One of the most mindblowing demos ever was many years ago with a simple setup of Audio Physic Virgo 3 speakers run through monocoupled Burmester 956 amplifiers, no subwoofers. These speakers I think have LR2 crossovers, quite simple…. Detail and imaging was simply unreal insane
Maybe it’s not about crossover topology, but execution of EVERYTHING in the design of the speakers
Maybe it’s not about crossover topology, but execution of EVERYTHING in the design of the speakers
You are right, there is a lot that needs to be right to get good imaging and sound stage. I have built several active DSP controlled speaker projects and pilot setups the last 4-5 years and have found that:
- The speaker should have a smooth on-axis SPL, DSP and good drivers helps
- A wide baffle aka Grimm LS1/BBC helps avoid reflections from front wall
- Put bass driver low to get gain from floor and avoid reflections
- Sealed enclosure gives tight and good bass
- No sharp directivity transitions between drivers, cross lower
- Point source, full-range/coax is really good for the top end
- Phase linear filters (not corrected afterwards as Dirac) are really good
I really think that using DSP control and high quality drivers, a two-way is enough as the top full-range easily can be linearised by the DSP.
There Is a host of factors aren´t there:
- a Wide baffle, like Sonus Faber Stradivari, insane imaging
- a Narrow baffle like Audio Physic Virgo 3, insane imaging
- Wide or Narrow baffle, both can work 😕
- Radiation pattern
- Distortion
- Power compression
- Sensitivity
- Cabinet resonances
- Maximum volume
- Bass Extension
- Diaphragm breakup
- Cabinet reflections being radiated back through diaphragm
- Suspension unlinearities
- Steep crossovers? 😕
- Shallow crossovers? 😕
- Diaphragm material, soft or hard? 😕
- Sealed enclosure works really well
- Bass reflex works very well, like Vienna Acoustics "Die Musik"
- Sealed or bass Reflex? 😕
- Point source is crazy good, Like Piega Master One
- Line source is crazy good, like Adyton Gracili or Piega Master Line Source 2
- Point source and line source is both crazy good 😕
Where to go from here? 😕
Last edited:
He he... yes it is a bunch of compromise. Always. 😀
The thing is to find the ones you need to please you.
Not easy but a bit of retro engineering on models you liked help to discern what does matter to your own preference AND how to achieve it! And sometimes very different approach might bring same outcome!
But tbh, i'm sure you are on right track: keep an open mind and identify the differents path which lead to what you want.
The thing is to find the ones you need to please you.
Not easy but a bit of retro engineering on models you liked help to discern what does matter to your own preference AND how to achieve it! And sometimes very different approach might bring same outcome!
But tbh, i'm sure you are on right track: keep an open mind and identify the differents path which lead to what you want.
I’m the same - confused! There seems to be a well defined recipe for getting a speaker that is capable of a large soundstage, but the formula for imaging is not clear to me. If wide baffle was the answer then that’d be great, but it’s obviously not.
Very good point, and as "stretchneck" says, it's confusing. But that is the beauty of loudspeaker design, it's complex fun and there is no such thing as the ideal speaker. One example is wide vs. narrow baffle, we have read from the glossy magazines that narrow produces superior imaging, right?
Inspired from the Grimm LS1 (wide), I designed and built two full-range-driver speakers, one narrow and one wide. The wide has just as good imaging as the narrow one. But the wide has improved upper bass since it lowers the baffle step frequency. However, if I might give one advice, it would be to build a simpler rather than overly complex speaker, and DSP control helps in that regard.
Wide baffle with large roundover and a set of other parameters including size of driver, xover freq and directivity management MAY lead to a good imaging IF the location within the room and the room itself are well managed too.
Good 'Imaging' comes from low interaction with the room ( Early Reflection management) which counterinteract what Stretchneck call a large 'soundstage' ( which in my view is a 'corrupted' one as the envelopment you get from ER is a distortion of the message).
That said you could have both effect together by treating ER ( make them the less annoyable possible for a good imaging: absorption and create an RFZ) and at the same time use some loudspeakers delayed in time ( but not arbitrarly! There is some time window upon which they are not harmfull to imaging) to bring back the feeling of 'envelopment'.
Take a look at Wesayso approach for example ( he use a second pair of delayed wideband to reintroduce 'ambiance' within his room AND he implemented absorption at ER location based on D.Griesinger research).
