..The direct sound is a 1-2 feet bubble around the speaker
..this also happens up into the midrange. A directive loudspeaker in a domestic setting up to a certain freq. isn't really directive.
If you really want to achieve your stated goal with respect to the room then you should be going near-field dipole (stereo) up to almost 300 Hz.
..as for the rest of the design with your goals, I'd probably be looking at a wide-mouth Unity/Synergy design with a large absorptive "frame" around the horn's mouth.
Like Paul's Unity S2 prototype here (but with a frame):
Red Spade Audio: DIY point source horn
Note that he sells an elliptical version (..but it's not cheap).
Last edited:
Hi Scott. Good thought.
But, its covered.
The modal range of standing waves is covered by the sub only.
Cause what happens then, around 100Hz and up, is not a standing wave, but early intense reflections between speaker<->first wall OR narrow walls (floor<->cealing 3 meters only)
Dips cant be FIRed, peaks are too slim. So what happens when u put the speaker on the wall? One is gone. Rooms got 6 walls.
So I put the speaker into the corner, 3 are gone, response gets lot better -> use a corner horn
Still not perfect. Cealing is the first n earliest. Lets stack more Khorns then, vertically up. Cylinderwave, cealing gone, only 2 left.
Well now it gets dramatically better. Lets fill the cealing with foam, let the other 2 walls come very late as theyre on the other end of the room, hit the foam on the front part of the room, and gone. Tiny decay, short n weak, no modes left.
Thats why i run the Khorns that unusual high. Which requires some mods in the foldings and phaseplug. Cause K's often get very dirty above 200, some even >100.
cheers
Josh
But, its covered.
The modal range of standing waves is covered by the sub only.
Cause what happens then, around 100Hz and up, is not a standing wave, but early intense reflections between speaker<->first wall OR narrow walls (floor<->cealing 3 meters only)
Dips cant be FIRed, peaks are too slim. So what happens when u put the speaker on the wall? One is gone. Rooms got 6 walls.
So I put the speaker into the corner, 3 are gone, response gets lot better -> use a corner horn
Still not perfect. Cealing is the first n earliest. Lets stack more Khorns then, vertically up. Cylinderwave, cealing gone, only 2 left.
Well now it gets dramatically better. Lets fill the cealing with foam, let the other 2 walls come very late as theyre on the other end of the room, hit the foam on the front part of the room, and gone. Tiny decay, short n weak, no modes left.
Thats why i run the Khorns that unusual high. Which requires some mods in the foldings and phaseplug. Cause K's often get very dirty above 200, some even >100.
cheers
Josh
Hmm. Different people hear differently, but I can't agree, not even a little with that statement. Turn any speaker away from you in any practical domestic room that is as damped as you can -- you will still hear the speaker playing almost as loudly, subjectively, as if it were pointed toward you. Off-axis is always a problem in any listening room (and anechoic chambers don't sound very good or comfortable).
Also, people move their heads quite bit while listening (at least I do) and perception of the source location of a sound and its reflections can be very sensitive. As far as I know, the only way to work around that is with a coaxial arrangement that has (at most) smoothly varying directivity. You can't just consider only the acoustic pressure sources make directly at the listening seat. I suspect that with the system you propose that it wouldn't be hard for a blindfolded listener to point with some accuracy at the location of each of all those drivers with their different positions and shifted radiation patterns.
Well well. Youre saying the right things. Ive been thinking how to solve this.
I cant deny, itll be challenging to compeed with the focus and lobing of a constant directivity point source.
1. what i meant, in a high reflective surrounding the off axis sound is important for the on-axis lis.pos. as well. sure. But in a anechoic chamber it would be completely irrelevant.
But the an. chamber sounds inconvinient. As you say.
So I made a live end - dead end installation.
In my case, that would limit the vert./hor. dispersion to reflective surfaces from normally +-180° down to +-about 50° around the axis. Thats a lot less to focus on for optimisation.
