Toole shows in his review (fig. 7) that the first 3 reflections in a "typical" room are already at or above "image spread/shift" threshold for speech.
Quote:
Originally Posted by weltersys
I used a Grundig YB 400PE in varying positions around a room, Graaf uses a wide dispersion co-ax (I forgot which) and I don't have the equipment to produce the nifty wavelet sweep which revealed Elias' FCUFS speaker's narrowing and disjointed HF response.
The dispersion of the FCUFS would definitely change the "spacious" effect, a wide HF dispersion speaker reveals it's floor location, a narrow HF dispersion has a source image varying from floor to ceiling with frequency.
Graaf''s wide dispersion co-ax, a 130 mm KEF UniQ has a -6 dB dispersion of around 160 degrees at 8000 Hz, and still is 90 degree at 16000 Hz, as can be seen in his post #2732 (and previous ones I forgot).
In a FCUFS, this wide high frequency dispersion insures the off axis HF is at or slightly louder than the longer path length reflected sound.
My hearing "zeros in" on the location of the initial HF sound if it is of sufficient level compared to reflected, later arrival time HF.
Graaf seems to hear differently.
Elias' wavelet shows the HF reflected sound at a louder level than the direct sound above 3000 Hz, which is quite different than the presentation that the KEF UniQ or the wide dispersion Grundig YB 400PE have used as a FCUFS.
To get an idea what a speaker similar to Elias' test showed, with a more narrow HF dispersion would sound like as a FCUFS, today I compared the 400PE using a board as a waveguide to direct the HF to the ceiling.
This set up did result in a "spacious" distribution of HF, not specific in location to just the ceiling reflection as I had previously stated.
Although vocals and instruments fundamental tones still seemed to come from the floor location, (I could close my eyes and spin left or right and still reliably point to the speaker location) now the location of the cymbals, fricatives and sibilance seemed to come from an indeterminate, disjointed location.
I agree with you, none of this sounded desirable to me compared to a "normal" ear height location of the speaker.
Art
Originally Posted by weltersys
I used a Grundig YB 400PE in varying positions around a room, Graaf uses a wide dispersion co-ax (I forgot which) and I don't have the equipment to produce the nifty wavelet sweep which revealed Elias' FCUFS speaker's narrowing and disjointed HF response.
The dispersion of the FCUFS would definitely change the "spacious" effect, a wide HF dispersion speaker reveals it's floor location, a narrow HF dispersion has a source image varying from floor to ceiling with frequency.
Marcus,
Graaf''s wide dispersion co-ax, a 130 mm KEF UniQ has a -6 dB dispersion of around 160 degrees at 8000 Hz, and still is 90 degree at 16000 Hz, as can be seen in his post #2732 (and previous ones I forgot).
In a FCUFS, this wide high frequency dispersion insures the off axis HF is at or slightly louder than the longer path length reflected sound.
My hearing "zeros in" on the location of the initial HF sound if it is of sufficient level compared to reflected, later arrival time HF.
Graaf seems to hear differently.
Elias' wavelet shows the HF reflected sound at a louder level than the direct sound above 3000 Hz, which is quite different than the presentation that the KEF UniQ or the wide dispersion Grundig YB 400PE have used as a FCUFS.
To get an idea what a speaker similar to Elias' test showed, with a more narrow HF dispersion would sound like as a FCUFS, today I compared the 400PE using a board as a waveguide to direct the HF to the ceiling.
This set up did result in a "spacious" distribution of HF, not specific in location to just the ceiling reflection as I had previously stated.
Although vocals and instruments fundamental tones still seemed to come from the floor location, (I could close my eyes and spin left or right and still reliably point to the speaker location) now the location of the cymbals, fricatives and sibilance seemed to come from an indeterminate, disjointed location.
I agree with you, none of this sounded desirable to me compared to a "normal" ear height location of the speaker.
Art
Last edited:
Here's the directivity pattern of the speaker I used in the flooder test
An externally hosted image should be here but it was not working when we last tested it.
In acoustics generally.
More particularly with respect to ASW (..which obviously modifies source size). (..of course you knew this, but it probably didn't occur to you.)
A good overall "primer" with respect to papers on this subject (as often the case), can be found on Griesinger's page:
http://www.davidgriesinger.com/
Specifically scrolling-down to the listing and selecting:
"The psychoacoustics of apparent source width, spaciousness & envelopment in performance spaces"
Unfortunately, the paper has some serious formatting problems that make it a "rambling read". You can read it several times over (..and should), and still only grasp or retain a portion of the material. Because of the formatting it makes reading it prone to misinterpretation (..even though it's written in a reasonably "lay" perspective).
Last edited:
Study or not,
-actually it's a vertical effect as well, though vertical effects are more prone to listener differences than horizontal ones. (i.e some people process the vertical better than others, and some do it much worse.) Processing is more difficult because ILD differences are low (or said somewhat differently signal correlation between L & R ears is higher).
And the vertical effect does NOT just operate at high freq.s, rather it usually *dominates* at high freq.s, in the same manner that direct sound usually dominates source position to the source rather than a reflection.
Last edited:
Interesting, that shows a bit wider dispersion than I expected considering the ceiling reflection was louder than the initial impulse in the wavelet chart.
Still much narrower than Graaf''s KEF UniQ which is -6 dB dispersion at around 160 degrees at 8000 Hz, yours seems around -6 at 45 degrees at 8000 Hz.
What was the direct distance from the FCUFS to the mic, and approximate off axis angle, and what was the reflected distance length?
Note the dB scale in the wavelet plot was 20dB while in the freq response 70dB.
The listening (and measurement) distance was about 2 m. While sitting in the listening position I could see the speaker on the floor at about 25 degree angle down.
