one element flooder which is horizontally omnidirectional would produce some very early 1st order side wall and 1st order front wall reflections and a strong 1st order contralateral side-wall reflection plus 1st order rear wall reflection much louder than in case of a flooded directional array
Well, obviously it depends how one success in integration into the corner.
Contralateral reflection will reamain, sure, but at higher freqs not much if practically at all depending on the size of the driver, and anyway less than normal stereo speakers toed in.
yes, I presume imperfect integration into the corner and rather wide dispersion, perhaps wrongly, because, interestingly, with higher directivity, like in the original "Fostex 8" flooder" what flooder achieves is in fact a kind of a virtual ceiling speaker because as measurements and listening tests have shown ceiling reflection dominates over the direct sound
this is a virtual ceiling speaker with one very important advantage over a real one - a real ceiling speaker produces a very strong 1st order floor reflection, a virtual one - none
Last edited:
More than once I've been thinking about ceiling speakers, and were pondering why not hang flooders. At least below about 1kHz it does not matter if a reflection comes above or below as long it has the same incidence angle relative to sagittal plane.
High freqs are different of course, and I think something coming from above is not so harmful than coming from below, since evolutionary we didn't expose much to above sounds at savannah so they are not strong indicators.
- Elias
yes, and therefore a real ceiling speaker is a bad idea and a virtual - ceiling reflection of a flooder - is good
accidentally I came across an interesting thread at Swedish forum: http://www.faktiskt.se/modules.php?...ic&t=42223&postdays=0&postorder=asc&list=full
James Croft participates in the discussion and he declares that Snell "Type 1" was his design and explains where the idea came from:
overall interesting discussion, most of it in English, the guy appreciates advantages of Beveridge placement (discussing Swedish Bremen speakers) and of flooder-type arrangement which he calls "a speaker boundary coupled close to the floor" while on the other hand he appreciates Dr Geddes' Summa as well
and here from a thread from the same forum documenting DIY-ing some Carlsson style flooders:
http://www.faktiskt.se/modules.php?name=Forums&file=viewtopic&p=502903
James Croft participates in the discussion and he declares that Snell "Type 1" was his design and explains where the idea came from:
overall interesting discussion, most of it in English, the guy appreciates advantages of Beveridge placement (discussing Swedish Bremen speakers) and of flooder-type arrangement which he calls "a speaker boundary coupled close to the floor" while on the other hand he appreciates Dr Geddes' Summa as well
and here from a thread from the same forum documenting DIY-ing some Carlsson style flooders:
http://www.faktiskt.se/modules.php?name=Forums&file=viewtopic&p=502903
Thanks for that link, I found it interesting. I wish that it were in English because I would like to present my side of Mr. Crofts comments since I don't agree with him on every aspect of his approach to waveguides.
no language problem really, I am sure that all Swedes there understand English
Mr Croft's post from 2011-01-28 09:23 is particularly interesting, he comments on the fact that there were soundstage height issues with Snell "Type 1":
now according to Mr Croft:
in the light of the foregoing and IF Mr Croft is right then it is easy to understand why the soundstage height is unrestricted in the case of the original flooder that is a floor coupled upfiring single 8 inches whizzer cone fullrange driver, it suffices to look at directivity plot of a typical driver of this kind:
obviously most of the energy of the 1<4 kHz frequency band reaches the listener via the ceiling reflection - so the ceiling reflection effectively becomes a virtual ceiling tweeter - a virtual one because it generates no its own early 1st order floor reflection
the frequency of the "crossover" between the floor coupled "midwoofer" and the "virtual tweeter" is determined by listener's distance, ceiling height and the angle at which the fullrange unit fires upwards
now according to Mr Croft:
in the light of the foregoing and IF Mr Croft is right then it is easy to understand why the soundstage height is unrestricted in the case of the original flooder that is a floor coupled upfiring single 8 inches whizzer cone fullrange driver, it suffices to look at directivity plot of a typical driver of this kind:
An externally hosted image should be here but it was not working when we last tested it.
obviously most of the energy of the 1<4 kHz frequency band reaches the listener via the ceiling reflection - so the ceiling reflection effectively becomes a virtual ceiling tweeter - a virtual one because it generates no its own early 1st order floor reflection
the frequency of the "crossover" between the floor coupled "midwoofer" and the "virtual tweeter" is determined by listener's distance, ceiling height and the angle at which the fullrange unit fires upwards
I had similar experience with my monoflooder test.
If the speaker was facing directly or almost directly the ceiling, it was impossible to locate the speaker and sound was coming from front but not from low.
But if I turned the speaker a bit facing more to the listening area the speaker immediately become localised or the sound was coming from low angle.
