step response is a important part to show speed of speaker that is good enough for ITD. See measures

[thewas]

On your impulse response can see at 2.1 ms around the early reflections with at least 30%. so all after 2.1 ms have much more errors. on my measures can see at 0.8 ms only 10% reflections on kali. this is 3 times less noise of room and is alot because it reduce also the room influence when hear. on the jbl it is 20% room reflection noise. the kali measures are older

Different rooms and listening distances, desktop vs classic listening setup, that's why, like also many others told you, such measurements at the listening position don't really tell much, even less when made at different positions and rooms.

I test only 2 way speakers. maybe you can show 2 way speaker results ?. that the step response show usefull on 3 way speakers i have never tell and i write earlier that i mean 2 way speakers. dont forget that i only have test and measure 2 way speakers.

here are my impulse responses. On kali can see that the first peak up reach only 70%. the second reach -100%.

On the JBL and your IN8 can see the first peek reach 100% but the second a little less. Or maybe thats a measure argument that the kali sound so worse.

Your topic is about the importance of step response and now suddenly you write about the impulse response? I have measured several 2-way loudspeaker but not the 2-way Kalis, you can ask though Amir or Erin who have measured them and didn't post them as they have very limited information value as they depend also on the frequency amplitude response. Just a hint, a better view for time/phase related issues is the group delay.

 

[bennybbbx]

Different rooms and listening distances, desktop vs classic listening setup, that's why, like also many others told you, such measurements at the listening position don't really tell much, even less when made at different positions and rooms.

the uploaded audio files from MTM, eris 3.5 and kali lp6 the speaker and microphone stand on same place. the diffrence is not so much that it can explain this few stereo width.

Your topic is about the importance of step response and now suddenly you write about the impulse response?

I ask you if you know a reason wy the kali reach with thefirst peak not 100%. In the audio lexikon is tell that a step response is more easy to read.
in the impulse response can not so easy see which have the fastest attack time and filters with fewest phase shift

I have measured several 2-way loudspeaker but not the 2-way Kalis, you can ask though Amir or Erin who have measured them and didn't post them as they have very limited information value as they depend also on the frequency amplitude response. Just a hint, a better view for time/phase related issues is the group delay.

group delay of speakers is here not show. the directivity of speaker i think not important because the room size wide is mostly hear with frequency below 460 hz. have your hear the example that when make 460 hz mono it get very small ?. or tell what you hear on this example https://www.audiosciencereview.com/...for-itd-see-measures.28585/page-2#post-996188

you can also post other 2 way speakers. need not the kali

EDIT: the group delay and the phase is very sensible due to room reflections. so i need window it. in GD can see that the kali change alot in group delay from 800 hz to 1.4 khz. this is bad for ITD. the jbl have a larger groupdelay but it change not so much

 

[markus]

are you a book seller ?.

Damn, is it that obvious?

Looks like the horse isn't really thirsty. By for now bennybbbx

 

[kyle_neuron]

I ask you if you know a reason wy the kali reach with thefirst peak not 100%. In the audio lexikon is tell that a step response is more easy to read.
in the impulse response can not so easy see which have the fastest attack time and filters with fewest phase shift

I don't know why I'm bothering because this has been said by so many others already....
This is because the initial peak of an impulse response is dominated by the high-frequency content. A speaker with more high-frequency content will appear to have a larger or narrower initial peak. This translates to the step response, too. Unless the frequency response and level of the speakers being compared is matched - including the room influences - then the data is not comparable in a scientific manner.

the directivity of speaker i think not important because the room size wide is mostly hear with frequency below 460 hz.

You are wrong.

 

[bennybbbx]

I don't know why I'm bothering because this has been said by so many others already....
This is because the initial peak of an impulse response is dominated by the high-frequency content. A speaker with more high-frequency content will appear to have a larger or narrower initial peak. This translates to the step response, too. Unless the frequency response and level of the speakers being compared is matched - including the room influences - then the data is not comparable in a scientific manner.

So you see that step response show usefull ?. or did you think this cann in impulse response better see ?. the rise time is slower when have a steep LP filter or low crossover Freeq and together speaker FR and LP filter FR result in the speed of the raise time. the speaker is in pysical a low pass filter too. slow bass/mid sspeakers have edge frequency arond 5 khz. and above they drop fast

You are wrong.

Depend on how far away the speakers are. What hear distance to speakers you have ?. I have the kali on desktop 70 cm away and the JBL now 50 cm away. they are near field monitors.
Have you hear the video audio example with the below 460 hz make mono before you judge ?. What do you hear ? if you hear much stereo width lost then room influence is less and also influence of directivity of tweeter or highr freq as 460 hz.


about room influence in this example the speaker at 1 meter distance have same volume as with 0.5 meter. so the microphone measure 6 db less level at 0.5 m . to have same level need increase speaker level 6 db. then the room noise is even more 6 db louder as with 0.5 m hear distance

0.5 meter distance

1 meter distance

 

[kyle_neuron]

Yet again you aren't reading what other people are writing, let alone considering it. Show me the step response of two speakers with normalised magnitude response.

