Then my guess is that:
1. you have a tweeter installed, and it's operating pretty low in freq.
2. the fullranges on the side are "cut" pretty low in freq..
-really what I want to see is just the dispersion pattern for one full range driver (either left or right) measured from the loudspeaker's frontal 0 degree axis to 90 degrees off-axis (or in other words the driver's 0 degree axis). No tweeter, no other driver contribution. No eq..
In other words starting from a "base" to solve the problem.
Why would that be relevant?
Well, the guy that gave me the box wore a Dastar - how could he be wrong?
Or this would be correct only if you had a very big head. Normally 17 cm is the average value.
Ear-drum to ear-drum is only 14cm whereas the effective distance is greater than 17cm. But is this really relevant?
OTOH, maybe you don't want to mimic the stereolith.
The stereosphere™ will be truly unique, designed from the ground up to "make the dream come true™". Did I already mention that?
These should work well for this purpose:
IKEA | Serving pieces | Serving dishes & bowls | BLANDA MATT | Serving bowl
IKEA | Serving pieces | Serving dishes & bowls | BLANDA MATT | Serving bowl
it is called stereosphere™ and it is much better than stereolith
don't listen to them! go Markus! go! go! go!!! make the dream come true™!
That's the spirit!™
Hi Scott,
Why do you think direct sound have much (any) importance in this principle? 90 degrees off axis for the driver will be direct sound for the speaker at the listener position. In parallel with the another active thread I would like to express my view: 'flat' is not needed for stereo system
Aargh, No! In this case we don't want to minimize diffraction. The bigger the baffle the bigger the step, or the bigger the baffle the bigger the left right separation, which is good for the purpose. I think ideally this speaker should be built into an Ambiophonics baffle
- Elias
Why do you think direct sound have much (any) importance in this principle? 90 degrees off axis for the driver will be direct sound for the speaker at the listener position. In parallel with the another active thread I would like to express my view: 'flat' is not needed for stereo system
Aargh, No!
In this case we don't want to minimize diffraction. The bigger the baffle the bigger the step, or the bigger the baffle the bigger the left right separation, which is good for the purpose. I think ideally this speaker should be built into an Ambiophonics baffle - Elias
Last edited:
Hello Radugazon,
I think it's best in the case of this loudspeaker to look at binaural impulse responses measured with a dummy head inside the room and at the listening position to really see how it works. You already demonstrated what will happen in case both channels (L and R) are contributing to the sound with phase or delay panning with the Lissajous scope figures. Indeed, phase/delay panning tilts the propagating wavefront at the listening position which is perceived as ITD. This is a direct sound phenomena.
BUT, what will happen if only one channel is executing a sound and another channel is mute? This is equivalent to hard amplitude stereo panning. Normal stereo has sound coming from one loudspeaker. What will happen with this loudspeaker? Direct sound is straight in front. Now we need those binaurals to see ..
- Elias
I think it's best in the case of this loudspeaker to look at binaural impulse responses measured with a dummy head inside the room and at the listening position to really see how it works. You already demonstrated what will happen in case both channels (L and R) are contributing to the sound with phase or delay panning with the Lissajous scope figures. Indeed, phase/delay panning tilts the propagating wavefront at the listening position which is perceived as ITD. This is a direct sound phenomena.
BUT, what will happen if only one channel is executing a sound and another channel is mute? This is equivalent to hard amplitude stereo panning. Normal stereo has sound coming from one loudspeaker. What will happen with this loudspeaker? Direct sound is straight in front. Now we need those binaurals to see ..
- Elias
Hi Markus,
Scott already gave you a lot of very relevant indications.
One of them was the width of the baffle, or better said, the inter acoustic center distance.
It's here that the original stereolith gets very special:
inter acoustic center distance = inter aural distance
This seems to be a detail but is the core of the design and has unexpected consequences when you analyze it, both drivers running and both ears listening.
Then, it was a good idea to use Swiss originated cardboard, but your width is far excessive.
Or this would be correct only if you had a very big head
OTOH, maybe you don't want to mimic the stereolith.
Hi,
You measured it without listening first!?
You don't happen to have a dummy head?
- Elias
You measured it without listening first!?
You don't happen to have a dummy head?
- Elias
Hi Scott,
Why do you think direct sound have much (any) importance in this principle? 90 degrees off axis for the driver will be direct sound for the speaker at the listener position. In parallel with the another active thread I would like to express my view: 'flat' is not needed for stereo system
Aargh, No!
- Elias
It depends on the bandwidth. It should be relatively flat to *almost* 2 kHz (..at what is 90 degrees of of the *driver* axis).
Above that you are correct - flat isn't needed or desirable for this type of system. Just what IS required however, is debatable and probably personal. Certainly to obtain better image positioning will require something moderately "near" the average for any given *fundamental* (..and likely close harmonic).
Nope.. a bigger baffle is wrong. Look a little more closely at the stereolith.
For this design you specifically do NOT want to have separation between L & R much below 2kHz. Moreover, IF possible, you do NOT want a secondary diffraction "peak" causing another "locational cue" in the 1-2 kHz range (..or really anywhere near it).
What I've suggested is the starting point for something very similar, yet *better*.
I'm talking about the side baffle. I hope there is no misconception which baffle I'm talking about. The front baffle facing the listener should be as it is. Bigger side baffle would increase left right separation in the room, but I think they needed to optimise their own cost in manufacturing
- Elias
- Elias
Nope.. a bigger baffle is wrong. Look a little more closely at the stereolith.
For this design you specifically do NOT want to have separation between L & R much below 2kHz. Moreover, IF possible, you do NOT want a secondary diffraction "peak" causing another "locational cue" in the 1-2 kHz range (..or really anywhere near it).
What I've suggested is the starting point for something very similar, yet *better*.
I'm talking about the side baffle. I hope there is no misconception which baffle I'm talking about. The front baffle facing the listener should be as it is. Bigger side baffle would increase left right separation in the room, but I think they needed to optimise their own cost in manufacturing
- Elias
yes, "baffle" is relative to driver. The side panel(s) of the loudspeaker.
It has nothing (or nearly nothing) to do with cost (with respect to its size). It has everything to do with cost with respect to its *shape*.
You can make a larger baffle as Radugazon has done, but it's less effective to this design. You can also use larger diameter drivers (as he has also done), but again - its less effective to this design.
The real *price* is limited low freq. extension, dynamics, etc..
Last edited:
Maybe you are not thinking about the side wall reflections, or yes? The size of the side baffle has a relation there. I think side wall reflections are the key of the operation of it's principle.
The Radugazon's system.. It is not ordinary In his wonderfull presentation can see the room size which is huge. There must be a correct portion of temporal aspects of direct sound and first lateral reflections in order the summing localisation to work. I'm afraid in too large space it just don't happen, or the time windows don't match.. The train leaves before the passengers arrive.
- Elias
The Radugazon's system.. It is not ordinary
In his wonderfull presentation can see the room size which is huge. There must be a correct portion of temporal aspects of direct sound and first lateral reflections in order the summing localisation to work. I'm afraid in too large space it just don't happen, or the time windows don't match.. The train leaves before the passengers arrive. - Elias
Posts Avalanche !!!
Markus, if the brothers Wright have been listening to all the good advices "heavier than air? this stuff will never flight..." you see what I mean. So, keep alive your pioneer mind and go go go !!!
Anyway, it's good to look at the other experiences. Walter S. claims that the width has to be 17 cm. Is it only to reproduce the sound field as recorded by a standard dummy head ? I guess this is only the obvious brochure argumentation.
Could it have something else ?
Markus, if the brothers Wright have been listening to all the good advices "heavier than air? this stuff will never flight..." you see what I mean. So, keep alive your pioneer mind and go go go !!!
Anyway, it's good to look at the other experiences. Walter S. claims that the width has to be 17 cm. Is it only to reproduce the sound field as recorded by a standard dummy head ? I guess this is only the obvious brochure argumentation.
Could it have something else ?
Maybe you are not thinking about the side wall reflections, or yes? The size of the side baffle has a relation there. I think side wall reflections are the key of the operation of it's principle.
The Radugazon's system.. It is not ordinary
- Elias
Side wall reflections *can* add or detract, but the basic principal doesn't rely on it. As I suggested before, direct sound dominates. You can use a system like this outside or even in a anechoic chamber and the effect is still there - "images" don't "collapse" to the loudspeaker. (..I've mentioned this, Graaf has, and Radugazon has.) Of course any room reflections can add or detract generally as well, and yes he might have some problems due to it's large size.
