Agreed. But DSP allows a lot of things - like driver protection thru hardy out of band rejection. With DSP it can be hard to tell what's going on, so I don't know what the slopes are.
Amen. Even a hack like me been able to build limiters with open architecture DSP, that are both frequency and level dependent.
No telling what manufacturers can do, who really know what they are doing,
Someday I'm gonna tackle level dependent sliding xover frequency...well, some truly bored day. Lol
Yeah Barry can be a bit of **** but some of his remasters are excellent I have to admit…..His take on Close to the Edge is near perfection on the Maggie’s.
Last edited by a moderator:
I thought I replied to this message, but I don't see that. My memory is failing... 🙁. Anyway....
Thanks for all the information. I was, when I started this, considering ATC, and am currently using 2 in my system. However, further discussion led me to consider other options like Volt, Satori and Bliesma. You got an interesting story with ATC, and I have a smaller one that still makes me inclined toward it. Maybe I should break away from it and move on.
I now see your point on Satori, and maybe also on Volt 2". I will have to think more about more options. Maybe ATC can be used again with narrower range. Actually I can test it in my current system after upgrading the DSP/electronics part. Maybe I should use Satori from 800Hz, or Volt 2" from 600Hz, or Volt 3" from 500Hz. Bliesma 74s could be interesting. I do have a microphone, but not sure how precisely I can measure it in my environment. I think I like the objective neutral sounding ones are good to me. I've listened to flat-tuned speakers too long, so systems deviating a lot from that makes me suffer.
So, let me flip-flop and lay out my current idea again:
2x 8" cone woofer in sealed enclosure
1x 5" ~ 7" cone midbass in a separate enclosure
1x 2" ~ 3" dome midhigh in another separate enclosure
1x 20mm ~ 30mm dome tweeter in 4th enclosure
Yes for wideband, and largish, but not a room - it's a living space between 2 small walls(1 with huge mirror!), 1 long walls with window, and left side open.....
I guess now I am more clear and in a better position to help. Yes, it is good to be in this forum. 🙂
I am glad to see someone doing/thinking the similar things here. Maybe it's natural for me to get attracted to and to drool on the idea of 4-way design. 😉
RIght, please don't! 🙂
What do you mean by 'complementary'? Talking about the information, what is your DSP H/W platform? I modified the F/W of a 16-channel surround processor, as my main audiophile source is the well-recorded classical music BDs with fantastic surround tracks. But it's quite hard to work out something with it, so I am planning to move to Mac with some S/W DSP platform like CamillaDSP.
And Alex, sorry to hear about your DSP unit. Did the fire damage the digital board? I wonder if it could be revived with fixing burned parts(likely power supply/analog?).
I envy that you could avoid the range with your other design choices/parameters. I seem unable to avoid it unless I use expensive ATC/Volt/Bliesma 3" drivers. I will keep thinking about it.
A very interesting listening room picture! it is also a bit like mine in a way, because I am using MG 1.7i as the surround/alternative stereo front in my 5.1 system with my 3 front speakers using ATC/Volt midranges. 🙂
It is obviously a very unusual system, or a combination of speakers, and I can't say it's the best combination. It works okay for playing out ambience/reverb in the recording, or the rear directional sound in movie audio. And... this managed to pass the WAF test!
Hi Jay,
I am late to the party.
I have used a 4 way system for over 20 years and have not had the need to change it yet.
The key to the attached system is not magazine curves but its ability to reproduce brutally accurate transient detail and envelope the listener a glorious immersive experience.
The image is for illustration only.
In essence it’s how the system delivers the sound.
All the drivers are 95db sensitivity or greater.
It’s a great way to evolve and finally cone up with what meets your aspirations and expectations.
Your SVS subs could be the low bass. An 8-12” mix bass, a 6 inch cone on lower midrange, a 2” dome upper mid and your choice of super tweeter which adds air.
You will need a descent crossover software like SoundEasy to work through your crossover options after taking enclosure measurements.
The trick is to come up with a baffle layout that does not compromise the smooth naked response of each driver or the diffraction which will spoil the sound stage and imaging.
I would start out with textile domes and paper cones.