 

[bennybbbx]

Yet again you aren't reading what other people are writing, let alone considering it. Show me the step response of two speakers with normalised magnitude response.

other people write theory and do not confirm how many % the step response change when increase Bass by 15 db. i do tests also with phase linear filters and notice change of 15 db in FR cause only 12 % of longer step response. if you speaker have diffrent crossover and diffrent db/filter is good when this can see because this slow the whole system down.

About normalised magnitude for step response i think about too. this should add in the measure software. Or that in measure software can enable a HP filter of 100 hz. so bass influence is gone. or add a 100 hz HP filter before. maybe there is step response measure software out that do normalize it. have you ever see a step response from the Klippel system ?

Maybe Klippel create better step responses and only Arta and REW are not good and do not normalize frequency. i try in REW measure only from 100 hz but this do not make the step response shorter

see here for step responses of small and big speaker

they do many step responses and they look very fast on speakers with much bass too.


https://www.stereo.de/hifi-test/produkt/kef-r3-1577 42 hz -3db

https://www.stereo.de/hifi-test/produkt/kef-r11-1733 they have 36 hz - 3db

 

[bennybbbx]

Show me the step response of two speakers with normalised magnitude response.

ok, here it is. JBL 104 Eq so it have 39 hz - 3 db. kali have -12 db around(can not see because measure begin with 40 hz. on this freq and Kali is still much slower. should i put my kali on stative and measure it too with EQ that it reach 39 hz -3 db ?. I think this is not need, because then the
Kali step response get even more slower. I measure this on JBL around 10 cm away from speaker. when i go more away from speaker then room nodes reduce the bass but step response is then a little smaller

 

[kyle_neuron]

other people write theory and do not confirm how many % the step response change when increase Bass by 15 db. i do tests also with phase linear filters and notice change of 15 db in FR cause only 12 % of longer step response. if you speaker have diffrent crossover and diffrent db/filter is good when this can see because this slow the whole system down.

Other people are providing the results of scientific studies over decades that explain the cause of the phenomena you’re attempting to observe.

You are posting measurements that include so many uncontrolled variables that there’s no chance to determine anything explicitly.

Unless the responses are normalised, the step response will simply represent the differing frequency responses, same as the impulse response.

 

[bennybbbx]

You are posting measurements that include so many uncontrolled variables that there’s no chance to determine anything explicitly.

then tell which variables this are. room reflection can not see in impulse response when measure 10 cm away. 10 cm measure is good enough because we want only know speed of woofer/mid
I can also the kali measure in nearfield and no influence of room. see the 1.5 cm measure of kali. it is same slow but can not good compare because the tweeter is not measure.

There can not think in all or nothing category. this extrem large bassboost increase the step response from the JBL 104 with 100 µsec to 600 µsec(20% level the woofer start to end when woofer reach 20%) . When you see kali step response in compare is 900 µsec. thats 50% slower. also the 100 µsec raise time to treach peak is more. and for sure when i EQ the kali to have 39 hz - 3db then the step response of kali get even more slower. or do you not believe this ?

the JBL 104 have 1725 crossover. it is 15% higher freq as the kali. if the slower step response come from slow driver, or low crossover frequency or the very steep filter doesnt matter, the result count because i can not change it.

maybe you not want that step response is show, then maybe you find it better when grup delay is show ?. I think with group delay can see a bad speaker for ITD too. the delay from 150 hz to 1200 is not so important. important is only that the time in this frequency range change not much, because left and right speaker play often diffrent wavelength on left and right speaker.

you can see on the genelec page the direct sound and reflect sound depend on distance. at 0.5 m always direct sound dominate much. when have the speakers more than 2 meter away or 9.8 feet then the reverberate sound dominate. see in table

https://www.genelec.com/correct-monitors Direct Sound Dominance​

The balance between direct and reverberant sound has a profound influence on how your mixes will sound. The table shown will help you identify the optimum range of listening distances for the Genelec range.

 

[ctrl]

There can not think in all or nothing category. this extrem large bassboost increase the step response from the JBL 104 with 100 µsec to 600 µsec(20% level the woofer start to end when woofer reach 20%) . When you see kali step response in compare is 900 µsec. thats 50% slower. also the 100 µsec raise time to treach peak is more. and for sure when i EQ the kali to have 39 hz - 3db then the step response of kali get even more slower. or do you not believe this ?

When you match the frequency responses of the speakers so that they are congruent, do you realize that you must not use phase linear EQ, only simple PEQ?
Otherwise the speakers are not comparable.