It didnt have the "in head" localization like the an.chamber anymore, but the stage lacks depth and gets very close. Thats where i have small 1x1 meter diffuser behind each main with the slot tweeter making the diffuse stage for depth. And that one's completely steerable on its own, in level + time + response. Like a "shape your reflection" toy.
2.
It still leaves the +-50° angles around the axis that passes my listening.pos. and echo behind me from the "live end". But small angle is less critical. First point is the higher the frequency the more lobing. Thats why i use the steep lowpass (72dB/Oct. does not ring but only delay) to limit the lobing. The highpass complement would normally ring when steep but not delay. Thanks to subtractive xover i can xchange that into delay without ringing, and still get my transient perfect. Its just not so steep anymore (22dB/Oct), requires strong drivers, well as you see i got heavy stuff, crossed high, theyll manage with ease!
3. The offset ... vertical jumping phantom sources. I made some ideas and scribbles. Always on paper im not good in CAD yet. So, next post, explaining the look and why.
I cant deny, itll be challenging to compeed with the focus and lobing of a constant directivity point source.
1. what i meant, in a high reflective surrounding the off axis sound is important for the on-axis lis.pos. as well. sure. But in a anechoic chamber it would be completely irrelevant.
But the an. chamber sounds inconvinient. As you say.
So I made a live end - dead end installation.
In my case, that would limit the vert./hor. dispersion to reflective surfaces from normally +-180° down to +-about 50° around the axis. Thats a lot less to focus on for optimisation.
It didnt have the "in head" localization like the an.chamber anymore, but the stage lacks depth and gets very close. Thats where i have small 1x1 meter diffuser behind each main with the slot tweeter making the diffuse stage for depth. And that one's completely steerable on its own, in level + time + response. Like a "shape your reflection" toy.
2.
It still leaves the +-50° angles around the axis that passes my listening.pos. and echo behind me from the "live end". But small angle is less critical. First point is the higher the frequency the more lobing. Thats why i use the steep lowpass (72dB/Oct. does not ring but only delay) to limit the lobing. The highpass complement would normally ring when steep but not delay. Thanks to subtractive xover i can xchange that into delay without ringing, and still get my transient perfect. Its just not so steep anymore (22dB/Oct), requires strong drivers, well as you see i got heavy stuff, crossed high, theyll manage with ease!
3. The offset ... vertical jumping phantom sources. I made some ideas and scribbles. Always on paper
im not good in CAD yet. So, next post, explaining the look and why.If you've actually measured it under the conditions you've stated (with the results you've mentioned): then
But if it's just a thought/modeled exercise, then I wouldn't bet on it.
In any event, a good nearfield stereo dipole is still the superior approach for a range of reasons - the most important is if you get the null just right and have a quality "shearing" effect across your H/torsoRTF.
If you've actually measured it under the conditions you've stated (with the results you've mentioned): then
Im not a dreamer
Heres my single K nearfield in mouthno gate, no smoothing, FIR filter applied (!), after first mods (tbc)
??? xplain more simple pls
Have a look at this link. 20 audio outputs. There are plenty more like it a various price levels.
Focusrite Clarett 8Pre 18x20 Thunderbolt Audio Interface | Sweetwater
Focusrite Clarett 8Pre 18x20 Thunderbolt Audio Interface | Sweetwater
I don't quite get that, unless you are saying to counteract room modes you pretty much need your head in the speaker? Also, are you saying only a slotted OB works well this close?@scott, if u got modes, its not close enough. Several meters are far field. U only listen to modes. The direct sound is a 1-2 feet bubble around the speaker
Try to put it behind ur head and ull notice why i choose a slim slot.
Yes, in your head, youre right. Thats annoying.
First page, drawing of my sofa, there you see the solution.
Im not a dreamer
no gate, no smoothing, FIR filter applied (!), after first mods (tbc)
??? xplain more simple pls
for me this year 2017 speaker desing is all heresy
Four Audio | PPA-Serie
video 0:30-1:14min
about the vertical offset issue....
