Hard to Impress?:
I’m not that hard to impress, I just think I know very clearly what I like and what I don't like and I don’t want to go for suboptimal quality 🙂
What Impressed me extremely much are the speakers below, I will write more about them further down.
About driver quality:
How much does driver quality affect the end result? I have no idea, I am not a professional in this field and I can’t say. What I remember is that Dunlavy picked pretty standard low price drivers because it’s what he liked the most. I also remember some pretty in depth comparisons by Danish hi-fidelity magazine quite a few years back where they found the standard simple Seas drivers to be the best ones, even beating some of the upscale Seas Excel products in their opinion. Based on measurements and listening impressions.
About that speaker with photo below… this is a one off speaker made with extreme quality of components everywhere, Accuton drivers all the way, €10k diamond tweeters (per pair) from Accuton, the crossover is housed in an external box the size of a pretty hefty bookshelf speaker, I think overall price of components exceed €35k. But here it is: The music coming out is completely unreal, the violin’s are so true to life that I need to grasp for air, you can sense the life of the vibrations of the violin cabinet, you can probably clearly spot the differences between different Guarneri violins. Grand piano…. Extremely nice.... It is not only my words, everyone I met who listened to these speakers provide the same feedback. This is with a simple Roon streamer and a quite simple Cayin tube amp.
Well, here is extreme component quality, cabinet is built like a solid brick of wood, every single part in the speaker is made with no compromises and love all the way.
I guess driver quality will make some significant difference at the end of the day, and maybe this enable DIY’ers to make genuine statement speakers without taking up a second big credit.
I never liked products with Accuton drivers, always found there to be an edge somewhere, some harshness; But this is non existent here 😛
€35k in components though is stretching very very very far, I assume these speakers would be > €130K retail product. I compared them directly to Sonus Faber Aida A/B testing, there is no compare, clearly betters the Aida with wide margin!
But for €4k to 6k of components we are still in statement speaker area 😛
Heck, even the lower priced design with cost of €2.5k is probably bringing us to statement speaker land too 😛
IMHO I think it's not because that Kii 3 is so neutral…. For three decades I lived with the Duntech PCL-15, which should be completely time/phase coherent and a frequency response very very flat. I really did not ever manage to find that they have any sound of their own, simply I find them to reveal what’s on the recording within the physical limits of what they can do. Yes, I love them till the end of the world. (Unfortunately they’re out of service now)
Measurements of Duntech PCL-15 below:
I experience the Kii as not really neutral speakers, I feel the life of music is being killed somewhere in the chain of the speakers, I dont't sense the nerve and life of music and I can’t manage to listen to a song until the end…. So that’s not for me, sorry guys
Here's the speakers in discussion, with crossover at the back:
And crossover:
Hmmmmm, not very simple, lot of components here...
Last edited:
I heard there is one build of Ellipticor-3 in Netherlands and one in Australia
Seems like not so many around
There should be more CNO-4 around, there is one guy in Norway building these at the moment ....
Seems like not so many around
There should be more CNO-4 around, there is one guy in Norway building these at the moment ....
Last edited:
Nice to see you Haraldo🙂 I have also listened to the Kii with BXT. I found it neutral, presenting an impressive and presice sound stage. However, I perceived the bass and midbass slightly "dull, lifeless". Could it be the sound cancelling concept that "sucks" up energy?
Similarly to the Kii, the Grimm LS1 also has a cardoid sound pattern, but it uses a wide baffle to achieve this. Have anyone of you guys heard or compared them to the Kii?
Similarly to the Kii, the Grimm LS1 also has a cardoid sound pattern, but it uses a wide baffle to achieve this. Have anyone of you guys heard or compared them to the Kii?
Thx mate, Likewise 😎
I reckon Kii is a pretty complex animal and it’s probably pretty easy to mess up with those DSP settings, it can be that my suboptimal experiences was just poor demos wih some messed up settings, overcompensating DSP kills life … I will try again another time I reckon.
I never listened to Grimm, but looks interesting 😛
I reckon Kii is a pretty complex animal and it’s probably pretty easy to mess up with those DSP settings, it can be that my suboptimal experiences was just poor demos wih some messed up settings, overcompensating DSP kills life … I will try again another time I reckon.
