• 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!

Importance of impulse response

fineMen

Major Contributor
Joined
Oct 31, 2021
Messages
1,504
Likes
679
You assume I am supposed to be unaware of these things — because I have not mentioned it.

And now the accusation of “misusing scientific terms in hyperbole”… eh, ok.
...
I’m hoping you’ve got an interesting answer for us.
I was hoping the case was settled as far as I'm concerned. I'm actually and positively not accusing You personally of anything. You don't refer to my argument I stated in reference to Your explications. Is that an interesting answer?
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
I'm actually and positively not accusing You personally of anything.

Well, I appreciate the clarification.

I was hoping the case was settled as far as I'm concerned.

You raised issues... in this case, those who take time to read and analyze measurements, perhaps, even working to improve them… information which can be gleaned from an impulse response.

I do not remember a resolution you've provided.
 
Last edited:
  • Like
Reactions: MAB

kongwee

Major Contributor
Joined
Jan 22, 2022
Messages
1,024
Likes
276
Shorter impulse response is better for me as I look at stereophile for measurement. They just sound better as I audit them. Start and stop the sound fast. More details and transparent. High transient music like a orchestra staccato hit. You can listen how fast it can go silent. You can listen hear how fast is transitioning from loud to silent. This is how I use this graph. Know ASR don't feel it is important is flat FR is all you need.
 

Killingbeans

Major Contributor
Joined
Oct 23, 2018
Messages
4,089
Likes
7,552
Location
Bjerringbro, Denmark.
Start and stop the sound fast. More details and transparent. High transient music like a orchestra staccato hit. You can listen how fast it can go silent. You can listen hear how fast is transitioning from loud to silent.

IMO, that's the same kind of disconnect between physics and layman intuition you also get when people claim to hear "stairsteps" in the outputs of DACs.
 

kemmler3D

Major Contributor
Forum Donor
Joined
Aug 25, 2022
Messages
3,022
Likes
5,644
Location
San Francisco
IMO, that's the same kind of disconnect between physics and layman intuition you also get when people claim to hear "stairsteps" in the outputs of DACs.
I dunno, I think the difference between a (say) 2ms and 15ms IR is pretty audible, although it will mostly be heard as FR and not decay. That said if the IR is like 100ms then yeah, you can hear that too if resonances get really bad... maybe... sometimes you can hear how long the IR is.
 
Last edited:

kongwee

Major Contributor
Joined
Jan 22, 2022
Messages
1,024
Likes
276
IMO, that's the same kind of disconnect between physics and layman intuition you also get when people claim to hear "stairsteps" in the outputs of DACs.
There are some physical you can hear. In fact more damp of the room acoustic, it is easier to hear the loud to silent transition between brand of loudspeakers. However, not everyone like fast damping impulse respond, just like flat FR.
 

fineMen

Major Contributor
Joined
Oct 31, 2021
Messages
1,504
Likes
679
..., just like flat FR.
Not quite. I think regarding the 'pulse' it became pretty much clear, it's a personal thing. Some say it's a belief, some say it was imagination arcing from fancy pictures in magazines to a personal listening experience, but nobody still claims that there were explanatory models based on science (except psychology, maybe).

Regarding 'flat', it depends. If it was an original recording from the farther past, one would prefer a speaker that resembles the original monitor back in the day. If it was a remastered version, one might prefer a more contemporary speaker. In the first case the 'flat' is ruled out even for the 'flat FR' guy. The speaker has always to match the recording, not a standard. Having a standard means, one same speaker type for all recordings, that's all so convenient!

Same, basically, with the 'pulse'. But its hard to grasp. Because the industry is lying at You deliberately, as to keep You a) unsatisfied and b) confused. Are You?
 
Last edited:

kongwee

Major Contributor
Joined
Jan 22, 2022
Messages
1,024
Likes
276
Same, basically, with the 'pulse'. But its hard to grasp. Because the industry is lying at You deliberately, as to keep You a) unsatisfied and b) confused. Are You?
You have the option to go out and listen more. In show or retail or money back online. You can listen to them and how do you related to the graph or reviews. I alway recommend to go out and listen more, especially you have a setup. I mean in show you can listen to $100k speaker. You don't need to own them too.
 

Killingbeans

Major Contributor
Joined
Oct 23, 2018
Messages
4,089
Likes
7,552
Location
Bjerringbro, Denmark.
I dunno, I think the difference between a (say) 2ms and 15ms IR is pretty audible, although it will mostly be heard as FR and not decay. That said if the IR is like 100ms then yeah, you can hear that too if resonances get really bad... maybe... sometimes you can hear how long the IR is.

Yes, but a textbook perfect impulse response would imply a frequency + phase response with a bandwidth way beyond what humans can hear?

I can't help thinking that when people see something approximating a neat spike, they imagine all kind of things about "speed" and "transients", just because of that graphical suggestion. I've no doubt that impulse responses can be useful, but you'll have to unwrap the information it gives you. Not just go looking for correlations based on possibly false intuitions.
 

kemmler3D

Major Contributor
Forum Donor
Joined
Aug 25, 2022
Messages
3,022
Likes
5,644
Location
San Francisco
a textbook perfect impulse response would imply a frequency + phase response with a bandwidth way beyond what humans can hear?
It implies infinite bandwidth with perfect phase response.
when people see something approximating a neat spike, they imagine all kind of things about "speed" and "transients"
I mean, it does basically approximate anything you'd want to call "transient response". And, although you rarely see it, if there were somehow an IR that totally lacked anything like a spike shape, you'd definitely (pretty much by definition) not be able to hear any crisp transients in the output. I know people hate the "slow" and "fast" terms, but I'd be comfortable calling a speaker with a badly smeared IR "slow"... if only because I can't think of another speaker metric for which a speed judgment would even make sense.