The volt BM2500.4 would be my pick for the mid bass and lower mids. It’s as dynamic as hell and efficient. The Volt or Bilesma 2” upper mid textile dome and the Bilesma 1” textile dome.
Game, set and match on a well thought out baffle(s).
http://www.troelsgravesen.dk/The-Loudspeaker-1_Uladzimir.htm
Here is a similar design by Troels who was inspired by the 4 way JBL’s. Troel’s feedback was it was the best system he had come up with at the time.
http://www.troelsgravesen.dk/The-Loudspeaker-1_Uladzimir.htm
Attachments
With some objectives and an outline speaker configuration we can look at what a high performance design looks like and possibly sounds like in some cases. Then examine possible pros and cons of changes.
The wide beam tower 4 way that has become something of a reference in the last few years having performed well in blind listening tests as well as "measurement tests" is the Revel Salon 2. 3 x 8" woofers, 6.5" lower mid, 4" upper mid and 1" tweeter. The baffle is narrow and shaped with a waveguide on the tweeter. Crossover frequencies: 150, 575, 2.3k Hz. Drivers have hard cones/domes.
Another wide beam tower 4 way with a high technical performance is the Vivid Audio Giya series. 2 x 7" (G2) or 2 x 9" (G1) woofers, 5" lower mid, 2" dome upper mid, 1" tweeter. The baffle is narrow and shaped. Crossover frequencies: 220, 880, 3.5k Hz. Drivers have hard cones/domes.
The reason for the hard cones/domes is that the narrower passband of a 4 way enables the motion to remain pistonic while providing a big enough frequency gap to the resonances to reduce them to inaudible levels with wise crossover and motor design. This is pretty much ideal and is why 4 ways that emphasize high technical performance will tend to use hard dome/cone materials. Soft materials with damped resonances in the passband can be a good choice for 4 ways that are looking for high SPL or 3 ways where things are often too tight not to drive the resonances a small but audible amount at higher SPLs. Soft cone/domes are easier to work with but at the price of a small drop in performance due to the presence of resonant motion in the passband. Again resonant motion may be a good thing for an audiophile speaker rather than a high fidelity one.
The shaped baffles and careful blending of the different radiation patterns through the crossover regions is what leads to a smooth horizontal beamwidth (placing the drivers close tends to damage the vertical least). The use of FIR filters will enable even better control of the beamwidth than the analogue filters used in the two commercial speakers but it requires knowing the radiation pattern of the drivers on the shaped baffle (and how to design FIR filters). This can be achieved with 3d simulation or a lot of time and effort building and measuring. It is doable by keen DIYers although I am not aware of examples. Anyone? It is likely coming though with the current enthusiasm for BEM and DSP. A chamfered baffle rather than flat is likely to be a significant step in the right direction if one then measures and tunes a FIR crossover. Some illustrations here and here.
I agree with Macka.
If you want to keep the cabinet narrow and cant use a 10 inch driver then this is a good option.
SVS up around 70Hz.
Twin 8 inch bass drivers up to around 300 Hz /400 Hz
6.5 inch midrange up to around 1,600Hz.
Small dome up to around 10KHz to 12 KHz
Tweeter above 10K.
If you can go to slightly wider baffle, then I also agree with Macka, the gorgeous Volt 10 inch 2500.4 is a fabulous driver.
Sealed or in open baffle it really is up there with the very best when covering from 70Hz all the way up to about 1,200 Hz before its off axis performance falls away (like all 10 inch drivers do).
Re the Volt 10 inch -
This is from about 20 years ago... The Volt 2500.4 used in an open baffle covering from 70 Hz up to 800 Hz (anywhere between 300Hz and 1,200 Hz was amazing)with a Manger driver above and a sealed 12 inch "Co- Drive" unique twin coned carbon piston beast) bass.
Source and electronics were £100 K worth of modified battery powered DEQX, with external NOS DAC's, A to D etc, etc... Wow... The system reduced a dozen drunk Norwegians to tears on opera and had them ready to invade Russia on the "Gladiator" sound track!
PS Note the curved sides on the open baffle which staggers the effective baffle width eliminating the step... Variable Path Length or VPL... Ooops!
