Hi Juhazi
Well i look back to this threat with a smile. It was a bit naive
Some things didn’t work out as i planned:
-i tried too much to combine exciting drivers with their own character together hoping to get all benefits in one package but it got a mishmash instead - i very much need to admit the earlier posters were right, it has too many ways. Not because of offset or lobing like i thought! A completely new problem came in: using a 7-8-9 way system involves narrow passbands. And the closer lowpass and hipass meet, the longer the bandpass starts resonating. Especially when the passpand is only 1 octave. It would need to be precisely canceled put by linearphase matching xover. Which for such long reverbant tales only work sufficiently in anchoic conditions... - the 700-1500hz channel esp with that horn was totally nuts
Later then i visited Mr. Klaus S. from Germany - owning Goto Horns. It totally blew me away in such a shock that i had to unfortunately completely re-think my way on how i approached building speakers so far. It made me loose interest in this specific concept, and until now i did not make a new masterplan. However, Goto is out of my league. I need to think of a solution...
For the time being i want to listen to music. I was sawing sweating soldering sweeping calculating for 2 years and almost didn’t listen to music. I need a break! For so some months I will be resting with a ready made speaker from the shop.
I will certainly come back to this in a revised way. Also I decided to completely rebuild and extend my listening room first. I desire way more space, more damping, and nicer design.
Cu soon!
Cheers Josh
Well i look back to this threat with a smile. It was a bit naive
Some things didn’t work out as i planned: -i tried too much to combine exciting drivers with their own character together hoping to get all benefits in one package but it got a mishmash instead - i very much need to admit the earlier posters were right, it has too many ways. Not because of offset or lobing like i thought! A completely new problem came in: using a 7-8-9 way system involves narrow passbands. And the closer lowpass and hipass meet, the longer the bandpass starts resonating. Especially when the passpand is only 1 octave. It would need to be precisely canceled put by linearphase matching xover. Which for such long reverbant tales only work sufficiently in anchoic conditions... - the 700-1500hz channel esp with that horn was totally nuts
Later then i visited Mr. Klaus S. from Germany - owning Goto Horns. It totally blew me away in such a shock that i had to unfortunately completely re-think my way on how i approached building speakers so far. It made me loose interest in this specific concept, and until now i did not make a new masterplan. However, Goto is out of my league. I need to think of a solution...
For the time being i want to listen to music. I was sawing sweating soldering sweeping calculating for 2 years and almost didn’t listen to music. I need a break! For so some months I will be resting with a ready made speaker from the shop.
I will certainly come back to this in a revised way. Also I decided to completely rebuild and extend my listening room first. I desire way more space, more damping, and nicer design.
Cu soon!
Cheers Josh
Thank you for update and conclusions! Sounds like a 4-way is ideal solution for a hifi-speaker in a living room. Most practical would be a 3-way. With horns this gets difficult or impossible to achieve in low frequencies.
The challenge is always the wide span of wavelengths and spl to be handled harmoniously. We must leave space for crossings, sharp filters like those used in PA systems will separate sound sources too much at short distance. Shallow slopes lead to a mess of interferences specially off-axis, and the nearness of boundaries makes these so bad. Even a horn doesn't give enough directivity.
One thing to consider is, how high directivity is good. I dont want to have a control room or headphone-like intensity and nearness of sound with loudspeakers. Neither rock club -like high spl levels with bass hits on my chest and cymbals making my ears bleed. At home I mean, but of course someone else is haunting for just that. When I want that, I go to a concert!
Stereophonic sound reproduction is about making an illusion. The preferred illusion is an individual choice, and we can play with loudspeaker type and room acoustics to achieve that, up to some degree. Another way to go is an anechoich chamber and multiway sound system with dsp, adjustable delays etc.. Still, published and distributed music comes mosty as 2-ch stereo.
Headphone listening is a different world, that I don't like at all.
The challenge is always the wide span of wavelengths and spl to be handled harmoniously. We must leave space for crossings, sharp filters like those used in PA systems will separate sound sources too much at short distance. Shallow slopes lead to a mess of interferences specially off-axis, and the nearness of boundaries makes these so bad. Even a horn doesn't give enough directivity.
One thing to consider is, how high directivity is good. I dont want to have a control room or headphone-like intensity and nearness of sound with loudspeakers. Neither rock club -like high spl levels with bass hits on my chest and cymbals making my ears bleed. At home I mean, but of course someone else is haunting for just that. When I want that, I go to a concert!
Stereophonic sound reproduction is about making an illusion. The preferred illusion is an individual choice, and we can play with loudspeaker type and room acoustics to achieve that, up to some degree. Another way to go is an anechoich chamber and multiway sound system with dsp, adjustable delays etc.. Still, published and distributed music comes mosty as 2-ch stereo.
Headphone listening is a different world, that I don't like at all.
Last edited:
Hm im rather not going down that radical to 3-4 ways.
Maybe depends on the circumstance. Most people cover 8 octaves. Then 4 ways may suit well. If you have limited ressource (space, money etc) even 3 ways can cover 8 octaves pretty good. Depends which compromise you choose.
If you compromise your living room and wallet - one compromise we all pick - you may want to hear all 9 octaves. And maybe even cover 10 octaves not to have the rolloff audible, and its phase distortion.
What counts are the width of the passbands. I experienced no problem down to 2 octaves per way before it started ringing. That goes well along with many who say horns play best for 2-3 octaves.
I have a 4 way horn running right now. And I compromise on crossover and bandwidth. But its smaller, cheaper, simple Its 15" in khorn, 8" in fundamental horn, 2" comp.dr. radial horn, slot-tweeter. For the future I see a 5 way suiting best to me. There can be more channels (rear tweeter, multiple subs etc) but max 5 ways xover.
