The Official HORNLESS Compression Driver Thread

Hi Jon,

I like these subjectiv benchmark between two drivers which share... nothing but their good reputation !

If you EQ the both with the same targett curve, slope, band width & XOs, what are the différences between the ATC and the Beyma listened at the same SPL ?

more details/clearness?
more speed ?
more slam ?
more dynamic (subjectiv feeling it goes louder faster and the damping allows even fastly low details between two musical phrases) ?

Have you the same feeling at two average spl level (casual, then loud) ?

In a second time, it could be interressant to check by curiosity if a compression driver with open back could only beam as a equivalent 2" classic dome for instance (if so you should not use it as low as 450 Hz but maybe it can swap with sucess the treble driver and share also the upper mid : pics & nulls différences are littlier and littler when the frequencie increases... iirc).

Is it not also the difference with the material of the radiating surfaces ? metal vs coated fabric dome ?
 
Last edited:
Hi Jon,

I like these subjectiv benchmark between two drivers which share... nothing but their good reputation !

If you EQ the both with the same targett curve, slope, band width & XOs, what are the différences between the ATC and the Beyma listened at the same SPL ?

more details/clearness?
more speed ?
more slam ?
more dynamic (subjectiv feeling it goes louder faster and the damping allows even fastly low details between two musical phrases) ?

Have you the same feeling at two average spl level (casual, then loud) ?

In a second time, it could be interressant to check by curiosity if a compression driver with open back could only beam as a equivalent 2" classic dome for instance (if so you should not use it as low as 450 Hz but maybe it can swap with sucess the treble driver and share also the upper mid : pics & nulls différences are littlier and littler when the frequencie increases... iirc).

Is it not also the difference with the material of the radiating surfaces ? metal vs coated fabric dome ?


Once EQd and SPL-matched, the Radian V.S. ATC... and pretty much anything including a Dayton woofer (!) are ''impossible'' to identify in a blind test. 🙁

http://www.diyaudio.com/forums/mult...t-midranges-shocking-results-conclusions.html

Subjectively, that's another matter! 😉
 
^^

my conclusions regarding that midrange drivers blind test:

1. Auditory capacities of humans are massively overestimated by audiophiles (and probably by most humans as well)

2. Frequency Response is King.

3. Once EQ'd, a 10$ midrange can mimic a 1500$ midrange, if within mechanical/electrical limits.

4. DSP/EQ/in-room measure tools might be the best investment an audiophile can make in our era.

5. Others will have to continue spending hundreds and thousands for a natural uncorrected FR.
 
^^

my conclusions regarding that midrange drivers blind test:

One that fires like a laser vs one that has wide dispersion (ie CBT36) will sound very different even if their on axis frequency is the same. Just using those as extreme examples. This is confirmed in Harman's blind listening tests as well; listening to speakers behind a curtain, not recordings of speakers.
 
One that fires like a laser vs one that has wide dispersion (ie CBT36) will sound very different even if their on axis frequency is the same. Just using those as extreme examples. This is confirmed in Harman's blind listening tests as well; listening to speakers behind a curtain, not recordings of speakers.

like explained in the thread, and the other one before: we have tested many different drivers with very different properties and got the same results: No one was really able to spot them in the blind test, even by using different music and in a controlled-environnement where the listeners were able to focus and concentrate even more than most normal listening conditions.

Matchup 1 was a ''cheat'' with 1/2 octave missing on one of the drivers, that's the only way we were able to get some positive identification results (and not that much!) 😱

So, i'm taking pretty much anything with a grain of salt, now. Subjectively speaking i may THINK i can hear something, but if i had to prove it with a blind test... Maybe i wouldnt be able at all.

One thing for sure, though: the Radian 950PB's output capacity is very high and that's my best option to match my line array.
 
I don't know if your measurements are made at 1 m on axis, mic on the floor or at listening position or even if the blind test is still ok with different drivers having all different off axis patern but the game could finally be to find the best curve !

You inject the curve in the DSP and you forgett the price of the drivers ! It changes the game or move it to a different playing field : the room and matching the drivers for a good off axis overlapping... It's perhaps the game of the good designers already? Choose the good box and drivers for the most common rooms then apply the best TCO you can for the best faisable margin or the marketing targett of the consumers ?!

Was it the purpose of Linkwitz ? Or a natural evolution of the markett ? Look the buzz of last Elac products for instance !

Seems it's a different game between passive and active speakers ? More difficult to play with passive : time & cost development being more expensive...