I never listened to Grimm, but looks interesting 😛
You will also note that CNO-4, alongside The Loudspeaker, are Troels current reference. Apparently CNO-4 looses out on delivery of some detail retrieval compared to the Ellipticor, but the CNO-4 does other things special and is very enjoyable when I asked Troels on the subject.
CNO-4 is also much cheaper!
Yeah I seen that …. CNO-4 is about half price and proudly with Norwegian drivers
(except for the subwoofer)
I asked Troels and he seem reluctant to say that Ellipticor-3 is any better than CNO-4
Gravesen also seem to offer some edge coating for the drivers that further improve performance
(except for the subwoofer)
I asked Troels and he seem reluctant to say that Ellipticor-3 is any better than CNO-4
Gravesen also seem to offer some edge coating for the drivers that further improve performance
It looks good, and I like that Troels use 1.order filters to keep the impulse response intact, as one coherent wavefront. It could be fun to build the speaker, but as a DIY-er I would of course do some modifications🙂
* Ditch the seas 8", make it a threeway, save money
* Go full-digital DSP controll all the way, all three channels
* Replace 1. order with steep filters, high pass the 6" should keep distortion low
* Lower treble XO for improved dispersion toward the 6"
* Flush mount treble on the baffle (use DSP delay)
* Make the bass sealed encolusure, reduce "ringing" and Group Delay
* Keep it impulse response correct through FIR filters
My goal would be to keep the speaker low in distortion and phase correct, although at a reduced price tag using modern technology. FIR filters are important in this respect, which rules out Hypex.
Last edited:
I really like your thinking @alfegutt … I thought on some of the same things, but not as extreme as you 😛
Maybe fully active with AudioVero computer based xover
If passive, I like the shallow filters
I’d like to keep the 8”, will give them more ”weight” in the midbass
Maybe hipass on the mid and midwoofer, to lower the load on midwoofers, improves midrange
I have a crazy big Adyton poweramp, speaks against the fully active solution
Ditch the ported sub, instead use two Revelator 28W woofers on the side in closed enclosure back to back…. Think volume should be big enough, PEQ the subwoofers to lift the bottom end ….
Another thought, make them passive with external crossover, and very easy to allow for making them active in future
Grounding of driver chassis seem to have a profound effect, would like to try 😛
Maybe fully active with AudioVero computer based xover
If passive, I like the shallow filters
I’d like to keep the 8”, will give them more ”weight” in the midbass
Maybe hipass on the mid and midwoofer, to lower the load on midwoofers, improves midrange
I have a crazy big Adyton poweramp, speaks against the fully active solution
Ditch the ported sub, instead use two Revelator 28W woofers on the side in closed enclosure back to back…. Think volume should be big enough, PEQ the subwoofers to lift the bottom end ….
Another thought, make them passive with external crossover, and very easy to allow for making them active in future
Grounding of driver chassis seem to have a profound effect, would like to try 😛
Last edited:
The Revelators are more than capable - absolutely. The thing is that Troels´ construction is admirable. You really need to be an expert to construct a low distortion, high SPL passive speaker, which is phase coherent. I certainly can´t, and there aren´t many around, Vandersteen is one of them.
However, using DSP and FIR filters, it get´s simpler as we are not restricted to shallow filters, this helps a lot. The drivers are much better protected, operating in their passbands only, getting rid of lobing issues, tilted baffles and distortion. Through this, we can move from four way to three way (I built a simpler two-way for myself)
But there is no free lunch, care must be taken to avoid pre-ringing, and we need a convolver to run the filters, examples are MiniDSP and CamillaDSP. Thankfully we also have the amazing re-phase software to create any type of filter.
However, using DSP and FIR filters, it get´s simpler as we are not restricted to shallow filters, this helps a lot. The drivers are much better protected, operating in their passbands only, getting rid of lobing issues, tilted baffles and distortion. Through this, we can move from four way to three way (I built a simpler two-way for myself)
But there is no free lunch, care must be taken to avoid pre-ringing, and we need a convolver to run the filters, examples are MiniDSP and CamillaDSP. Thankfully we also have the amazing re-phase software to create any type of filter.
Last edited:
Agree on everything
I had discuss with Joachim Gerhard, he svear on using AudioVero as the best sounding DSP solution he came across, it also does steep filters that are phase coherent and seemingly sound surreal good…
Taking an already proven design it should be possible to do ….
Also why I would like to have the passive option, in order to have a benchmark to compare
Two different sets of terminal?
For active and passive use?