Stretchneck, we had same discussion 4 or 5 years ago! 😉
In this discussion i constantly asked for definition of terms used, and i will stil do. 🙂
I own both kind of loudspeakers too, and there is still some difference to them regarding imaging: the wider ones are less pinpoint imaging. It comes from the way diffraction is managed and at which freq it happens.
I like both rendering but not for same genre of music: with acoustic ( including amplified music like R&R or R&B ( old school) , funk,...) i prefer the larger one's as they mimic MY feelings of a real event.
For electronic music the more pin point imaging CAN be more interesting in that they locate more easily on the (abstract) sonic scene.
All in all i prefer the larger one but could live with the smaller one too...
Good 'Imaging' comes from low interaction with the room ( Early Reflection management) which counterinteract what Stretchneck call a large 'soundstage' ( which in my view is a 'corrupted' one as the envelopment you get from ER is a distortion of the message).
That said you could have both effect together by treating ER ( make them the less annoyable possible for a good imaging: absorption and create an RFZ) and at the same time use some loudspeakers delayed in time ( but not arbitrarly! There is some time window upon which they are not harmfull to imaging) to bring back the feeling of 'envelopment'.
Take a look at Wesayso approach for example ( he use a second pair of delayed wideband to reintroduce 'ambiance' within his room AND he implemented absorption at ER location based on D.Griesinger research).
Stretchneck, we had same discussion 4 or 5 years ago! 😉
In this discussion i constantly asked for definition of terms used, and i will stil do. 🙂
I own both kind of loudspeakers too, and there is still some difference to them regarding imaging: the wider ones are less pinpoint imaging. It comes from the way diffraction is managed and at which freq it happens.
I like both rendering but not for same genre of music: with acoustic ( including amplified music like R&R or R&B ( old school) , funk,...) i prefer the larger one's as they mimic MY feelings of a real event.
For electronic music the more pin point imaging CAN be more interesting in that they locate more easily on the (abstract) sonic scene.
All in all i prefer the larger one but could live with the smaller one too...
Last edited:
Alfegutt,
I agree about your opinion of the simpler the better. At least to evaluate some effects.
Then, i think the most important for a good global rendering is to have a valid approach of the physical layout of the loudspeaker ( box dimension, location of drivers, number of them and how they interact ) rather than focus on the type of drivers ( their 'inner' mechanics, the membrane material, etc,etc) at first.
Once you have something which goes with the physic at play ( rather than against it) it is time to find the 'best' driver for the application.
Iow i'm in same camp as member Tmuikku approach.
I agree about your opinion of the simpler the better. At least to evaluate some effects.
Then, i think the most important for a good global rendering is to have a valid approach of the physical layout of the loudspeaker ( box dimension, location of drivers, number of them and how they interact ) rather than focus on the type of drivers ( their 'inner' mechanics, the membrane material, etc,etc) at first.
Once you have something which goes with the physic at play ( rather than against it) it is time to find the 'best' driver for the application.
Iow i'm in same camp as member Tmuikku approach.
If I may add, there's a big difference between a wide and controlled directivity speaker's (large) soundstage and the method of absorbing early reflection energy and adding a Haas kicker.
A speaker with wide but controlled directivity can produce a large soundstage. But it will always be wide, on every song played. It is like having the same sauce on every song.
While it can be lots of fun and energetic, it won't fit all kinds of music or music production in my opinion.
I've had both types to find out what I liked. When I first absorbed early reflections I was shocked by the increase of imaging precision. But I also lost that somewhat "live vibe". Another thing it uncovered is the influence of cross talk on stereo. That's a whole other story.
The plan was there from the start to add a virtual Haas Kicker, something I first heard/read about on an automotive audio forum. Basically it gives you the tools to make a small room sound larger.
When I started playing with ambience speakers, I tried many things as the source of the ambience. I downloaded actual impulses from (some well known) rooms and spaces and used that as a back drop. It worked quite well, but reminded me of the wider soundstage I did have earlier when I was still without absorbing panels for early reflections (can't use ER, krivium owns the trade mark 😉). It had the same feeling of adding a similar effects on every song. The same sauce recipe.
In the end I used a combination for my Haas Kicker, where the main property is L-R and R-L, band passed and delayed with a bit of L+R mixed in (otherwise the center would be too dry sounding, there's freedom to play with different delay here) augmented with a snuf of reverb, to mimic that larger space and provide envelopment. In my case I use Lexicon's Random Hall algorithm. As that is what Griesinger had mentioned in his papers for a processor he developed for Harman. I guess it was also part of Logic 7?