So how moved the line today for the sound enthusiasts we are ? We often talk now about patern, controled directivity, still about efficienty, phase and impedance plot.... OB vs Horn vs planars vs waf & therapy because blind test... but we have still tons of different speakers and spent a life and great money for this quest ! 🙄

I confess I am the first always to hesitate when it comes to choose between two colors : "choisir, c'est renonçer !" :crazy: (choucroute or cassoulet is always a torture in the restaurant when it comes to choose...) But there are such gaps between two designs that I'm asking myself if the simple "trade off" term can always be involved as an excuse ?!


However it seems to me things have evolved towards better sounding & cheaper diy speakers ! I'm forwarding to see if cheap DSP will invade our DAC, speakers to reach our goals or if the blessed diamonds married with toxic Beyrilium continue to be the dream of everyone ? 😕

At least, always a pleasure and refreshing moment to follow your experiments, even if I take it with a grain of salt, maybe I'm simplist but I believe it goes towards a good direction !
 
Last edited:
I don't know if your measurements are made at 1 m on axis, mic on the floor or at listening position or even if the blind test is still ok with different drivers having all different off axis patern but the game could finally be to find the best curve !

You inject the curve in the DSP and you forgett the price of the drivers !


All that blind test says is once equalized, different drivers can mimic each other's ''sonic signature'' to most human ears.

Will it be the same in stereo ? I don't know, but my guess is we'll have a lot of listeners still confused about which is which.

One interesting test would be to make the same midrange blind test but with half dozen drivers spreaded on 1 meter or so (mini line array) compared to a single driver, and then observe if people can, blindly, spot the ''bigger'' source. And if so, from what distance...
 
I just see SB Acoustic have published curves at 37 cm ! Urk !

Hey : you may like a Forsell DAC with new IanCanada multi dac inputt board !

But I really don't know if you can EQ a MiniDSP to sound like a Forsell DAC ?! IMHO all can not be measured or if so there are problems to match some measurement with the translation on how it sounds in real life to tailor as one would like !

But all these experiments goes towards a better understanding for sure ! All these experiments should havea double blind test in an other room as well!
 
  • Like
Reactions: grec
But I really don't know if you can EQ a MiniDSP to sound like a Forsell DAC ?!

🙂

Haha, i wish, but no.

EQ act on frequency response, not on the source quality.

By example: usually the high-end converters are silent. It's very obvious (blind test kind of obvious) with high sensitivity drivers. Right there: no mistake possible.

DAC's blind test is the only blind test so far i successfully pass. I failed midrange blind test but i got the DACs one... Go figure! 😱
 
Just removed my hornless 950PBbe and replaced it by a more conventional but still decent 3fe22 midrange...

Oh boy, i already miss the Radian.

The very first (bad) impression concerns the change in the bass perception: i lost the ''speed'', the ''attack'' and precision of the overall bass presentation. It now feels duller, boomier, not as vivid as it was before.
 
That's what I thought 🙂

I stay with my AYA II TDA1541A S1 for the moment ! I'd like to know how these difference of transcription are translated in term of cone movements ?! Often between two dacs there is more or less outputted informations... but détails and clearness is not all in the story about finding a musical DAC !

But all should be measurable, I'm just not convinced by the fact you can for the moment measure all with a scope ! Many things but not all ! Or we don't know yet all the things to measure and where to measure it maybe ?! (there are too much difference between DACS whatever the price and the engineers are all coming from the same schools all other the world and learn the same things) ! All the story is not said yet ! It's not as easy as : you have money, buy the ferrari it's the best you can have !
 
Hey, try the Radian 750 with no back cover to see if it can swap also the upper driver (before the 950 comes back from box holydays)! (but you can't have a so low 450 Hz with a 1.5" open compression imo... Many seems to like the Bastanis speaker sound and I surmise it's not about the big cones in OB but more because their 1" OB compression without backcover ? Perhaps, perhaps not !
 
Hey, try the Radian 750 with no back cover to see if it can swap also the upper driver (before the 950 comes back from box holydays)! (but you can't have a so low 450 Hz with a 1.5" open compression imo... Many seems to like the Bastanis speaker sound and I surmise it's not about the big cones in OB but more because their 1" OB compression without backcover ? Perhaps, perhaps not !

what makes the 950 special is the 400-800hz region. Can't have that with the 760NEO. I tried to EQ-push it and even though it's 7mil-shim-equipped, its obviously out of his comfort zone: i would not even use it below 800-900hz, especially not hornless...