This is from about 20 years ago... The Volt 2500.4 used in an open baffle covering from 70 Hz up to 800 Hz (anywhere between 300Hz and 1,200 Hz was amazing)with a Manger driver above and a sealed 12 inch "Co- Drive" unique twin coned carbon piston beast) bass.
Source and electronics were £100 K worth of modified battery powered DEQX, with external NOS DAC's, A to D etc, etc... Wow... The system reduced a dozen drunk Norwegians to tears on opera and had them ready to invade Russia on the "Gladiator" sound track!
PS Note the curved sides on the open baffle which staggers the effective baffle width eliminating the step... Variable Path Length or VPL... Ooops!
Attachments
Cool discussion,
Lots of relevant input.
Has a Wilson Audio or JBL L250 Ti sloping baffle been discussed?
It’s difficult to appreciate one listeners conscious preferences from another.
But l have found well heeled audiophiles to proclaim they will pass up a $70K system after auditioning the JBL system in the attachment above.
I think Eric Alexander of Tektron has a similar philosophy with his patented Tympany ring radiator array. Get the efficiency up and the distortion down from the lower mids to the upper mids. Anything above 7000 hertz is Air. Using a dedicated driver eliminates all the problems two way woofer - horn systems have with polar response and mass roll off above 3Khertz, diaphragm break up ect.
Lots of relevant input.
Has a Wilson Audio or JBL L250 Ti sloping baffle been discussed?
It’s difficult to appreciate one listeners conscious preferences from another.
But l have found well heeled audiophiles to proclaim they will pass up a $70K system after auditioning the JBL system in the attachment above.
I think Eric Alexander of Tektron has a similar philosophy with his patented Tympany ring radiator array. Get the efficiency up and the distortion down from the lower mids to the upper mids. Anything above 7000 hertz is Air. Using a dedicated driver eliminates all the problems two way woofer - horn systems have with polar response and mass roll off above 3Khertz, diaphragm break up ect.
One thing worth considering is the height of the driver array?
With a long array how do you point each driver for the same path length and on axis to the listening position?
An arched array consisting of modules with round overs (aka Wilson) could help with fine tuning the spatial coherence.
The mid bass- lower mid could be mounted inverted like the Dynaudio Contour or a variation of Wilson.
Keeping the floor cancellation out of the mid bass and lower significantly improve the fundamental for the significant overtones and harmonics.
It’s interesting that Wilson overlaps their midrange drivers according to JA Stereophile.
Jay,
When you get around to narrowing down your selection l’d be happy to receive some driver baffle measurements and simulate some possibilities in Leap 5. REW can incidentally export complete files in an email with multiple measurements. These files can then export data in FR and phase data points for a Leap simulation. Leap has an excellent suit of passive and active filter designs.
The newer upgraded Hypex Fusion amps have an FIR DSP crossover function.
I do believe however the selection of drivers will be the major force behind the delivery of the results in this type of multi way system.
Looking at the simplified diagram attached keeping the upper bass - lower mid driver away from room boundaries can remove 1/4 wave length reflections that cause 1/2 wave length cancellations. These cancellations result in notches or troughs in mid bass. Ideally we want the direct mid bass wave front that extends up to 1000 hertz in the 2500.40 to arrive without destructive room interference.
With a long array how do you point each driver for the same path length and on axis to the listening position?
An arched array consisting of modules with round overs (aka Wilson) could help with fine tuning the spatial coherence.
The mid bass- lower mid could be mounted inverted like the Dynaudio Contour or a variation of Wilson.
Keeping the floor cancellation out of the mid bass and lower significantly improve the fundamental for the significant overtones and harmonics.
It’s interesting that Wilson overlaps their midrange drivers according to JA Stereophile.
Jay,
When you get around to narrowing down your selection l’d be happy to receive some driver baffle measurements and simulate some possibilities in Leap 5. REW can incidentally export complete files in an email with multiple measurements. These files can then export data in FR and phase data points for a Leap simulation. Leap has an excellent suit of passive and active filter designs.
The newer upgraded Hypex Fusion amps have an FIR DSP crossover function.
I do believe however the selection of drivers will be the major force behind the delivery of the results in this type of multi way system.
