• 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!

Step response question

@bennybbbx
Here is the Alpair 10p at 1kHz - it seems to be a bit slower (comparatively speaking)
1615529065627.png


I am not aware of any commercially available full range speaker at that price point - I don't think it exists
The cheapest I saw is from Heco (you can google it) - but I would rather go for DIY
btw I have a pair of 10" Audio Nirvana Classic ferrite for sale that I can ship to you to Germany in case you are interested :)
 
the alpair 10p look not so good. seem have resonances. but 10 inch is too large for me. I look for a 6.5 inch replacement for the kali. it is then also intresting what sound better. the 6.5 inch in fullrange or the fullrange upto 1.5 khz(kali crossover) with the tweeter. I think most important for stereo is speed in the range 200 - 1.5 khz. so i can compare. its also usefull to record and do blind tests to compare. I have in my speaker quicktests a testsong piece record first left channel, then right channel. and because i use DAW and same speaker same microphone it is sample exact.

maybe waterfall on REW is more usefull to show speed at more frequency. there can in perspective setting set the slice to show. see settings i use. i use the 16. slice at front. important to see speed is set the rise time to 0.1. shorter values seem not possible.
16. slice 01 ms riste time.jpg
 
Last edited:
just came to my mind that I also have a Morel TSCM 634 (also for sale), here are the measurements:

1kHz

1615556776454.png


15kHz

1615556813279.png
 
the morel is very expensive, you have really a good speaker with the 12 inch. the result look not so good as the 12 inch. maybe it is also usefull to see a slice at 2.5 khz.

i thought about 2 systems for 200 eur max. the eris 3.5 cost pair 99 eur btw. I try to remove in the kali the woofer, to measure the screw distance, but its bad the screw wood in it so i can not get the woofer outside and need remove the screws in the wood. but my screwdriver is too short and 1 screw is very hard, can not loose and the other screw is near bass reflex port so i need this very long screwdriver(need longer as 30 cm). I have a ratched but with extension its too too short. see photo of it. also it seem the woodpiece screwed wood is clued. there can see shine in corners. so seem not possible to remove the old. really bad.

eris 3.5 spectro 2.5 khz.jpg
eris 3.5 spectro 1 khz .jpg
lp6 open.JPG

BTW: there is new rew version REW RC6. it normalize now correct and my speakers and headphones are at peak 100% too.
kali spektro 1 khz.jpg
kali spectro 2.5 khz.jpg
 
Last edited:
I find a active wide band speaker behringer c5a auratone. maybe there are precise step response or waterfall somewhere ?.
 
I don't have measurements about the Auratone but I do have measurements (with Dirac) about the passive Avantone Mixcubes; here you go

1kHz:
1615629940021.png


2.5kHz:

1615629991547.png


15kHz:

1615630015900.png
 
at 1 khz they look slower as the Eris, but at 2.5 khz they look faster. and at 15 khz they look a little faster as the eris with the tweater.

I read a compare between the auratone https://www.amazona.de/comparison-review-auratone-5c-avantone-mixcubes-behringer-c50a/4/ so seem the 5c is better. the behringer c50a they test not good and the mixcubes. but the behringer c5a have auratone speakers inside because they have a auratone label in front and the C50a have a Behritone label in front.

https://www.amazona.de/comparison-review-auratone-5c-avantone-mixcubes-behringer-c50a/4/

it seem really hard to find fast speakers. intresting that the 12 inch speaker is faster. maybe with 12 inch the speaker need not move so much and it give advantage. its also possible that in the Alpair 10p at 1kHz measure there happen room reflection. i see simular in my headphone measure

cheap 10 eur headphone panasonic rp th030 give this results. I put headphone on table and hold microphone near pads. so it is possible that microphone get lots reflections from table. several measure give simular results. the 15 khz measure is very simular to speaker. i measure also with 96 khz. make not much diffrence. so its no 44 khz limit in sight. here are headphone results. maybe i should try again and put foam under the headphone and table
rp th030 15 khz.jpg
rp th 030 1 khz.jpg
rp th 030 2.5 khz.jpg
panasonic RP TH030.jpg
maybe i should do measure of headphone again and put foam on the table.

