Look at the directivity pattern of human voice, that should explain. For instruments, a good reference: Jurgen (with Umlaut) Meyer-Acoustics and the Performance of Music. Our speech organ is a horn crossed with 4th order BP.
What are you trying to say? Directivity does influence the way a sound decreases in intensity with distance in the far-field?
Maybe I have been missing it, but has somebody said something about the ceiling induced reflexions ? And please, don't tell me this is not relevant for our auditory system.
an example : my room is wide, long and high, the ceiling is 2 slopes 25° 5.5o m high in the middle, this ceiling is a real membrane (thin braided bamboo), and indeed it's like absent.
But somewhere, there is a bridge, 2.50 m high, 2.00 m wide, hard wood. Just by speaking or clapping hands under it while walking, it's easy to feel the flutter and its sharp delimitations.
So, is it an issue ? Or because of the standard 2.50 m ceiling height nobody feels it anymore ? Or is it an other effect of my perverted hearing ?
an example : my room is wide, long and high, the ceiling is 2 slopes 25° 5.5o m high in the middle, this ceiling is a real membrane (thin braided bamboo), and indeed it's like absent.
But somewhere, there is a bridge, 2.50 m high, 2.00 m wide, hard wood. Just by speaking or clapping hands under it while walking, it's easy to feel the flutter and its sharp delimitations.
So, is it an issue ? Or because of the standard 2.50 m ceiling height nobody feels it anymore ? Or is it an other effect of my perverted hearing ?
I think it is rather that by then it became feasible with single dynamic driver
flutter surely is an issue, late reflections and echoes can easily become a problem, I think everyone agrees
OTOH first ceiling reflection is a quite separate non-issue
No, you misinterpreted what I meant to say. Same amount of spectral energy does travel further for a horn because it is compressed onto a smaller area and hence louder. -compared to a 4pi radiation of lower spectra where dimensions of the mouth lose directivity- And there is greater difference in relative hearing ability between low and mid spectra for different levels.
What the author at Deutsche Welle probably meant to say is that a certain portion of spectra can end up at the mouth like plane waves, so hence the <6dB propagation loss.
Last edited:
I have noticed that Holm has a multi window approach although I'm not sure what he bases it on. Didn't MLSSA have a mode also?
I suspect that some external processing would be necessary to get exactly what we want, although it would be nice to have an automatic mode in one of the popular packages. Both Salmi and Kates suggest a combination of variable time windowing and critical band smoothing. Not too hard to implement.
David S.
I think it does. Both Salmi and Kates show "perceptual" frequency response curves of speakers in rooms and the floor bounce is a strong artifact in the measurement. Bech did his study in a simulated room in an anechoic chamber. He turned the level of typical reflections up and down and found the floor bounce was generally audible (more likely audible than many other typical room reflections).
I think it makes a strong argument for either some midrange directivity or boundary woofers, such as Allison or the AR9 used.
David S.
Thanks for the reference.
It looks very interesting (but will take a while to dig through!). He does appear to EQ the direct or early sound but he also uses a room curve, something needed for EQ when you only have steady state curves to work with.
I have to get a DSP system running!
David
Floor and ceiling reflections are mainly relevant for tonality and depth impression, but much less for the stereo imaging.
Don't clap yourself, but let someone else do it and listen from your usual position. How about the flutter echoes?
Next place someone next to the loudspeaker and let him/her clap. Do you still hear any flutter echoes at your listening position?
Rudolf
Practically speaking, both would be required ? Directivity of the floor located woofer doesn't really matter, but the high mounted midrange driver it hands over to needs to have enough directivity to minimize the comb filtering ripples that would otherwise form.
(And enough height to maximize the time delay of the reflection to help our perceptual mechanism to separate it better, perhaps putting the midrange above the ear level tweeter to gain additional reflection delay would be beneficial ?)
One other thing I was thinking about recently when working out the design of a 3 way with this configuration is that if the tweeter is at ear level, (say 95cm) midrange just below it, and woofer at the floor, that's a good 50-60cm centre to centre between woofer and midrange.
At a 250Hz crossover frequency (which I think is probably near optimal for this low woofer high midrange arrangement) that's a spacing of about 0.45 wavelengths. This is enough for significant vertical directivity to form at the crossover frequency due to driver spacing. Is this beneficial, or is it the wrong way to look at it ?