Then I don't know how well Snell speaker satisfied the purpose in the first place. Maybe it never become popular for this reason, who knows.
- Elias
If the speaker was facing directly or almost directly the ceiling, it was impossible to locate the speaker and sound was coming from front but not from low.
But if I turned the speaker a bit facing more to the listening area the speaker immediately become localised or the sound was coming from low angle.
Then I don't know how well Snell speaker satisfied the purpose in the first place. Maybe it never become popular for this reason, who knows.
- Elias
But I don't understand Swedish and could not figure out how to sign up or post.
google translate is pretty good at Swedish, at least good enough to guide Your through the sign-up process
this one looks like a suitable loudspeaker for a floor coupled up-firing configuration (FCUFC? 
) it's an 8 inches Tannoy concentric
ps.
for the listener's distance from the speaker of ~250 cm and ceiling height of ~250 cm as well the direct sound that reaches the listener is radiated by the speaker at ~70 degrees and the sound reflected via ceiling is radiated by the speaker at ~35 degrees
) it's an 8 inches Tannoy concentricps.
for the listener's distance from the speaker of ~250 cm and ceiling height of ~250 cm as well the direct sound that reaches the listener is radiated by the speaker at ~70 degrees and the sound reflected via ceiling is radiated by the speaker at ~35 degrees
Attachments
Last edited:
this one looks like a suitable loudspeaker for a floor coupled up-firing configuration (FCUFC?
ps.
for the listener's distance from the speaker of ~250 cm and ceiling height of ~250 cm as well the direct sound that reaches the listener is radiated by the speaker at ~70 degrees and the sound reflected via ceiling is radiated by the speaker at ~35 degrees
I have been thinking that we have on this forum reports from very diverse yet effective applications of the FCUFC principle including some speakers with much wider dispersion than 8 inches fullrange or coaxial
so perhaps the source of the Snell "Type 1"'s problem with elevation of the soundstage wasn't inadequate ratio of direct/ceiling reflected sound in the >2kHz band (that is of too much direct vs reflected) - after all the frequency band was handled by a small tweeter in "Type 1" so there certainly was a plenty of ceiling reflected sound
OTOH the crossover frequency was set at 2 kHz and the midfoower was 10 inches firing forward, just slightly tilted back, so most probably there was very little ceiling reflected sound in the 1<2 kHz band and in addition to that a very early floor reflection off the "wheelchair ramp" reached the listener, it didn't cause any frequency response related problems because the angles of the baffle and the "ramp" had been carefully calculated and matched BUT the problem of perception of the impulse reflected off the floor (the "ramp") remained - and the speaker thus was partly unmasked as physical source of sound at a particular height causing sometimes perception of the soundstage that was too low
Attachments
I'm not convinced yet. Looks like Snell "Type 1" tweeter has too much direct sound. This is similar to my flooder experiments where aiming the box more towards the listening area than the ceiling resulted localisation of the speaker at the floor or the direction of the sound was very low.
And, in my experience in the flooder type of a speaker the direct sound to reflection ratio should be very small. This could be listener dependent however ! If one looks results from any typical psychoacoustic test, therre are individual variances in the order of several dB. This difference could lead to quite different speaker design depending of the individual.
Maybe Snell designed the speaker mostly for himself, and found it to satisfy his own perception.
Similar possible occurance comes in mind with Walter Schubach with Stereolith speaker. Possibly
- Elias
And, in my experience in the flooder type of a speaker the direct sound to reflection ratio should be very small. This could be listener dependent however ! If one looks results from any typical psychoacoustic test, therre are individual variances in the order of several dB. This difference could lead to quite different speaker design depending of the individual.
Maybe Snell designed the speaker mostly for himself, and found it to satisfy his own perception.
Similar possible occurance comes in mind with Walter Schubach with Stereolith speaker. Possibly
- Elias
Last edited:
what kind of speaker You used for Your tests?
tilted position results both in higher direct/reflected ratio and audibility of a floor reflection
This could be listener dependent however ! If one looks results from any typical psychoacoustic test, therre are individual variances in the order of several dB. This difference could lead to quite different speaker design depending of the individual.
Maybe Snell designed the speaker mostly for himself, and found it to satisfy his own perception.
Similar possible occurance comes in mind with Walter Schubach with Stereolith speaker. Possibly
- Elias
yes, that's possible
So if you listen to ceiling speakers standing on your head it will not sound like a radio ?
I don't think so, You would have rather to sit on a ceiling upside down, like a bat
It was a normal 2-way 12 l box with 1" dome tweeter and 6.5" bass.
There could have been floor reflection with tilted position, but the speaker was also on the floor with tweeter height about 20 cm. In addition there was carpet below the speaker and between the listening position.
On the other hand Snell type 1 also has floor reflection for the tweeter.