So. I made a test with two mini speakers, both fullrange, stereo-triangle very small.
1. stereo horizontal triangle 30cm, fullrange: localization in the middle. Of course. No surprise
2. same as (1) but vertical. I noticed the phantom source gets sharper.
3. same as (1) but changing the distance. The closer i get (10cm) the blurrier it gets, the further away the sharper the center
4. Horizontal triangle 30cm, left speaker low / right high-passed.
Perception is kind of weird. A bit diffuse. But mostly it comes from the right side.
5. like (4) but vertical, gets a lot sharper. less blurr. closer worse, further (1 meter) its like a point source sharp on the tweeter.
My conclusion, nothing new but nice to see live, vertical offset is half that bad, and especially on long distance. The brain cant catch the fake. The focus should be on high freq. drivers (getting them close).
I heard big horns like Avantgarde etc which have small issues with precise localization vertically (jumping up n down). Thats too much then.
But smaller ones, like the usual standing speaker with 1 meter overall height of sources, but half meter for mid-high, are fine. We all know such.
But heres an interesting speaker i heard, huge offset, BUT the sounds all sharp centered on the tweeter:
And why is that?
Instead of designing bottom up - low mid high...
They designed low high mid. In principle. More relevant there with so many XO.
Well if u think of having the center between low-mid where the tweeter is...
Where else normally you build bottom up the center BELOW the tweeter...
Inspired me
So, heres a proportional correct drawing of the components in place. 1 meter width, >2m high:
The bottom horn is still to be built. Others are resting in the basement.
Tweeter is about 1 meter heigh from the floor.
You see, the mid-high spectrum is mostly within a 70cm range. Not that bad.
So. I made a test with two mini speakers, both fullrange, stereo-triangle very small.
1. stereo horizontal triangle 30cm, fullrange: localization in the middle. Of course. No surprise
2. same as (1) but vertical. I noticed the phantom source gets sharper.
3. same as (1) but changing the distance. The closer i get (10cm) the blurrier it gets, the further away the sharper the center
4. Horizontal triangle 30cm, left speaker low / right high-passed.
Perception is kind of weird. A bit diffuse. But mostly it comes from the right side.
5. like (4) but vertical, gets a lot sharper. less blurr. closer worse, further (1 meter) its like a point source sharp on the tweeter.
My conclusion, nothing new but nice to see live, vertical offset is half that bad, and especially on long distance. The brain cant catch the fake. The focus should be on high freq. drivers (getting them close).
I heard big horns like Avantgarde etc which have small issues with precise localization vertically (jumping up n down). Thats too much then.
But smaller ones, like the usual standing speaker with 1 meter overall height of sources, but half meter for mid-high, are fine. We all know such.
But heres an interesting speaker i heard, huge offset, BUT the sounds all sharp centered on the tweeter:
An externally hosted image should be here but it was not working when we last tested it.
And why is that?
Instead of designing bottom up - low mid high...
They designed low high mid. In principle. More relevant there with so many XO.
Well if u think of having the center between low-mid where the tweeter is...
Where else normally you build bottom up the center BELOW the tweeter...
Inspired me
So, heres a proportional correct drawing of the components in place. 1 meter width, >2m high:
The bottom horn is still to be built. Others are resting in the basement.
Tweeter is about 1 meter heigh from the floor.
You see, the mid-high spectrum is mostly within a 70cm range. Not that bad.
No. Processor is bored, 9% load. That FIR stuff is primitive! Like elementary school math for grown up. Modern Intel is bored.
It all runs normal. My vinyl, my tape, my DAC, into the analogue preamp with a analogue volume knob, stereo XLR out into the "fir construction" which gives 16 outputs into the amps. Easy.
That "FIR construction" just runs, no need to ever touch it.