I find that there is a dirac PC software. I give up using such measure software because it sound not good for me, because it correct left and right speakers much diffrent and use small filters and higher level cut . I have a video do that use diffrent smoothing. I think the REW corrections with 1/2 octave smoothing sound best for me. here you can hear the diffrence between much and less smoothed corrections.
i think best sound the version with less eq but wider bands. In the video is minimum phase disable. so its linear phase. but diffrence is not much hearable. I use the rew EQ settings in my t.racks dsp and measure with EQ enable and correct with hand and until it sound best i allow +/- 1.5 db corrections from linear. I think this video show for me wy i not like such correction software. i have ik multidia arc 2.5 and i test mathaudio roomeq and the sonarworks. maybe i diarc 3 have a usable demo(a vst plugin to test the correction) i test it too some day
 
Last edited:
@bennybbbx
I agree that wideband EQ should definitely sound better, however I suggest trying Dirac Live 3 (they have a 14-day free trial, zero risk)
I have also tried all the others mentioned by you but only Dirac can correct the step response
It not only corrects the frequency response but also the impulse response and step response!
You will very clearly see and hear that if you measure your Eris or Kali with REW before/after Dirac - there will be a huge difference
When you use Dirac make sure that the measurement "cube" is at least 1 metre big
Let me know if you tried it, I would be curious to hear your opinion and experience
 
When it can test then i test it. In the free 14 day trail version i can measure i think, but how should i correct my speakers and hear ?. I think there need dirac hardware that it support. or can it output a filter response and i can use a convolution effect ?. on dirac page i read there need dirac live. so i now install it and it say no devices found.

I think the speed of speaker no software correct. but it is possible to enhance speed i think in same way as it is on TFT displays called overdrive. maybe a quick and easy way all audio measure software can show to show fast speaker or slow speaker is zoom the step response so only the rise time is show and measure the speaker 1.5 cm(when have more ways) away. here i have such step responses post. there can clear see which is the faster in rise time.the ilould mtm have a strange rise curve, or maybe it is because the MTM have a grill before speakers. this can not remove

maybe you can show such step response measure too(with or without dirac maybe). Y
rise time rp th 030 headphone.jpg
rise time kali lp6.jpg
rise time iloud mtm.jpg
rise time eris 3.5.jpg
you can use your existing measures because you have 1 way so no tweeter that change the real speed.

EDIT: i also read about FAST (Fullrange Assisted Subwoofer Technology) speakers. this are bass and very small(and hope fast) wideband speaker. systems are 2 way and crossover is around 300 hz. but seem only for DIY
 
Last edited:
Regarding FAST, I personally prefer the sound of larger full range drivers as I mentioned above since they have "larger" sound
You need to try for yourself and see which one you will prefer :)

Regarding Dirac you need to download two components
One is Live: https://artifacts.connect.dirac.com/public/accord/release/win64/diraclive-v3.0.13-setup.exe

And the other is the VST Processor: https://shop.dirac.se/order/checkout.php?PRODS=28606608&QTY=1&CART=1&CARD=1

Then you can use the VST Processor in any DAW (I use Jriver) and just follow the manual for the measurement

Then once you optimized your system with Dirac you can measure what you have done by passing the signal through Dirac in REW - you can do this the most simple way by using Jriver's WDM Driver: it will show in Windows as a device that you can select in REW and then the sound will go through Jriver and all the VST plugins that you configured there (including Dirac)
So this way you can make direct comparison with/without Dirac (or anything else for that matter)

This is a Logitech Z623 2.1 step response without Dirac vs with Dirac, look how much smaller the ringing is with Dirac enabled:

1615670467287.png


This is the speed at 2.5kHz:

no Dirac:

1615670502712.png


with Dirac:

1615670534022.png


The time is the same but the % gets lower by 10% and the second peak is also lower - to me this means increased precision

Tomorrow I will send some comparison for the Audio Nirvanas too
 
So this is where it gets really interesting!
Nubert nuVero 140 step response measurements

With Dirac vs without Dirac:

1615711518454.png


The difference is day and night - not just visually but audibly too
You will see the same level of difference with any 2 or more way speakers

Now let's look at the speed!