 

[bennybbbx]

When you match the frequency responses of the speakers so that they are congruent, do you realize that you must not use phase linear EQ, only simple PEQ?
Otherwise the speakers are not comparable.

my test are done with PEQ in the t.racks dsp. when use phase linear EQ then the 80 100 80 peak get smaller and not so round as in my post. see my previous post about diffrence in step response with lineae phase filter or mininal phase filter.

here can see more in detail the FR with and without EQ

 

[bennybbbx]

Just a hint, a better view for time/phase related issues is the group delay.

ok, i remove the JBL 104 and put the kali again on the stative and try to measure that there come a correct group delay. it work 8 cm away from speaker ok i notice and tweeter can see too. but kali have really heavy group delay i do with and without bassboost and i try same on other speaker on other position. My measurement look ok. i get large valley on 650 hz and large peak at 750 hz on both speakers and diffrent position.

the jbl 104 is much better and the highlightet line in this screenshots. I have also a 10 inch guitar amp with box without tweeter. It is measure on the FX return that go to the Tube power Amp of this box. this have in the range from 650 hz to 750 hz much less diffrences that Kali. i think diffrences are important because left and right speaker play often diffrent wavelengths on stereo

so group delay show too. midrange is very slow and unprecice and bad for ITD. maybe there is better as JBL 104 that is not so expensive. maybe there can see group delay of more speakers somewhere in good resolution ?

on this is the guitar box speaker

 

[thewas]

ok, i remove the JBL 104 and put the kali again on the stative and try to measure that there come a correct group delay. it work 8 cm away from speaker ok i notice and tweeter can see too. but kali have really heavy group delay i do with and without bassboost and i try same on other speaker on other position. My measurement look ok. i get large valley on 650 hz and large peak at 750 hz on both speakers and diffrent position.

At which height do you place the mic? Also you could you upload the mdat file?

 

[bennybbbx]

At which height do you place the mic? Also you could you upload the mdat file?

It was 55 cm from desktop plate distance . I use the mems microfone from arc 2.5 this is less sensible to surounding noise becauee it is not unidirectional. I use with calibration file. if you want calc reflections remember the sound go always 2 ways. to desktop plate then from desktop plate. so direct noise go after 8 cm to microphone. the refelections after 110 cm

But i have also large stative 1,2 meter high. i can put them on this so near to a wall is 1 meter if this help. but you see on my speakers they have much better group delay because the speaker are faster in step response

 

[thewas]

At what height relative to the tweeter and woofer? Desktop placement is the worst for measuring the anechoic behaviour of loudspeakers, if you want that you must use narrow and high stands in the middle of the room to have a the first reflection as long far as possible, see here https://www.audiosciencereview.com/...ents-spinoramas-with-rew-and-vituixcad.21860/

 

[bennybbbx]

At what height relative to the tweeter and woofer? Desktop placement is the worst for measuring the anechoic behaviour of loudspeakers, if you want that you must use narrow and high stands in the middle of the room to have a the first reflection as long far as possible, see here https://www.audiosciencereview.com/...ents-spinoramas-with-rew-and-vituixcad.21860/

on the kali it was at height from the light. so it is in the middle of tweeter and woofer. that there are no problems of the room you can see because the FR of kali look good from tweeter too with the 8 cm measure. and you see the tweeter group delay after 1.5 khz do much less Jitter as the mid range. the kali have a wavguide and the waveguide direction fit so microphone is inside waveguide range when use 8 cm. the 1.5 cm measure i have done in middle of woofer. you see lots high loss in FR

 

[thewas]

on the kali it was at height from the light. so it is in the middle of tweeter and woofer. that there are no problems of the room you can see because the FR of kali look good from tweeter too with the 8 cm measure.

There always reflections, if you have them directly on the desk they are just very short so mix up with the direct sound impulse.

 

[gnarly]

So you see that step response show usefull ?. or did you think this cann in impulse response better see ?.

Well, my 2c.........i have to say i don't think step responses are all that useful.

Sure, they are a plot that shows both magnitude and phase (frequency response) in a single graph, which gives them some unique value similar to a Nyquist plot.
But they are difficult to read because both mag and phase have been lumped together.

And step response contains zero information that is not already contained in everyday mag and phase Bode plots, where i can easily see what's going on individually with mag and phase.

Try making tuning decisions, EQs, xovers, time alignments,... based on step response ......good luck.
It's back to good old transfer function Bode plots for real world work, huh?
(Or impulse response if looking for horn reflections, multiple time arrivals, and such.)

But step response ????, ....imho, it's mostly about illustrating some supposedly deeper understandings about mag and phase....
IOW, mostly marketing and techno babble.

 

[bennybbbx]

Well, my 2c.........i have to say i don't think step responses are all that useful.
Try making tuning decisions, EQs, xovers, time alignments,... based on step response ......good luck.
It's back to good old transfer function Bode plots for real world work, huh?
(Or impulse response if looking for horn reflections, multiple time arrivals, and such.)

and with which do you time alignement of tweeter and woofer ?. I use for this the step response. I know no other solution. I have buy a ribbon tweeter that i can put on my other speakers so i can play with crossover and delay between tweeter and woofer. and i notice that steep crossover sound worse and large delay between woofer and tweeter too. I think the JBL can even more sound better when there is shorter delay between tweeter and woofer. also there are many Hifi magazine that show step responses. group delay is better for speaker speed to see, but the time between woofer and tweeter can not see. look my group delay screeenshots