What's funny is that in the production / DAW world, impulse responses are typically used as special effects or to simulate reverb. They are also used to capture and simulate the FR of guitar speaker cabinets, very frequently. After a while you develop a sense that you can tell what an IR sounds like by looking at it, although that may be an illusion... for the really extreme cases you definitely can. I would like to point out before anyone gets upset that measured Hi-Fi speaker IRs are never extreme cases.
 

Killingbeans

Major Contributor
Joined
Oct 23, 2018
Messages
4,089
Likes
7,552
Location
Bjerringbro, Denmark.
And, although you rarely see it, if there were somehow an IR that totally lacked anything like a spike shape, you'd definitely (pretty much by definition) not be able to hear any crisp transients in the output.

Absolutely. But if you measured the frequency + phase response instead, it would also be wonky as hell?

That's what I'm trying to make clear to @kongwee

You can't have a nice frequency + phase response and a lousy IR from the same measuring point, or vice versa. They are two sides of the same coin.

He/she hasn't found a source of "hidden information" in the IR. The poor souls who look at frequency + phase has the same data in their hands.
 

kimmosto

Active Member
Joined
Sep 12, 2019
Messages
213
Likes
498
I'd be comfortable calling a speaker with a badly smeared IR "slow"... if only because I can't think of another speaker metric for which a speed judgment would even make sense.
Loudspeaker with bad timing (long group delay at upper bass..low mid, bad step and impulse response and ETC) sounds primarily weak, relatively thin/bright and sound stage could be scattered. "Slow" and "long decay" are quite bad terms.
 

kemmler3D

Major Contributor
Forum Donor
Joined
Aug 25, 2022
Messages
3,022
Likes
5,644
Location
San Francisco
You can't have a nice frequency + phase response and a lousy IR from the same measuring point, or vice versa. They are two sides of the same coin.
Yes, certainly. They are both derived from the same measurement and you can derive one from the other.

However, I actually think an IR plot is easier to read in the case of big phase or GD variations. Multiple peaks or wide peaks are easier to spot (for me) than a phase plot that wraps 10 times or whatever.
 
Last edited:

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Loudspeaker with bad timing (long group delay at upper bass..low mid, bad step and impulse response and ETC) sounds primarily weak, relatively thin/bright and sound stage could be scattered. "Slow" and "long decay" are quite bad terms.

Since you mention “sound stage”… does this mean you listen in stereo as well to evaluate speaker performance? I ask because there are those who only do mono listening test evaluations for a variety of reasons.
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Since you mention “sound stage”… does this mean you listen in stereo as well to evaluate speaker performance? I ask because there are those who only do mono listening test evaluations for a variety of reasons.
I actually audition headphones one ear at a time. Superior way of judging SQ.
 

gnarly

Addicted to Fun and Learning
Joined
Jun 15, 2021
Messages
992
Likes
1,390
Yes, but a textbook perfect impulse response would imply a frequency + phase response with a bandwidth way beyond what humans can hear?
I'd say an academically theoretical perfect impulse response would imply a frequency + phase response with a bandwidth way beyond what humans can hear.
But that's a big so what imo....
Academically, theoretically, the lake rises when we pee in it.........

What matters is the shape of the impulse within human hearing...20Hz-20kHz more than suffices....
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
I actually audition headphones one ear at a time. Superior way of judging SQ.

You may be right for the most part — and it can be faster or often be convenient — but I reckon there are potentially some things you might not notice or miss.
 

kemmler3D

Major Contributor
Forum Donor
Joined
Aug 25, 2022
Messages
3,022
Likes
5,644
Location
San Francisco
What matters is the shape of the impulse within human hearing...20Hz-20kHz more than suffices....
Yes, for some reason I couldn't quickly google up a picture of a band-limited impulse (say 0-20khz) but yeah, it wouldn't look like a vertical line. Just take a single dirac-type sample at 96khz and lowpass it at 20khz or so... should be close enough for government work. Might be good to keep that handy as a reference. Someone could probably generate one, I need to get back to work...
 

NTK

Major Contributor
Forum Donor
Joined
Aug 11, 2019
Messages
2,661
Likes
5,821
Location
US East
Yes, for some reason I couldn't quickly google up a picture of a band-limited impulse (say 0-20khz) but yeah, it wouldn't look like a vertical line. Just take a single dirac-type sample at 96khz and lowpass it at 20khz or so... should be close enough for government work. Might be good to keep that handy as a reference. Someone could probably generate one, I need to get back to work...
OK. I bite. For numerical computation reasons, frequency scaled by 1000, i.e. 0.02 -> 20 Hz and 20 -> 20 kHz. Time scale is in milliseconds.
System: 2nd order HP, Q = 0.7, at 20 Hz and 2nd order LP, Q = 0.7, at 20 kHz.

imp_resp.png
 

kongwee

Major Contributor
Joined
Jan 22, 2022
Messages
1,024
Likes
276
That's what I'm trying to make clear to @kongwee
I don't understand why mixed flat FR and impulse respond. I alway related impulse respond to speed. Fast start and stop. Of course, you can mixed up with FR. Like a diamond tweeter that excited up to 40kHz. People will only interested in 2000hz and 15khz. You have a 2000Hz hit sound clip respond and look at how fast it is died down compare to soft dome. Of course, if you don't damp the internal cabinet, diamond tweeter will continue to ring as hell. Because, diamond tweeter is super light weight it can excited up to 40kHz. Same as beryllium tweeter. But some ASR will scream at you can't hear 40kHz. However, people don't scream at ribbon tweeter that can hit 40kHz too. Finally, you have super tweeter at 100kHz, which don't make sense if you do not know how to use it.
 
Top Bottom