One other important thing to mention is that the ambience backdrop should be decorelated. I let it deffract/reflect off of things close by by aiming the ambience speakers out to the sides and this way having no direct path to the listener.
This gives a huge stage if the source asks for it, but it remains small if the source is small. It lives and breath's with the music if one can describe it like that. To have imaging + stage combined is a real joy.
Some folks will say this isn't true to the source. But nothing is, as long as we listen in a living, breathing home. I've called it MyFi in another thread. My preferred way to hear the soundtrack of my life. And I have tried many ways! I've made it my personal study case, making me listen to a lot of music (ohh the horror 😀).
A speaker with wide but controlled directivity can produce a large soundstage. But it will always be wide, on every song played. It is like having the same sauce on every song.
While it can be lots of fun and energetic, it won't fit all kinds of music or music production in my opinion.
I've had both types to find out what I liked. When I first absorbed early reflections I was shocked by the increase of imaging precision. But I also lost that somewhat "live vibe". Another thing it uncovered is the influence of cross talk on stereo. That's a whole other story.
The plan was there from the start to add a virtual Haas Kicker, something I first heard/read about on an automotive audio forum. Basically it gives you the tools to make a small room sound larger.
When I started playing with ambience speakers, I tried many things as the source of the ambience. I downloaded actual impulses from (some well known) rooms and spaces and used that as a back drop. It worked quite well, but reminded me of the wider soundstage I did have earlier when I was still without absorbing panels for early reflections (can't use ER, krivium owns the trade mark 😉). It had the same feeling of adding a similar effects on every song. The same sauce recipe.
In the end I used a combination for my Haas Kicker, where the main property is L-R and R-L, band passed and delayed with a bit of L+R mixed in (otherwise the center would be too dry sounding, there's freedom to play with different delay here) augmented with a snuf of reverb, to mimic that larger space and provide envelopment. In my case I use Lexicon's Random Hall algorithm. As that is what Griesinger had mentioned in his papers for a processor he developed for Harman. I guess it was also part of Logic 7?
One other important thing to mention is that the ambience backdrop should be decorelated. I let it deffract/reflect off of things close by by aiming the ambience speakers out to the sides and this way having no direct path to the listener.
This gives a huge stage if the source asks for it, but it remains small if the source is small. It lives and breath's with the music if one can describe it like that. To have imaging + stage combined is a real joy.
Some folks will say this isn't true to the source. But nothing is, as long as we listen in a living, breathing home. I've called it MyFi in another thread. My preferred way to hear the soundtrack of my life. And I have tried many ways! I've made it my personal study case, making me listen to a lot of music (ohh the horror 😀).
Last edited:
Hmmm... I might just not agree here, because it depends on which drivers we use. A wide baffle can have a wide directivity all way, but it can also have a wide directivity, and controlled tapered directivity from midrange to treble. And I agree, pintpointing and a precise soundstage will benefit from this. One example is the acoustical adjustable lenses on Bang & Olufsen speakers.
Another simpler way of tapering in the directivity on a wide baffle is to use a fullrange driver for the top end, a lot easier for the DIY-er than a lense. As the driver starts to beam (depending on driver diameter), the directivity tapers into a more narrow pattern.
Hi Alfegutt,
I understand Wesayso comment differently: when he talk about wide and controled directivity, i think he talk only about directivity, not the width of the box.
A commercial example could be Atc monitors, which have (very) wide directivity all the way up by splitting freq range with drivers used only within their optimal Ka. In this manner it is controled but wide radiation.
ATC speakers: engineers reveal 7 design secrets in exclusive Q&A - Acoustic Frontiers
I agree about fullrange. Coax are another way to deal with this too.
I understand Wesayso comment differently: when he talk about wide and controled directivity, i think he talk only about directivity, not the width of the box.
A commercial example could be Atc monitors, which have (very) wide directivity all the way up by splitting freq range with drivers used only within their optimal Ka. In this manner it is controled but wide radiation.
ATC speakers: engineers reveal 7 design secrets in exclusive Q&A - Acoustic Frontiers
I agree about fullrange. Coax are another way to deal with this too.
Last edited:
Correct, I wasn't talking about the differences between a wide or narrow baffle, just talking about wide directivity speakers in general. Don't get me wrong though. It can work quite well with certain material. I just think that for studio oriented work better solutions are available.
Both wide baffle and narrow baffle can work well for imaging precision. Even better if one avoids diffraction as much as possible. Giving the speakers a good chance in the disappearance act. Nothing left but the music.
