The Quest for a near perfect 'Step Response'

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,822
Points
113
Location
Mumbai
I was reading a couple of articles in Stereophile by John Atkinson about 'Measurements of Loudspeakers' and was diving deep in this section about 'Step Response'. In this article a near perfect step response is shown as below,

28t974z.png


Only a handful of loudspeakers measured by him (around 10 out of 350) is having a near perfect step response. (Only 3% of the speakers fall in this category) Quad, Thiel, Dunlavy, Spica, and Vandersteen.... to name a few. I am not seen any reputed Indian speaker manufacture publish this data (Rethem or Cadence) or reviewed by someone else - atleast not known to me

This is a typical loudspeaker step response (97% of the speakers fall in this category) as below,

i4et76.png


I re-reviewed my own designs for the 'Step Response' and found them to fall in the not so 'Perfect' step response (97% category). Below are the step responses of my own designs build using 'Dayton Reference Woofer and Tweeter'. Though they sound very good but does not have the near perfect step response.

Dayton Book shelf:

zsjtit.jpg


Dayton Floor Stander:

1jxm45.jpg


This made me to tweak my own floorstander which has a Peerless India 5.25" Kevlar (Originally made for JPW) and Vifa-Tymphanay Tweeter -BG/OG series. I have redone the entire crossover to try and achieve maximum Phase coherence as possible and the results are amazing to speak.

Peerless + Vifa (After Modifications to crossover):

29vb8d0.jpg


This speaker now sound very nice with no clutter in the mids, smooth rounded highs with excellent detail and very real, articulate and pleasing sound with no over or under emphasis of any frequencies. This speakers now do full justice to the music by adding nothing or taking away anything from the music. Music as it should be.

So are Peerless India / Vifa drivers good.

Yes and No.

Yes if you design properly with them giving all attention to measurements and details.

No if you give a damn about phase, FR, impedance and measurements in general.

So are Dayton drivers good. Yes. They are very forgiving and do not sound very bad even if you give a damn about measurements and design with your ears subjectively only.
 
Hari, that article also mentions phase and time coherence are not all that important. Most listeners do not identify with phase coherence as per research

Also article mentions phase and time coherence cannot compensate for lack of correct tone, colorations, frequency extremes, etc.

So it's very important to understand what the customer wants and design speakers accordingly.

It's not that loudspeaker designers are not aware of these basics. It's more about what to put in the speaker design at a price point which will appeal to the audience.
 
Hari, that article also mentions phase and time coherence are not all that important. Most listeners do not identify with phase coherence as per research

Also article mentions phase and time coherence cannot compensate for lack of correct tone, colorations, frequency extremes, etc.

So it's very important to understand what the customer wants and design speakers accordingly.

It's not that loudspeaker designers are not aware of these basics. It's more about what to put in the speaker design at a price point which will appeal to the audience.

Yes Prem, I am aware of that as I have read that full articles many times as i 'Believe' in measurements. Even my own designed speakers do not fall in the 3% category but do sound quite good. More over if you listen to these speakers one may not even know if they are with perfect step response or not.

As you are aware, I also do some kind of research in these areas and post when ever I find something interesting and worth mentioning within my knowledge and capacity. Even my own signature speaks volumes about that - Our Passion. Your Advantage.

Achievement of a near perfect step response is very difficult and is a dream of many loudspeaker manufactures and they strive for it. No wonder only a few has achieved it till date no matter what the price point of their speakers are. IMO any manufacturer who achieves near perfect step response will blow their horns and mention it with pride. Its an achievement for them again IMO.
 
Last edited:
I measured the Excess Phase of my speaker two days ago and found that the arrival time for the woofer and tweeter are not completely in phase in the vertical plane. There was a shift of around 32.9 deg in the two. I was considering many options to bring it as close to zero (ideal) and tried some options of putting some foam around the baffle, some leather etc. But it worsened the situation.

This is the excess phase without Phase Plug on tweeter:

25k0569.jpg


Finally thought of putting a phase plug on the tweeter with caution.There was a bit of risk in putting the phase plug on the tweeter as that could damage the tweeter permanently beyond repairs, but i was okay with it. Hence added this phase plug and later measured them. The phase difference now between woofer and tweeter is now around 24.9 deg.

Phase Plug added on Tweeter:

2mcc4uq.jpg


This is excess phase with Phase Plug on tweeter:

2cx9lhd.jpg


Also there was a considerable imporovement in the step response as below compared to the first post,

biuees.jpg



I think i must have somewhere messed up in calculating the correct offset required between the woofer and tweeter causing this time delay between the woofer and tweeter. Now the only option remains is to introduce a time delay to the tweeter which will minimize this offset effect. The trick is to calculate and implement the exact time delay required to have complete phase coherence between the woofer and tweeter which i am currently working on.

Thanks for looking.
 
Tried applying time delay, but doing in a passive network was a challenge and it disturbed filter phase, impedance phase and overall linearity in my simulator.

Applied logic and tilted the baffle by raising the rear spike by a few mm to bring the woofer and tweeter center in one plane (My tweeter is below the woofer causing me to tilt forward rather than backward). Again measured the excess phase and step response. This tilt now keeps the phase angle between woofer and tweeter to around 4 deg in the critical 300Hz to 10Khz region. Also the step response shows a better right triangle than the previous observation.

mhuj6e.jpg


zv1wjp.jpg


So how does it sound with these objective achievments. Well i can just say that I have not heard anything like this before.

I can now formally close this project and sit back, relax and listen to music now (till now i was listening to speakers) leaving everything behind.

Any FM keen to discover this speaker are welcome after PM
 
The downward tilt on my speaker was causing me to slouch a bit to have the correct imaging. Hence i have now optimized the tilt for normal seating position at my ear level. Also have toed-in the speakers for razor sharp center image focus and depth. Now dont have to slouch.
Below is the new step response and excess phase,

2agv1j.jpg


30arqmx.jpg



I have recalled some of my friends to return their speakers for remodification of their crossovers with the new technology. One of them have returned his speakers from North India and I will be carrying out the modifications in them too. Since they are the most acclaimed peerless 6.5" kevlars and Peerless 1" Tweeter used by our diy community, I will post the details in this post for the review of the diyers. Cheers.
 
Very much intrigued by your posts, Hari, can you help me with understanding what an imperfect step response sounds like and how a perfect step response would sound like.

I am not too great in Elec Engg so treat me as a 1st year Engg grad with a good background of maths.
 
IME, the most striking difference is a huge reduction in high frequency noise - meaning no extended sibiliance, no grain, no fatigue at high SPL levels. Just sweet sounding and making you to extend your listening session. I heard music till 2 AM after time aligning my speakers.

Also the image focus is razor sharp at the center, the bass is very tight and extended without boom or chesty feeling - reduced low frequency distortions. Just accurate and tight. Also the instrument timbers are more natural and the transient attacks very smooth and perfect.

These drivers are costing around 5K (both woofer & tweeter) and so much is the improvement effect, i cant imagine how it will sound on expensive drivers.


Before the change, the high freq had an impression of a grainy and more bright feeling. This was due to the early arrival of the tweeter sound before the woofer due to the phase lag of the woofer. Also the transient sounds were less than perfect and i wanted that to end soon in the music. The vocals sounded a bit compressed and bright and not very natural as it should have been.
 
Last edited:
Get the Award Winning Diamond 12.3 Floorstanding Speakers on Special Offer
Back
Top