All my hornless enthousiasm comes from the results i have with the 950. The 400-600hz is like nothing i heard before, especially hornless where there is great openness and an absence of the typical horn sonic signature. All in all, awesome and unique feeling.

Also, i'm using 300db/octave or so slope, so i got the maximum energy right from 440hz (my prefered xover point so far), that might also help extract the most from it.
 
(UPDATE 17th December 2016 @ end of this post)


I have heard about (very) few people playing with CD in hornless config, but i had to try it by myself.

Well, i did.

I got my hand on two pairs of CD:

760NEO (3'' voice coil + 2'' exit)
950PB w/ Beryllium (4'' voice coil + 2'' exit)

For the sake of testing, the ''hornless plate'' is simply a plate made of 15mm thick bamboo. I'm planning to have something even ''more hornless'' with a non-magnetic steel plate (316 stainless).

So, the big question: DOES IT WORK ?

Short answer: yes.

Ok, but does it work GOOD: ?

So far, I LOVE IT.

Check the images....

That's the 950, crossover points 4120hz... 480hz (!)
Also check the EQ: instead of boosting the lower-end, i reduced what i consider the peak of the driver (-13db seems a lot, but... it works flawlessly)

Note: it shows two dips around 560hz and 1400hz that i didnt took care yet, but the important thing is i still have energy at the crossing point of 480hz, which is nothing short of amazing. Since i play at moderate volume (domestic, not stadium needs here) i don't ear or feel any problem whatsover (distortion, xmax limit, etc..)

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.


--------------

(update 17th december)

Quick notes about my impressions on both Compression Drivers in general and Hornless specifically:

PROs
.: Ultra detailed mid, you don't hear that kind of resolution, energy and micro-dynamics from any cone between 400hz and about 1khz.
.: Blend very nicely with ribbon tweeters
.: Hornless config delivers a great airy, open, 3D-like feeling and sonic presentation.
.: High SPL output potential
.: CD drivers like the Radian 950PB and some BMS can have low crossover points, such as 250hz
.: Made for very demanding PA, they permit agressive EQ corrections and lower xover point in domestic environnement.
.: No need for separate enclosure
.: Hornless = simple, compact
.: Changeable diaphragm in case you push one too far...

CONS
.: Most high quality CDs are expensive, especially the neodymium versions
.: Not because it's hornless they lose their high efficiency... and the noise that comes with it, if you have poor electronics...
.: Requires active config for optimal results (massive gain adjustments, EQ, steep xover slopes, etc..)
.: Could work in 2-way (CAST) but optimal results in 3-way or 4-way
.: I don't think the CD technology in general has much to offer passed 4khz in terms on subjective quality, even from the smaller ones. AMT and Ribbons are superior in the last 2 octaves, IMO.


I just want to say I have experimented with using a deqx and agree about midrange potential using a 2" exit compression driver (4" dome) hornless. I also agreee about CD's not sounding good over 4khz. I want to try using the my B&C950ti's in a shallow waveguide to match the directivity of the satori beryllium waveguide tweeter. SOunds like fun!!
 
Recently, testing a beyma tpl200 against a set of compression drivers, I came across a CD that to my ear seemed to sound better than the others. The 18 sound nd3st. To do the tests I used all the CD without a horn. This was because I wanted to avoid adding a second variable to the equation: the horn. I did not plan to use the drivers without the horn at all, I simply wanted to see how it sounded before adding the horn.
After several trials I decided to replace the tpl200 with the nd3st, at which point I set about finding a suitable horn for this driver. Of course, I started with the horns that were created just for it. Then I tried a variety of others. Incredibly (or maybe not) with the horn (any horn tried) this driver could no longer sound better than the tpl200 as it did before. Although it sounded fuller in the lower part of the band of use, it lost something that I have a hard time explaining but that made it superior to the tpl.
Doing blind listening with some friends, I noticed that the preference was not just mine: they too continued to prefer the sound of the driver without the horn. I must admit that I tested response and distortion of the hornless driver exclusively at 90 dB. At that volume the distortion was very low, the response a bit bumpy, but working on active was not a problem.
Trying to understand these strange results that went against everything I had ever studied I came across this forum.
My suspicions are focused on two aspects: a certain variable amount of sound "through a tube" introduced by the trumpet and the wider dispersion that is achieved without the use of the horn.
Can anyone help me understand what is going on? How is it possible that a driver that is meant to be used in a horn sounds better without it?
 
Last edited: