About the size of (HF) image, I also had some observations.
A triangle (again), can be one of the following situations (and between):
1. Very gently tapped (or even brushed?) -- the image would be fuzzy, like a smokey cloud without clear contour, quite large in a hard-to-define size.
2. Moderately hit -- there'd be bright tiny solid 'core' showing up on that 'cloud'. The first impression would be a very tiny image, but the 'cloud' does exist and firmly attached to the 'core'.
3. Hit hard -- the bright solid core grows in size considerably, to a degree that can be slightly larger than other physically bigger instruments. Also it seems stepping forward somewhat. At this moment the 'cloud' is largely supressed and not as easily identified as other situations.
I think the HF transient makes the image small still apply, as the situation 2 can have a higher attact/sustain ratio than 3 in these cases.
However, on the other hand, sustaining sounds (without obvious attack) can have very different image sizes, too.
For example, a flute (with very much HF portion) can sound very big -- as big as my DML panel in a recording I found. Considering the distance, that view angle is just too exaggerated. But it's not the same as the 'cloud with a core' model as the triangle, the sound of a flute is a whole, the texture is very uniform within the area. In other recordings, it can be a smaller size but still bigger than the physical (visual) impression.
Another example is trumpet. It's usually arranged at the further part of the orchestra, and almost always sounding solid and small (very loud, though).
Quite complicated in the acoustics of musical instruments. Or maybe I mixed up too many factors here...
A triangle (again), can be one of the following situations (and between):
1. Very gently tapped (or even brushed?) -- the image would be fuzzy, like a smokey cloud without clear contour, quite large in a hard-to-define size.
2. Moderately hit -- there'd be bright tiny solid 'core' showing up on that 'cloud'. The first impression would be a very tiny image, but the 'cloud' does exist and firmly attached to the 'core'.
3. Hit hard -- the bright solid core grows in size considerably, to a degree that can be slightly larger than other physically bigger instruments. Also it seems stepping forward somewhat. At this moment the 'cloud' is largely supressed and not as easily identified as other situations.
I think the HF transient makes the image small still apply, as the situation 2 can have a higher attact/sustain ratio than 3 in these cases.
However, on the other hand, sustaining sounds (without obvious attack) can have very different image sizes, too.
For example, a flute (with very much HF portion) can sound very big -- as big as my DML panel in a recording I found. Considering the distance, that view angle is just too exaggerated. But it's not the same as the 'cloud with a core' model as the triangle, the sound of a flute is a whole, the texture is very uniform within the area. In other recordings, it can be a smaller size but still bigger than the physical (visual) impression.
Another example is trumpet. It's usually arranged at the further part of the orchestra, and almost always sounding solid and small (very loud, though).
Quite complicated in the acoustics of musical instruments. Or maybe I mixed up too many factors here...
to Graaf,
I can't find a convenient postion to elevate the flooder tweeters. The available space around that area is indeed limited.
On the equipment rack, I think it'd be too close to the (black) mini-monitor and the edge of DML panel.
I've briefly tried putting them on the central sub. Sitting slightly lower than normal, my line of sight would be about the same height with WG's up-firing mouth. And the WG's were pushed against the TV and emitting toward the DML panel. Maybe because of this, there'd be some interferences... I'm not sure.
Compared with the normal setup, the overall presentation is somewhat congested. Narrower, shallower, smaller overall. I don't like it.
Ah, I ran out of space for options. The 12" WG is not small and can't be tucked anywhere. If it's elevated and still behind the OB, I guess there'll be no benefit, or a risk of interfereing with rear wave of mid...
I can't find a convenient postion to elevate the flooder tweeters. The available space around that area is indeed limited.
On the equipment rack, I think it'd be too close to the (black) mini-monitor and the edge of DML panel.
I've briefly tried putting them on the central sub. Sitting slightly lower than normal, my line of sight would be about the same height with WG's up-firing mouth. And the WG's were pushed against the TV and emitting toward the DML panel. Maybe because of this, there'd be some interferences... I'm not sure.
Compared with the normal setup, the overall presentation is somewhat congested. Narrower, shallower, smaller overall. I don't like it.
Ah, I ran out of space for options. The 12" WG is not small and can't be tucked anywhere. If it's elevated and still behind the OB, I guess there'll be no benefit, or a risk of interfereing with rear wave of mid...
I see but what interests me most is whether the images were positioned higher? Was it like in the case of:
?
Was it that:
?
In other words was there anything odd about soundstage height, unnatural elevation, sound sources floating beneath the ceiling etc. when the tweeters were on top of the sub?
One idea again) to critic.
If we do not have sidewalls available for such placement, and not place for SSS in the room (out of walls) , but we do have free front wall.
Put soffite mounted 3 speakers in front wall, central speaker - to dealy for 9 ms. Will that give the same sense of depth as reflected 9 ms sound?
Of course side walls reflection still will have to be dealed and treated.
If we do not have sidewalls available for such placement, and not place for SSS in the room (out of walls) , but we do have free front wall.
Put soffite mounted 3 speakers in front wall, central speaker - to dealy for 9 ms. Will that give the same sense of depth as reflected 9 ms sound?
Of course side walls reflection still will have to be dealed and treated.
Last edited:
You mean You put them on the <80 Hz sub just next to each other? without any angular separation really?
and still there was a soundstage? somewhat congested, narrower etc. but a soundstage?
then it seems that information above 3 kHz doesn't carry the essential information in Your system as far as soundstaging is concerned, spaciousness - yes, extension of the soundstage - yes, but localisation strictly speaking - apparently no, interesting
For three separate speakers for stereo using matrix, there are multiple of alternatives for matrix coefficients too.
See Gerzon and e.g. AES Preprint: "An Optimum Linear-Matrix Stereo Imaging System"
In order to simulate SSS with three speakers, the three speakers should be placed in an angle much wider than normal stereo triangle. Maybe 120 deg.
See also some diyaudio threads on 3 speaker stereo.
This is quite funny, because SSS is a simulation of a three speaker stereo in a room
Why would you want to simulate the same thing twice to arrive to the starting point ? Maybe for the purpose of reinventing the wheel perhaps
Just use either or !
not necessarily - You can use an array of small drivers and achieve the desired not too low directivity of L and R channels by means of deflectors - half-waveguides (half because only in the front) - could be hinged for convenient adjustment of the amount of attenuation of the L and R components towards the listener
Yes, more or less side by side, they were 55cm apart (c-t-c).
I've tried even closer, or even mono-tweeter setup
I think you were there in those OSD and Stereolith threads In those cases, localization was mostly fine, with one or two exceptions.
Charcoal,
If I understand in your room there is no suitable side walls for SSS and the room is wide ?
Then suggestion would be to use 3 separate speaker in a wide angle and with a matrix. See the Miles linked above. If you use matrix x=0.5 then you can use exactly the same driver configuration as in SSS.
See also CLS's thread on his 3 speaker matrix (also with matrix x=0.5) as the driver connnections are the same as well.
Matrix with x=0.5 is very easy !
If I understand in your room there is no suitable side walls for SSS and the room is wide ?
Then suggestion would be to use 3 separate speaker in a wide angle and with a matrix. See the Miles linked above. If you use matrix x=0.5 then you can use exactly the same driver configuration as in SSS.
See also CLS's thread on his 3 speaker matrix (also with matrix x=0.5) as the driver connnections are the same as well.
Matrix with x=0.5 is very easy !
however mono is different and better than narrow base stereo as You have experienced YourselfI've tried even closer, or even mono-tweeter setup
anyway, in Your current setup You get all the highs >3 kHz via secondary and higher order celing-floor reflections exclusively yet You don´t miss anything as far as quality of highs and details is concerned vs Your previous experience with high directivity system?
Last edited:
Hi,
It's a 3-ch linear matrix system. Left = L-0.5R / Center = 0.5 (R+L) / Right = R-0.5L
It was originally DML panels for all 3-ch, and then the side channels were replaced by the current OB (for its much narrower dispersion). So I've yet to tried OB alone except briefly tests of the xover tuning in this setup.
It's hard for me to go back to 2-ch stereo.
As to switching off the flooder tweeters, IIRC, the major effects are darker, less spacious. Of couse it's very obvious.
The midrange drivers on OB are all different and wired as parrallel with level shading (loudest at the bottom, quietest at the top). Top 2 are widerangers originally so they can play above 10kHz, the bottom one is a guitar driver and can play (breakup) to 5~6kHz or so. Above 3kHz, the attenuation is shallow (6dB/oct. electrically).
So, direct HF is not completely absent, it's just less than ordinary system.
A side not: the central DML sounds very diffused, too. Although with the size of major visual impact, its sound is very layback - very wide dispersion means only a tiny fraction of sound energy hits the listener directly, all the rest are splashing everywhere in the room.
It's a 3-ch linear matrix system. Left = L-0.5R / Center = 0.5 (R+L) / Right = R-0.5L
It was originally DML panels for all 3-ch, and then the side channels were replaced by the current OB (for its much narrower dispersion). So I've yet to tried OB alone except briefly tests of the xover tuning in this setup.
It's hard for me to go back to 2-ch stereo.
As to switching off the flooder tweeters, IIRC, the major effects are darker, less spacious. Of couse it's very obvious.
The midrange drivers on OB are all different and wired as parrallel with level shading (loudest at the bottom, quietest at the top). Top 2 are widerangers originally so they can play above 10kHz, the bottom one is a guitar driver and can play (breakup) to 5~6kHz or so. Above 3kHz, the attenuation is shallow (6dB/oct. electrically).
So, direct HF is not completely absent, it's just less than ordinary system.
A side not: the central DML sounds very diffused, too. Although with the size of major visual impact, its sound is very layback - very wide dispersion means only a tiny fraction of sound energy hits the listener directly, all the rest are splashing everywhere in the room.