I found a picture of the speaker (ALMOST the same) except the 6.5" driver was Seas P17REX. The tweeter and the arrangement was the same.
An externally hosted image should be here but it was not working when we last tested it.
It is a fine speaker.
But when I listened it placed on the floor it sounded like a radio, if such a comparison can be made
I don't believe in this concept. It can do some things better than conventional stereo triangle, BUT:
1) The speakers may be unlocalisable at high freqs, while conventional direct firing tweeters can lead to speaker localisation.
On the other hand, the same improvement can be made to normal stereo speakers by simply tilting the tweeters upwards to reduce the direct to reflection ratio !
2) The spaciousness can be higher than with conventional direct firing stereo triangle.
On the other hand, spaciousness can be increased in the listening room by introducing a lots of lateral wall reflections. It can be done using a special sideways firing speaker
or by appropriately toeing the speakers (according to some) or by adding extra speakers emitting 'surround' signal of sort.- Elias
That is a rather typical performance of many dome tweeters. I thought the UniQ would be better.
Are there any real world measurements of the UniQ available? I've seen only pretty marketing drawings so far.
Last edited:
Not that I'm aware of.
..actually in the context of arrays there might be some information.
Last edited:
There's some data of the LS50 at stereophile.com:
KEF LS50 Anniversary Model loudspeaker Measurements | Stereophile.com
Does this look like "-6 dB dispersion at around 160 degrees at 8000 Hz"?
3-8 kHz it looks like its almost -15db at 90 degrees.
Here is Soundstage's anechoic measurements:
SoundStageNetwork.com | SoundStage.com | NRC Measurements: KEF LS50 Loudspeakers
Quote:
Originally Posted by weltersys
KEF UniQ which is -6 dB dispersion at around 160 degrees at 8000 Hz
.
The "rather typical performance" of the KEF UniQ is over 3 times the beamwidth at 8kHz as Elias' speaker.
I'd say Elias' speaker is more typical in it's dispersion pattern than the much wider KEF UniQ, which seems to approach the theoretical maximum dispersion of 180 degrees up high.
As to which dispersion pattern is preferable for a speaker laying on the floor, I'd leave that up to those few that prefer the sound of speakers laying on the floor.
Art
Originally Posted by weltersys
KEF UniQ which is -6 dB dispersion at around 160 degrees at 8000 Hz
I'm sure Graaf can post actual measurements of his UniQ
.The "rather typical performance" of the KEF UniQ is over 3 times the beamwidth at 8kHz as Elias' speaker.
I'd say Elias' speaker is more typical in it's dispersion pattern than the much wider KEF UniQ, which seems to approach the theoretical maximum dispersion of 180 degrees up high.
As to which dispersion pattern is preferable for a speaker laying on the floor, I'd leave that up to those few that prefer the sound of speakers laying on the floor
.Art
Does the KEF LS50 look like Graaf's 130 mm KEF UniQ ?
At any rate, according to the Soundstage's anechoic measurements, at 8kHz the KEF LS50 is 83 dB on axis, 79 dB at 45 off (only -3 for 90 degrees of dispersion) 75 dB at 60 off (-8dB for 120 degree dispersion) 71 dB at 75 degrees off (-12 dB for 150 degree dispersion).
Seems to be more narrow than the KEF UniQ charts I based my comments on, dispersion is similar to Elias' speaker.
Attachments
Last edited:
Now that we're back on topic and real science is being investigated...
Back in the '80's I did alot of messing around with Snell Type Ones DIY. all kinds of various placement in several room sizes etc. Nice effect but back then most pro stuff for measurement was out of the question, so not much to show other than experience.
Trying to resolve the combing effect and proper balance of direct to reflected was crucial. Reciently while working on MLTL type, where I'd made a baffle plate for the tweeters (rt4001 AMT's) flush mounted direct radiated and two 6.5" peerless midbass's top mounted 48" off the floor up firing. Placement from experience said space the drivers according to the crossover point of 2200, so a symmetrical triangle, mids side by side. Now this is not FCUFS, but along the same lines. In my mockup this yielded a beautiful ambience while maintaining tight imaging. Since my project may take me where ever it leads, still in the works 7 months later... I have room to play, and all that experience from yester year. Partial FCUFS implementation may go into what all I'm... for the lack of a better word, Striving for.
Playing with this mostly active, lots of amps and lots of drivers to mess with and can measure finally!
Thought just hit me from my experiments from the past was that a curved baffle board, think paneling, curved upward to meet the ceiling and say a good meter wide improved the imaging in a real FCUFS setup. Something to go on
Back in the '80's I did alot of messing around with Snell Type Ones DIY. all kinds of various placement in several room sizes etc. Nice effect but back then most pro stuff for measurement was out of the question, so not much to show other than experience.
Trying to resolve the combing effect and proper balance of direct to reflected was crucial. Reciently while working on MLTL type, where I'd made a baffle plate for the tweeters (rt4001 AMT's) flush mounted direct radiated and two 6.5" peerless midbass's top mounted 48" off the floor up firing. Placement from experience said space the drivers according to the crossover point of 2200, so a symmetrical triangle, mids side by side. Now this is not FCUFS, but along the same lines. In my mockup this yielded a beautiful ambience while maintaining tight imaging. Since my project may take me where ever it leads, still in the works 7 months later...
I have room to play, and all that experience from yester year. Partial FCUFS implementation may go into what all I'm... for the lack of a better word, Striving for.Playing with this mostly active, lots of amps and lots of drivers to mess with and can measure finally!
Thought just hit me from my experiments from the past was that a curved baffle board, think paneling, curved upward to meet the ceiling and say a good meter wide improved the imaging in a real FCUFS setup. Something to go on