Looking at the simplified diagram attached keeping the upper bass - lower mid driver away from room boundaries can remove 1/4 wave length reflections that cause 1/2 wave length cancellations. These cancellations result in notches or troughs in mid bass. Ideally we want the direct mid bass wave front that extends up to 1000 hertz in the 2500.40 to arrive without destructive room interference.
Attachments
Last edited:
The issue I see/have is why limit the passband of the 8” woofers so much for the need for a midbass driver at all?…….its like eating at a fancy restaurant where the place is set with 3 forks…..more pretentious than useful IMO.
My plan would still be a three way….slim tower with low midwoofer center 18” off the floor and low passed 1st order followed by your choice of midrange driver and then your tweeter all topped off by the remaining 8” woofer on top. Depending on your listening height, you could even invert the mid/tweeter…..Linn has been doing this with much praise and success for decades. Point is the line source behavior is going to present better in most rooms as the drivers will sum sooner and more coherently.
But HERE is the most important part of this whole exercise……avoid the French fries and Ketchup performance of the typical modern 2way box that places the mid/tweeter crossover in the 2-3khz range……….that’s the best way to ruin all,of the detail you’d be getting from using a dedicated midrange driver. Mind you….it is a populist thing to do these days…..it’s pretty simple to make everything sound good……like a French fry……but nothing sounds great……no standout on the plate…..just the French fries and 3 forks to choose from.
This sounds crazy but please bare with me.... You might want to invest in a couple of plastic covered (reduced ringing see below!) coat hangers...
When I first got a DEQX which allowed me to time align driver centers I made a coaxial point source with a Volt 2500.4 and a Morel tweeter ( I cant remember which one, but it came with the Volt as part of a Wilmslow Audio "Vortex" kit) which sounded fantastic but looked.... Like a "Dalek chewing a wasp" as my wife called it! Way back then tweeters were big... Today you can get tiny tweeters / domes and 3D printing "cones" is cheap.
Your 2 or 3 inch domes would be ideal... Mounted in an "ice cream cone " shaped / 3D printed enclosure you can suspend the tweeter dead centre in front of the Volt (or another driver of your choice) with a the coat hangers as thin scaffolding bent all around the cabinet and screwed / bolted to the rear of the cabinet.
I am currently trying to fit a PHL mid into a cone and mount it front of a 15 inch Beyma 15P80 Fe/N...
If you were to really do it professionally, you could replace the coat hanger with nice copper / stainless steel etc. Or if you really want to go high tech, a polymer printed in a "DNA helix" type structure which is acoustically transparent, ultra strong and non resonant... Also looks cool as hell.... Wilson Audio would write 100 pages of marketing BS, show lots of photos of men in white coats holding microphones and tie the idea up in 10 years of IP / patent pending BS and charge an extra £100K!!
PS Volt plus small dome covering 70Hz to say 10KHz, then super tweeter perched on top of the cabinet, or baffle mounted but off set?
When I first got a DEQX which allowed me to time align driver centers I made a coaxial point source with a Volt 2500.4 and a Morel tweeter ( I cant remember which one, but it came with the Volt as part of a Wilmslow Audio "Vortex" kit) which sounded fantastic but looked.... Like a "Dalek chewing a wasp" as my wife called it! Way back then tweeters were big... Today you can get tiny tweeters / domes and 3D printing "cones" is cheap.
Your 2 or 3 inch domes would be ideal... Mounted in an "ice cream cone " shaped / 3D printed enclosure you can suspend the tweeter dead centre in front of the Volt (or another driver of your choice) with a the coat hangers as thin scaffolding bent all around the cabinet and screwed / bolted to the rear of the cabinet.
I am currently trying to fit a PHL mid into a cone and mount it front of a 15 inch Beyma 15P80 Fe/N...
If you were to really do it professionally, you could replace the coat hanger with nice copper / stainless steel etc. Or if you really want to go high tech, a polymer printed in a "DNA helix" type structure which is acoustically transparent, ultra strong and non resonant... Also looks cool as hell.... Wilson Audio would write 100 pages of marketing BS, show lots of photos of men in white coats holding microphones and tie the idea up in 10 years of IP / patent pending BS and charge an extra £100K!!