Member
Joined 2009
Paid Member
I didn't think it was possible to even consider the ultimate horn speaker without first making a pilgrimage to this site: GoodSoundClub - Romy the Cat's Audio Site
Attachments
Im not a dreamer
..??? xplain more simple pls
Nope. We were discussing room effects, so your measuring would need to be at multiple positions throughout the room, transitioning from nearfield to multiple positions farfield (..a gradient vertically and horizontally).
We are most strongly effected by pressure differences as they move from left ear to right ear (and vice versus), but as freq.s lower this becomes more difficult so we increasingly use phase to compensate. However, if you can transition the pressure loss at lower freq.s that has that distinct pressure difference between ears (..and not *too* much), then the recorded sound-field tends to "expand".
Or rather, forget that and try it out!
(..rotating the drivers until you hear a better sound-field, if you do - some people actually have learned listening behavior that is counter-productive.)
Last edited:
caT
Agreed. Some really good reading there. This is as good a place to start as any:
Macondo Acoustic System
Romy the Cat's Macondo Acoustic System
p.s. I have posted several times on his site. He has always treated me with respect. Others may have not been so fortunate !
Agreed. Some really good reading there. This is as good a place to start as any:
Macondo Acoustic System
Romy the Cat's Macondo Acoustic System
p.s. I have posted several times on his site. He has always treated me with respect. Others may have not been so fortunate !
Ill try. For sure. And then modify. First thoughts are direction to head to, then see where to optimize.
Whats the alternative? Downgrading to standard setup? What revolution u can expect then?
I can still take 1 driver out, move around, use steep NevilleThiele filter, etc.
Unless you have a stecific proposal ??
Ive been to Romy's side. Interesting stuff. I share lot of his opinions, not all.
Discovered his nice explanation why my resonant chambers (top horn on the drawing) sounds so nice but measures horrible.
But I'm also more familiar with modern stuff such as measuring and the correct psycho-acoustic interpretation of such (its long path to walk). Also correct use of FIR filter. Stressing correct. Because u can do a lot of things wrong being naive.
Once u understand that, theres no way back to analogue crossover or EQ.
Same as understanding non-linear distortion, theres no way back from multi-way horns.
Reg. the pilgrimage. I had several inspirations of speakers that are absolute "wow" to hear live, and i can only recommend to listen to:
Klipsch horn for its fundamentals
Vox Olympian - for its resonant chamber and tweeters
Fostex Laboratory Monitors - for its depth of stage and tweeters
Electrostatics for understanding how to use rear-reflection for advantage (stage depth)
Western Electric Mirrophonic - for the joy and ease
So. im trying to pick the best of each, and merge it with new ideas.
Lets see. Still hoping someone will say "better make this that way because..."
Whats the alternative? Downgrading to standard setup? What revolution u can expect then?
I can still take 1 driver out, move around, use steep NevilleThiele filter, etc.
Unless you have a stecific proposal ??
Ive been to Romy's side. Interesting stuff. I share lot of his opinions, not all.
Discovered his nice explanation why my resonant chambers (top horn on the drawing) sounds so nice but measures horrible.
But I'm also more familiar with modern stuff such as measuring and the correct psycho-acoustic interpretation of such (its long path to walk). Also correct use of FIR filter. Stressing correct. Because u can do a lot of things wrong being naive.
Once u understand that, theres no way back to analogue crossover or EQ.
Same as understanding non-linear distortion, theres no way back from multi-way horns.
Reg. the pilgrimage. I had several inspirations of speakers that are absolute "wow" to hear live, and i can only recommend to listen to:
Klipsch horn for its fundamentals
Vox Olympian - for its resonant chamber and tweeters
Fostex Laboratory Monitors - for its depth of stage and tweeters
Electrostatics for understanding how to use rear-reflection for advantage (stage depth)
Western Electric Mirrophonic - for the joy and ease
So. im trying to pick the best of each, and merge it with new ideas.
Lets see. Still hoping someone will say "better make this that way because..."
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.