Note: Radugazon has at least two systems that I'm aware of. When I referenced his loudspeakers I was referring to his testing of the stereolith loudspeaker, which is excellent (..and absolutely superb build for what he considers a mock-up) - but could be better in some respects to a stereolith-type design. (..of course it IS better with respect to low freq. extension, dynamics, etc..)
Last edited:
It's easy to make flawned simplistic theory, so let's go . BTW, Elias and Scott had put a finger on it's direction.
If we forget the side reflections (and the speaker works without them too), what's happening when only one driver is running?
On the connected side, ok, NP, but on the other side.... we have a passive radiator...... that's moving a lot, excited by the waves/pressure of it's brother. They move then "as" a dipole. ("as" only...).
Measuring on it's nearfield, the passive radiator is @-10dB from the active till 1000 Hz, and then falling abruptly. Can see the plots.
Notice that one driver active vs 2 drivers active is the same curve, except in the lower range under 200 Hz. Logic.
On the step plots, normal inversion of the passive driver. The step is very sharp for a passive radiator...
All this makes me think that what's measured is the sound waves of the active driver passing through the thin cone of the passive one.
----------------------------------------------
Now, with these datas, when both speakers are running, we can suppose that the same fraction of the controlateral channel is crossing the cone, kind of -10 dB till 1000Hz, not something that can be ignored.
When this derived signal arrives to the open air, it's after a delay fixed by the space between our drivers.
Thanks to Walter S., this delay is 17 cm or 0.49 ms, conveniently said our inter aural delay.
------------------------------------------------
So, as you are experienced guys you have seen in a flash the big allegory in the sky:
a signal < 1000 Hz, on the other side after THE delay
-/*/-
If we forget the side reflections (and the speaker works without them too), what's happening when only one driver is running?
On the connected side, ok, NP, but on the other side.... we have a passive radiator...
... that's moving a lot, excited by the waves/pressure of it's brother. They move then "as" a dipole. ("as" only...).Measuring on it's nearfield, the passive radiator is @-10dB from the active till 1000 Hz, and then falling abruptly. Can see the plots.
Notice that one driver active vs 2 drivers active is the same curve, except in the lower range under 200 Hz. Logic.
On the step plots, normal inversion of the passive driver. The step is very sharp for a passive radiator...
All this makes me think that what's measured is the sound waves of the active driver passing through the thin cone of the passive one.
----------------------------------------------
Now, with these datas, when both speakers are running, we can suppose that the same fraction of the controlateral channel is crossing the cone, kind of -10 dB till 1000Hz, not something that can be ignored.
When this derived signal arrives to the open air, it's after a delay fixed by the space between our drivers.
Thanks to Walter S., this delay is 17 cm or 0.49 ms, conveniently said our inter aural delay.
------------------------------------------------
So, as you are experienced guys you have seen in a flash the big allegory in the sky:
a signal < 1000 Hz, on the other side after THE delay Me too till I realized that this cheerful signal was now in phase and then no more free XTC 
---------------------------------------------------
Anyway I have since a few days a very iconoclastic operational solution under test, it works almost perfectly, but it's impossible to vulgarize it, I wouln't like to pervert the researches of Markus our sympathic new challenger.
Hey, I'm just teasing, hobby has to be fun !
plots : blue = both drivers; red = one driver; orange = passive
---------------------------------------------------
Anyway I have since a few days a very iconoclastic operational solution under test, it works almost perfectly, but it's impossible to vulgarize it, I wouln't like to pervert the researches of Markus our sympathic new challenger.
Hey, I'm just teasing, hobby has to be fun !
plots : blue = both drivers; red = one driver; orange = passive
Attachments
or this:
http://www.diyaudio.com/forums/multi-way/121385-loudspeakers-room-system.html#post1485807
it's a flooder but thinkbad tried them back-to-back too:
http://www.diyaudio.com/forums/multi-way/121385-loudspeakers-room-system.html#post1485801
best,
graaf
isn't our maximum ITD more like 0.63 ms which corresponds to around 22-23 cm?
21,6 cm from left entry of ear canal to right entry of ear canal to be exact. But why would this be relevant? The signals arriving at our ears won't be delayed by that amount just because two drivers are mounted to the left and the right of a 21,6 cm wide box.
By the way, the interchannel time differences in stereophonic recordings are even longer than 0,63 ms.