PS Volt plus small dome covering 70Hz to say 10KHz, then super tweeter perched on top of the cabinet, or baffle mounted but off set?
Andy makes some good points. It would be possible to get very high performance from a sub+2x8+3+1 system, especially with the right drivers... but a 4-way with sub+2x8+6+3+1 does offer the potential for higher performance.
His point about hard cone/dome drivers being used within their pistonic range is also worth emphasizing. I have found that in order to achieve very high resolution and clarity, I have to use drivers only within their pistonic range.
Yes. This is an area where simulation can help optimize the baffle shape and driver spacing. However, a person can generally get a pretty good result by just making the flat baffle area around the tweeter as small as possible, and using large, blended bevels or radius edges. Copying the baffle shape of something like Revel Salon, Rockport, Magico, or the Vivid Giya would get you close to optimal without the need for a lot of computations. ... And there are people on this forum who are good at the computation part, and they can help you with the baffle shape when the time comes for that. In my case, I find it easier and more enjoyable to build and test a prototype baffle rather than run BEM simulations. I do a lot of simulating at the start of a project, and once I make all my driver measurements, there is a lot more simulations to design the crossover. I spend enough time at a computer screen as it is, so I like the "build prototype and test" aspect.
j.
Well, waveguide are sure a nice thing ( and the ATC mid use it!). But... the idea to mount drivers in front of another bring it's own set of issues ( mainly diffraction).
It can work for sure, Gethain monitors are an example or Fulcrum's coax... but Gethain have almost 50years of trial and errors an D.Gunness implemented some complex FIR treatments for Fulcrum's range to work as intended. Not sure it is availlable to us mere mortals!
I'm with Andy and Jim about pistonic range: it often translate into 2-2,5 octave max range per drivers ( direct radiating).
I must say i don't share the idea wide baffle to be an issue either: as i live with both 'large' width an 'narrow' cabinet at home and spent (way too much) time in front of different flush/inwall monitoring: i'm not convinced narrow is better.
Rendering can be different with pinpoint stereo on smaller width baflle but i have yet to find real world condition where this exist*. And some inwall i heard were as 'good' in this as small standmount, so... i think it depend mainly about how cabinet diffraction is managed. 'Large' baffle can help bringing diffraction in a freq area where it merge with room own behavior. More accurately BSC at or below Schroeder freq...
* during recording session of acoustic instruments using stereo mic only i almost never heard such pinpoint imaging as rendered by some small width nearfields.
My own observations about this are inline with what Troels Gravesen expressed with his experiment with wide baffles ( A Cappela project) inspired by Sonus Faber 'Stradivari'.
http://www.troelsgravesen.dk/Acapella_WB.htm
Last edited:
@krivium - I think a wide baffle is a very viable option, it is just harder to position optimally in most domestic listening rooms. When I simulate wide baffles, like 50+ cm wide, I can get good results. Depending on the builders tool and skills, a wide baffle with a simple rounded edge might be easier to construct than a narrow baffle with highly profiled and sculpted low-diffraction shapes.
Soffit mounted or in-wall infinite baffle designs may be the ultimate in speaker positioning performance, and if not the ultimate, then at least very nearly so. But now we are designing the entire room around the speakers.
The Sonus Faber Strad you mentioned is a perfect example of a well executed wide baffle. I only heard them once, and they sounded so nice.
Another wide baffle speaker that I was quite familiar with back in the 1980s was the Boston Acoustics A400. An affordable speaker (for a large 3-way with 2x8 woofers), I thought it sounded better than anything else in its price range.
So yes, wide baffles can be quite excellent, the only real downside is the challenge in positioning them.
Soffit mounted or in-wall infinite baffle designs may be the ultimate in speaker positioning performance, and if not the ultimate, then at least very nearly so. But now we are designing the entire room around the speakers.
The Sonus Faber Strad you mentioned is a perfect example of a well executed wide baffle. I only heard them once, and they sounded so nice.
Another wide baffle speaker that I was quite familiar with back in the 1980s was the Boston Acoustics A400. An affordable speaker (for a large 3-way with 2x8 woofers), I thought it sounded better than anything else in its price range.
So yes, wide baffles can be quite excellent, the only real downside is the challenge in positioning them.