I had discuss with Joachim Gerhard, he svear on using AudioVero as the best sounding DSP solution he came across, it also does steep filters that are phase coherent and seemingly sound surreal good…
Taking an already proven design it should be possible to do ….
Also why I would like to have the passive option, in order to have a benchmark to compare
Two different sets of terminal?
For active and passive use?
Last edited:
If a speaker is made active from the start, I assume the whole design can be different, driver selection, everything…. Because you can use completely different xover topologies and still be phase coherent …. Tweeter crossover can be much lower, yes? …. Difficult stuff this is 😛
Last edited:
This is an interesting comment - I have been thinking how driver overlap in the crossover region could affect imaging precision. The best imaging speaker I have ever heard have been Linkwitz LX521 which used LR4.
Does the wide overlap due to higher order crossovers 'smear' imaging, or is there something special about 1st order, as opposed to LR2?
I think the route to success here for you would be to set yourself goals, and decide on technology later - that is not to get into the tech first, examples of goals could be:
* Measures SPL from 30 hz - 15 kHz
* Have controlled directivity 300 hz. - 15 kHz
* Have low distortion even at higher SPLs
* I am using my Dynamic Presicon amp
* Be phase coherent, and low group delay
These examples in sum will demand large speaker, and the last goal will either result in a passive CNO-type, or a simpler DSP speaker corrected through FIR using separate amps for each channels. So your job would be to understand what do I want, and what is my DIY skill level?
* Measures SPL from 30 hz - 15 kHz
* Have controlled directivity 300 hz. - 15 kHz
* Have low distortion even at higher SPLs
* I am using my Dynamic Presicon amp
* Be phase coherent, and low group delay
These examples in sum will demand large speaker, and the last goal will either result in a passive CNO-type, or a simpler DSP speaker corrected through FIR using separate amps for each channels. So your job would be to understand what do I want, and what is my DIY skill level?
My goal is passive crossover... I just wondered how LR4 compares with wide overlap of drivers on 1st or 2nd order with regards to imaging.
I think that good imaging is caused by several factors, where phase linearity is one of them. What we have discussed here is the crossover filter, which causes such unfortunate phase deviations, the higher order the worse it gets. LR2 is often considered to be a good compromise. But there are also other factors involved in getting a phase linear speaker - look at the measured step-response, it should ideally look like a triangle with one sharp rising front.
Stereophile measures the step response, and regarding the Dunlavy, which is phase coherent, John Atkinson wrote (2008):..... "Although the subjective effects of a loudspeaker featuring time-coherent behavior are not fully understood, my subjective impressions have been that such designs always feature superb imaging and soundstaging. The SC-IV is no exception"
Last edited:
I think you can get really good imaging also using LR4 filters, no reason to worry. Please note that it is the network and drivers acoustical slopes combined that causes phase challenges, not the network alone. And almost all other factors also get worse with shallow slopes.
All of this is difficult, but living with these phase coherent Dunlavy designed shoeboxes for some decades, there is something about them that you just don´t get with more sharp crossovers, I think... Can I prove it, NO. But there seem to be something just more pure "organic" with genuine true phase coherent speakers, like "less mechanical" in loss for a better word.
Imaging is difficult to say, there are some early Audio Physic speakers that are not phase coherent that are incredible imaging champions....
Of course many more like the ones mentioned above 😛
Are the Troals Gravesen speakers phase coherent, heck none of them are!
There is some 1st order electical slopes but it´s obvioulsy not first order acoustic, this is easily proven by looking at the tweeters which are connected in inverse polarity.... no way it¨s phase coherency from mid to tweeter!
Those speakers are pretty close and have very gentle filters, but not fully phase coherent
Still I believe these are insanely great sounding speakers, very much want to give it a try!
Imaging is difficult to say, there are some early Audio Physic speakers that are not phase coherent that are incredible imaging champions....
Of course many more like the ones mentioned above 😛
Are the Troals Gravesen speakers phase coherent, heck none of them are!
There is some 1st order electical slopes but it´s obvioulsy not first order acoustic, this is easily proven by looking at the tweeters which are connected in inverse polarity.... no way it¨s phase coherency from mid to tweeter!
Those speakers are pretty close and have very gentle filters, but not fully phase coherent
Still I believe these are insanely great sounding speakers, very much want to give it a try!
Last edited:
Guess there is only one way to find out 🙄
But I am not sure I would want to go down that path !