest to use few moments to study what step response and impulse response envelope ... It's not our problem and responsibility to filter or modify data in posts if you don't understand it.

Sorry, I'm a tech/math scientist by education. I'm entirely not o/k with Your presentation. Anyway, I already left this mindless thread. 'Won't save the world. Thank You.

 

[thanossapiens]

"Fast" is meaningless. Any transducer that can reproduce any particular frequency is by definition doing it at that frequency (speed=frequency). A speaker generating a 1kHz tone is vibrating back and forth at exactly 1 kHz. There is no such thing as a slow or fast 1 kHz tone.

I meant more in the sense of "rise time" or simply how long it takes to start vibrating at 1Khz.

 

[ChargedCapacitor]

I meant more in the sense of "rise time" or simply how long it takes to start vibrating at 1Khz.

That's not how it works, bud.

 

[thanossapiens]

That's not how it works, bud.

So companies that say stuff like this are basically lying?

 

[dfuller]

So companies that say stuff like this are basically lying?

I mean, that certainly looks like total nonsense. Wherever did you find that?

"fall delay" happens because nothing can start or stop moving instantaneously. "rise delay" looks pretty nonsensical. I guess the closest thing is group delay but even then...?

"Fast" is meaningless. Any transducer that can reproduce any particular frequency is by definition doing it at that frequency (speed=frequency). A speaker generating a 1kHz tone is vibrating back and forth at exactly 1 kHz. There is no such thing as a slow or fast 1 kHz tone.

The point that I'm interested to hear explained better is not speed, but how transient response and resonance after the transient affect the perception of "fast impulse response". I can imagine the magnet in a dynamic driver responding at a given frequency to a transient signal, but the propagation of the impulse throughout the speaker cone occurring over time. This is distortion of the fundamental into the time domain, either as a smearing of the fundamental or as some form of resonance. The addition of distortion is, indeed, expressed in the frequency response, but is still distortion, and could contribute to a perception of one speaker (for example, with a stiff, light, beryllium cone) as "fast" or "resolving", whereas another speaker (for example with a soft, paper cone and a stiff surround) as "slow", or "soft". Both speakers are generating, say, the same 1kHz transient at the same 1kHz frequency, but sound very different. One has a spiky impulse graph with immediate decay, and the other's is smeared out over time with a sloped attack, distortion and/or a trailing resonance/decay.

Why wouldn't graphing or quantifying these aspects of transient response be useful? Why wouldn't we want to do so and correlate the results, if possible, with either subjective or objective measures of fidelity? Do we (or can we) get a useful picture of this aspect through any other of our routine measurements? Surely THD percentage is too blunt a measurement to provide a useful analysis of this? Why aren't impulse response graphs or square waves useful in doing this?

Yes - that's why the goal is infinitely stiff, zero mass, and completely acoustically inert.

In the real world, all things have compromises. Beryllium is toxic to mine and refine, and insanely expensive despite its exceptional behavior, and the cheaper alternatives like aluminum are both heavier and have worse breakup mode behavior.

But really, most of what makes for "fast" or "slow" bass is the cabinet behavior. How the port is tuned (if present), how well-braced and nonresonant the cabinet is, the internal volume, how much absorbent material there is, etc etc.

For example: ATC intentionally overdamps their ports - they trade off maximal LF extension for shorter decay time as they preferred that outcome in that compromise. Others like Genelec prefer maximal LF extension over shorter decay time. They're both entirely valid approaches, it ultimately boils down to which works better for your tastes.

 

[kemmler3D]

So here's something I've been thinking about and I can't quite be sure of how it translates into a physical system.

In order to reproduce a perfectly sharp transient (say a click from a snare drum or cymbal strike) you need all frequencies at once, from DC on up. This is what's meant by "an" impulse response or Dirac function.

Now, in real life, is that still how it works with a multi-way system? If you put a click through the speaker, what is the contribution of the sub at 20hz? How much does it move?

Intuitively, the click won't sound right if you don't have full range reproduction, but intuitively, it also feels like the sub won't have time to start and stop moving... the impulse doesn't really make intuitive physical sense at all, but at the same time, we really do have very sharp transients that resemble an impulse in music sometimes.

Does any of this make sense?

 

[ChargedCapacitor]

So here's something I've been thinking about and I can't quite be sure of how it translates into a physical system.

In order to reproduce a perfectly sharp transient (say a click from a snare drum or cymbal strike) you need all frequencies at once, from DC on up. This is what's meant by "an" impulse response or Dirac function.

Now, in real life, is that still how it works with a multi-way system? If you put a click through the speaker, what is the contribution of the sub at 20hz? How much does it move?

Intuitively, the click won't sound right if you don't have full range reproduction, but intuitively, it also feels like the sub won't have time to start and stop moving... the impulse doesn't really make intuitive physical sense at all, but at the same time, we really do have very sharp transients that resemble an impulse in music sometimes.

Does any of this make sense?

Studies have shown that listeners don't have a preference between speakers that have time aligned driver impulses, and speakers without time aligned driver impulses. Does this answer your question? As long as the frequency response is even and follows the listeners preferred curve, how the impulse response looks is mostly unimportant.

 

[Trdat]

I’d like to see people post more complete IR data or even just a single sweep REW mdat file rather than just a picture of the FR magnitude graph most of the time. Yeah, context of the measurement be would appreciated as well.

Apologize for hijacking the the thread but it seemed appropriate to post here rather than start a new thread.

Finally after 4 years I finally realised how I can zoom into the impulse response and see the first 5 or so seconds and I can actually see the shape of the response.

I have attached the response of my Audiolense DSP tri-amp system with its crossovers done through the DSP. Compression driver and 15 inch mid and dual subs in a very small but heavily acoustically treated room. And no timing reference or loopback in measuring.

It looks good to me but I am struggling to fully understand if is properly time aligned, as on paper with Audiolense it should be. If anyone would like chime in...