1kHz, no Dirac:

1615711671906.png


With Dirac:

1615711701717.png


Again, day and night

Now let's see the same for the 12" Audio Nirvana

No Dirac:

1615711768304.png


With Dirac:

1615711789943.png


As you can see, with a full range the difference is not that huge, but there is a difference - and this is audible too
 
ah great that there is a VST processor. then i can test dirac. my room is not so good, because measure show much diffrence between left and right speaker in bass. in bass ears smooth around 1 or 2 octave. the ERB smooth filter in rew should simulate ears. when i use ERB filter my left and right speaker measures are not so much diffrent as the speaker correction softwares i previously test show.

best zoom step impulse and look at the rise time. the later ringing is maybe reverb or change by EQ. in 1 ms audio walk 38 cm. A speaker measure system notice in rooom at some frequency higher reverb and reduce the level at this frequency. so when measure later the reflections are less and also in response result a little. in rew you can output impulses from the EQ you use in REW EQ tab with menu "Export Filters Impulse Response as wav" You can look how EQ look in step response when you import this wav. The time after the rise time in step impuls then is first come high freq later low freq. but so real clear what it do it is not for me. because it do more slopes than i expect. see this impulse output . normaly asume that the 1 khz boost i do increase visible at 1 ms or 0.5 ms with max level. but as you can see it do not.

this example do 50 hz and 1 khz and 5 khz Q2 + 4 db EQ boost. step response of the filter impulse is this
filter 50 hz 1 k 5 k all 4 db q2_phase freq.jpg
filter 50 hz 1 k 5k all 4 db q2_1.jpg
filter 50 hz 1 k 5k all 4 db q2_2.jpg


and here is a step response that do HP. it look lots diffrent, but rise time is same. so i think when look at step response best ignore the tail and look at rise time. the rise of the impulse is reduce when use higher sample rate. I can use upto 96 khz. but speakers are so slow, i see not much diffrence
Hp 200 risetime.jpg
 
So this is where it gets really interesting!
Nubert nuVero 140 step response measurements

With Dirac vs without Dirac:

View attachment 118150

The difference is day and night - not just visually but audibly too
You will see the same level of difference with any 2 or more way speakers

Now let's look at the speed!

1kHz, no Dirac:

View attachment 118151

With Dirac:

View attachment 118152

Again, day and night

Now let's see the same for the 12" Audio Nirvana

No Dirac:

View attachment 118153

With Dirac:

View attachment 118154

As you can see, with a full range the difference is not that huge, but there is a difference - and this is audible too
this is really amazing the nuvero get lots faster in measure. i try dirac soon on kali. maybe kali sound much better. on the 12 audio nirvana there is only the fall time more constant. this not constant fall time is my guess happen due to room reflections. my headphone measure too have not constant fall time
 
I get dirac to work. this really hard. first it hang on export filter and now studio one 5.2 crash when i remove the dirac plugin. but ok it work enough to test. there happen what happen with all automatic measurements i see. you can try if your correction is ok when you play testtones. need not loud. when you move frequency from 100 hz to higher then hear if the testtone come from middle.the measure i do in dirac i need no testtones, the signals come from more left. but it is my room is worse. here is online testtone generator to verify if signal is when move the slider slowly up always in middle

https://www.szynalski.com/tone-generator/

if you use a DAW you can also use the moscilator plugin. the ikmultimedia arc 2.5 allow same correction for both speakers and mathaudio roomeq allow correction change for left right channel independent.

when some frequency level more left or more right and are not in middle, this do mono to stereo processors too. this create pseudo stereo and room sound bigger. but i try with more measure if i can get dirac measure better in the middle and i measure the kali. what happen to step response. this is only a early report, maybe you have some suggestions to solve this better. maybe there are some settings i not see ?.
dirac right.jpg
dirac left.jpg