Both wide baffle and narrow baffle can work well for imaging precision. Even better if one avoids diffraction as much as possible. Giving the speakers a good chance in the disappearance act. Nothing left but the music.
Hello everyone, I stumbled on this post while pondering which speakers to build... The CNO4 or the Ellipticor3. I am still confused. Some stuff I was reading about the ellipticor bothered me. Coming from the NEXT4, I have tendency to go for the CNO4... And what more, I don't really see a lot of the ellipticor drivers around. Here in Germany they seem to be a bit neglected.
Any kind of comment is very much appreciated!
Any kind of comment is very much appreciated!
hi @noamgeller, good to see your post.... and this thread going back to what it was originally about 😛
I have hardly seen any designs with the ellipticor drivers, there seem to be no other DIY designs and the only commercial product I can recall seeing is the Burmester BC150, which retail around €100.000.
This speaker though have custom 18cm ellipticor bass/mid and what I think is a Revelator W32 sub driver)
And BC150 is a 3 way, while Ellipticor-3 is 3.5 way and should give even more 😀
The Ellipticor's are insanely expensive drivers and I reckon this is some of the reasoning why they are so rare in use, but are they better than the Seas Excel's? Troels Gravesen state that it's a matter of taste.
Well, the CNO-4 is Gravesen's reference speaker at the moment 😛
What OI recon should be pretty clear is that by building Ellipticor-3 we are approaching absolute statement speaker area.
Both the Ellipticor-3 and CNO-4 has clearly inspirations from Richard Vandersteen, who has sgtrategic goal of having zero baffle designs, basically fully eliminating the baffle, for better imaging. These designs do not fully employ this to the extreme Vandersteen does but it is not so far away, see below (well it is not actually fully zero baffle)
I think I have a build on the plan but not before next year 😀
CNO-4 or Ellipticor-3? 🙄
I have hardly seen any designs with the ellipticor drivers, there seem to be no other DIY designs and the only commercial product I can recall seeing is the Burmester BC150, which retail around €100.000.
This speaker though have custom 18cm ellipticor bass/mid and what I think is a Revelator W32 sub driver)
And BC150 is a 3 way, while Ellipticor-3 is 3.5 way and should give even more 😀
The Ellipticor's are insanely expensive drivers and I reckon this is some of the reasoning why they are so rare in use, but are they better than the Seas Excel's? Troels Gravesen state that it's a matter of taste.
Well, the CNO-4 is Gravesen's reference speaker at the moment 😛
What OI recon should be pretty clear is that by building Ellipticor-3 we are approaching absolute statement speaker area.
Both the Ellipticor-3 and CNO-4 has clearly inspirations from Richard Vandersteen, who has sgtrategic goal of having zero baffle designs, basically fully eliminating the baffle, for better imaging. These designs do not fully employ this to the extreme Vandersteen does but it is not so far away, see below (well it is not actually fully zero baffle)
I think I have a build on the plan but not before next year 😀
CNO-4 or Ellipticor-3? 🙄
Last edited:
Hello Haraldo, thank you for the responed
Tough world we´re living in, having to choose between those two speakers 🙂
I am more and more leaning to the CNO4. I already know (Think I know) what to excpect from those drivers... and the cost is playing a roll as well...
I just read a post about a guy who has built the Ellipticor3 and after a while changed the crossover! sacrament! That freaks me out a bit...
Tough world we´re living in, having to choose between those two speakers 🙂
I am more and more leaning to the CNO4. I already know (Think I know) what to excpect from those drivers... and the cost is playing a roll as well...
I just read a post about a guy who has built the Ellipticor3 and after a while changed the crossover! sacrament! That freaks me out a bit...
FYI: german diy speaker magazine Klang und Ton made a speaker with Ellipticors:
Ella
Ehre, wem Ehre gebuhrt | hb.hifi.
Ella
Ehre, wem Ehre gebuhrt | hb.hifi.
Ouuuuchhh... I think the CNO-4 would be really close to the Elliticor-3 and it's half price, or maybe even less, are you considering a build in the near future?
Thx, well I am going floorstander I reckon 😛
Last edited:
Yes, I will probably buy the kit in 3 weeks time... so I have still time to read and explore. Waiting till my NEXT4 reaching they´r new owner.
Here is my NEXT4 project (sold):
NEXT4 DIY Loudspeaker Project - YouTube
Here is my NEXT4 project (sold):
NEXT4 DIY Loudspeaker Project - YouTube