With an even order filter they are also spaced far enough apart to put a significant dip in the power response at the crossover frequency. At high frequencies this wouldn't matter so much, but would putting a dip in the power response near the schroeder frequency cause an apparent lack of total energy in the room in the low midrange that wouldn't be there with close driver spacing ?
Last edited:
Excellent Rudolf, I was not aware of this distinction, and effectively, when the clap is done by an "assistant", the flutter disappears, only is noticeable a change in tune when compared with claps under high ceiling. My speakers are very far from this bridge, of course, my listening position too.
I agree. The solution employed in constant directivity cornerhorns is to blend the midhorn and woofer below 250Hz, which smoothes vertical modes. The horizontal reflections don't exist because the woofer and mid are acoustically close to the apex of the corner, and the verticals are mitigated by blending, much like a multisub configuration but spaced closer because of the wavelengths involved.
Where cornerhorns aren't possible, I rely on a flanking sub arrangement, which also blends two sound sources from the lower midrange down. It is essentially a subwoofer run a little on the high side, blended with the mains to smooth vertical modes, floor bounce and the notch off the back wall. Of course, the flanking sub has to be run fairly close to the mains - typically just a couple feet away, offset in all three dimensions - beside, behind and below the mains. It is usually close to the main speaker it is flanking, and symmetrical. This close proximity prevents localization cues; It's just far enough away to smooth the 80Hz to 200Hz range.
My take on the floor and ceiling bounce is that it is difficult to impossible to control this in a small room with directivty. In my rooms I put a large and thick rug on the floor and a difusser on the ceiling. These work well. It would be nice to be able to have a directivity control sufficient to control this problem, but that becomes problematic. The woofer isn't narrow vertically in any case, even if the waveguide is. And now the directivities don't match. I have just concluded that dealing with these reflections is best done in the room and not in the speaker.
---------------------
Dr Geddes: I am trying to work out the geometry on this. I beleive your waveguides have an included angle (or 6dB down points) of about 90 deg (correct me if I am wrong). So, how far back is your chair from the speaker cabinet and how tall is the ceiling and how high off the floor is your waveguide?
EDIT:
But an extra woofer near the floor should help you to mitigate at least the floor-bounce. You might be able to integrate it with the speaker stand . If such a solution might prove insufficient, then what about a line-array for use below about 500 hz?
The ADAM Olympus system should have pretty good vertical(!) control: Olympus Sound System - Description | ADAM Audio GmbH
But an extra woofer near the floor should help you to mitigate at least the floor-bounce. You might be able to integrate it with the speaker stand
. If such a solution might prove insufficient, then what about a line-array for use below about 500 hz?The ADAM Olympus system should have pretty good vertical(!) control: Olympus Sound System - Description | ADAM Audio GmbH
Last edited:
I think your view of the floor bounce problem being unsolvable in the speaker design is tainted by looking through the glasses of constant directivity designs.
As much as I like the concept and goal of constant directivity, the typical 2 way approach of constant directivity horn crossed over with a large mid-bass driver at the frequency where directivity is matched rather unfortunately puts the crossover in the middle of the midrange, lets say 1.5Khz.
You then have the same floor-bounce problems you have with a typical stand mounted 2 way design - put the speaker too close to the ground and imaging suffers, too high up and floor bounce is a problem.
To really solve the problem you need a crossover in the vicinity of 200-300Hz with the lower frequency driver at the floor and the higher frequency driver(s) a lot higher up.
Not really solvable with a 2 way CD design, and if you went to 3 way you then have the absurd problem of requiring two "woofers", one close to the floor which is actually producing the bass, and a second large one higher up which is only covering 300-1500Hz, but still has to be large to match the directivity of the horn...
Or perhaps a 2.5 way design ? Two identical woofers, floor height woofer covering 20-250Hz, upper woofer just below the horn covering 20-1500Hz, crossing over with the horn, that would eliminate floor bounce. If you matched it with the baffle step frequency there is your baffle step taken care of too.
Last edited:
Then this is not a "small room". It takes a lot of open space for this to occur. That not the "assumption" that we are dealing with here.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.