EDIT: i have done new measure and it sound in first moment better as without, but it is not correct. i notice at testtone 148 hz that testtone come near full from right speaker. when look at the levelmeter right speaker get a 9 db louder signal. in arc 2.5 there is option combined L/R correction. this is better see screenshot of dirac. i think a room correction software should have a option for diffrent smoothing settings and that the diffrence between l/R correct can limit to max 4 db. I hear the testtone clear from the middle without dirac. and with dirac it come near full from right and it is more diffuse not so clear to locate as without dirac. in a mix this strange pan of room correction is not so much hear, but when record solo intruments this can hear. but verybody should check self if he get problems. i think problem is my room because desktop stand on right side 60 cm near wall. on left side it is 1,5 meter. better center i can not. anyway without automatic room correction and hand optimized REW eq settings i get it good working in my room
dirac at 148 hz.jpg
 
Last edited:
here are step responses of rise time Eris 3.5 and iloud mtm measure 1.5 cm away from woofer with and without dirac. not much change. but really intresting whats happen then with kali if stereo width get larger.
mtm dirac off.jpg
mtm dirac on.jpg
eris 3.5 firac on.jpg
eris dirac off.jpg
 
Hi @bennybbbx
Try to measure in the listening position comparing Dirac vs without Dirac - at least this is how I do it

The issue you mention about certain frequencies coming from odd directions: that is definitely the room - I have the same on my left channel at certain frequencies below 100Hz
It can be corrected in my case using dual subs in Double Bass Array but that generates some other issues on the other hand...

It is very interesting what you have done in REW simulating the effect of EQ on the step response - I tried to do the same but did not succeed
Can you please tell me the steps so I could reproduce it myself too?
Many thanks
 
Hi @bennybbbx
Try to measure in the listening position comparing Dirac vs without Dirac - at least this is how I do it

The issue you mention about certain frequencies coming from odd directions: that is definitely the room - I have the same on my left channel at certain frequencies below 100Hz
It can be corrected in my case using dual subs in Double Bass Array but that generates some other issues on the other hand...

It is very interesting what you have done in REW simulating the effect of EQ on the step response - I tried to do the same but did not succeed
Can you please tell me the steps so I could reproduce it myself too?
Many thanks

this have ernestcarl explain here with screenshots. I did not know before too
https://www.audiosciencereview.com/...oes-it-really-matter.1999/page-12#post-693727

when it is only below 100 hz is much better. on my room this happen upto 300 hz and change stereo image alot.

When switch dirac on and switch dirac off. I have in my dsp all EQ and delay reset so the DSP do nothing. maybe you try your unprocessed measure in REW smooth 1 octave and disable allow narrow band. create the eq setting from Rew. then do the export as in the post of the filtersetting from ernestcarl. the wav you can load in and look or you can this load in an convolution effect or impulse loader to correct your speaker. then you can hear what you more like. wide band less level correct or small band much level correct as the automatic room corrections do. when you do wide band less level correct you get very few phase shift too. so phasse correct help nothing hearable. the example smooth video i do with linear phase. with mimnimum phase it sound even more worse
 
Last edited:
i get in mind what i upload was wrong. it was 1.5 cm with dirac and without. the measure in 50 cm away i dit not see now but it was too very simular. only the tweeter rise and woofer rise was more near. seem there is no jriver need. dirac now have a own virtual device. yesterday it work. today device is not find. i quit the dirac live processor and start again. and get the virtual device

dirac virtual device.jpg
mtm no dirac.jpg
mtm dirac.jpg
eris no dirac.jpg
eris 3.5 dirac.jpg
 
Last edited:
This looks great!!
Apparently it blended the drivers in your Eris pretty nicely!
How does it sound?

I personally need Jriver since I have other DSP plugins too that need to run (crossover, tube EQ)
 
Back
Top Bottom