I totally agree with what you say about avoiding too dry room acoustics and headphone alike sound. I dont like either. I just said that such utopia concepts run better in an anchoic room. Real world needs more "forgiveness" (like with everything).
cheers
josh
Maybe depends on the circumstance. Most people cover 8 octaves. Then 4 ways may suit well. If you have limited ressource (space, money etc) even 3 ways can cover 8 octaves pretty good. Depends which compromise you choose.
If you compromise your living room and wallet - one compromise we all pick - you may want to hear all 9 octaves. And maybe even cover 10 octaves not to have the rolloff audible, and its phase distortion.
What counts are the width of the passbands. I experienced no problem down to 2 octaves per way before it started ringing. That goes well along with many who say horns play best for 2-3 octaves.
I have a 4 way horn running right now. And I compromise on crossover and bandwidth. But its smaller, cheaper, simple
Its 15" in khorn, 8" in fundamental horn, 2" comp.dr. radial horn, slot-tweeter. For the future I see a 5 way suiting best to me. There can be more channels (rear tweeter, multiple subs etc) but max 5 ways xover.I totally agree with what you say about avoiding too dry room acoustics and headphone alike sound. I dont like either. I just said that such utopia concepts run better in an anchoic room. Real world needs more "forgiveness" (like with everything).
cheers
josh
I am getting in late to this thread but was wondering if you saw this site - check out the speaker the IMPERIA - 4 way horn system:
Oswalds Mill Audio | OMA
He claims to have solved the horn issues...... I have never heard them but would like to based on the good reviews. However many have promised much before as most of us know on this site.
Oswalds Mill Audio | OMA
He claims to have solved the horn issues...... I have never heard them but would like to based on the good reviews. However many have promised much before as most of us know on this site.
I spent years and 4 different drivers trying to get a woofer that had a matching "voice" for my horn until I finally hit on a winner. And that was for a 2-way! I can't imagine 5 way. Drivers with very similar response curves sound completely different. I know I'm in left field, but the coherence of my two way seems to far outweigh the negatives. Unfortunately a two way with horns that's flat and full range is extremely difficult with passive xo. I'm in the active camp. You've probably already done a simpler build? If not, don't discount them. I've heard a lot of speakers, and complexity has its disadvantages...it seems to often corrupt the delicate touch of the music and bring a sterility. My 2 cents, Craig
My 2way:
the "SPIEKER" my almost 20 year prototype finally done
What driver in K-Horn ?Passive or active? Care to draw crossover with types of CD drivers (impedances)? Re-discovering a wheel is not my strong suit
Hi Limono
Its Fostex PA38 in Khorn, crossed 200Hz.
Then Fostex FP203 in mini-Altec 816- alike horn, crossed 1000Hz.
Then Fostex FD200L in Fostex H220 horn, crossed 6000Hz.
Finally Fostex T825.
All crossover Active analogue Butterworth 2nd order. Together with 8 Amps in 1 case, complete DIY, Class AA, á 50 Watts, TIP SS.
Difficulties in this concept are Khorn below 50 Hz as well as the 200-XO having little overlap.
cheers
Josh
Dear cspieker
Back when transient response and phase where a miracle to me, I also had more success with less ways.
Once I discovered the power of DSP and learn to match things, 4-5 ways beat the simple ones easily by far.
I must admit, it took me years of try-hard studies though. Only for understanding crossovers.
Back when transient response and phase where a miracle to me, I also had more success with less ways.
Once I discovered the power of DSP and learn to match things, 4-5 ways beat the simple ones easily by far.
I must admit, it took me years of try-hard studies though. Only for understanding crossovers.
Hi! I agree with Patrick.
OMA is a pure lifestyle and design object. It is great in many aspects, but certainly not an engineering piece that addresses any problems rather than creates new. However, if you desire a piece of art in your loft, it may be the right thing to buy.
No offense, I am designer and like them a lot. It inspires and matches the current spirit of the time, almost setting a benchmark visually. But technically the concept is almost funny!
cheers
Josh
Don't know how I missed this thread....I was hoping to read a success story using steep FIR filters.
Have you heard more recent designs using wide bandwidth beryllium compression drivers on large waveguides like the Seos 30?
I would think running one conservatively from 1k on up with steep FIR filters would keep all objectionable distortions to a mimumum. Maybe add a nested super-tweeter in foam if desired/needed.
Have you heard more recent designs using wide bandwidth beryllium compression drivers on large waveguides like the Seos 30?
I would think running one conservatively from 1k on up with steep FIR filters would keep all objectionable distortions to a mimumum. Maybe add a nested super-tweeter in foam if desired/needed.
Last edited:
Hi Ernie
The problem is that steep filter ring. A brickwall e.g. can ring a second. And your guitar sounds like a bell.
Linearphase from FIR may change the phase of the ringing, but the duration is inevitable linked to steepness.
The steepest that is kind of ok to use in reality imho is Neville Thiele 2nd order.
It looks like a soft Bessel that exponentially transforms into a Brickwall.
see page 11-12 here:
Acourate crossovers
cheers
Josh
The problem is that steep filter ring. A brickwall e.g. can ring a second. And your guitar sounds like a bell.
Linearphase from FIR may change the phase of the ringing, but the duration is inevitable linked to steepness.
The steepest that is kind of ok to use in reality imho is Neville Thiele 2nd order.
It looks like a soft Bessel that exponentially transforms into a Brickwall.
see page 11-12 here:
Acourate crossovers
cheers
Josh
As CAD progresses, a sneak peak of my so called "vox phidias"
Your drawing reminded me of a once well-known acoustics company:
http://i39.tinypic.com/wuh09h.jpg
http://i44.tinypic.com/vnijwh.jpg
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.