I don't have any data to back this up, just my experience. But I think you can get great imaging and dispersion as well. A three way with lower crossover points that stays omnidirectional and then goes to a high dispersion tweeter would image well.
I think reducing diffraction and early reflections is the issue. The ambience of late reflections aren't an issue wrt imaging.
Also lobing and other crossover issues like dispersion mismatches hurt imaging.
I think reducing diffraction and early reflections is the issue. The ambience of late reflections aren't an issue wrt imaging.
Also lobing and other crossover issues like dispersion mismatches hurt imaging.
Last edited:
I have to have our main system in a quite small room, waveguide speakers right against the front wall. Imaging (L/R) was stellar and worked across multiple seats but ambience wasn't so great and the sound was somewhat like it was coming through the wall from the next room.
I put in some small wideband drivers (BMR types, wide coverage even in the top octave) facing the wall and ceiling behind them and drive them with a ~15msec delayed signal about 7dB down and from about 1500Hz up. Adds a nice ambience, doesn't hurt the imaging any that I can hear and significantly helps the depth illusion. Hard to say what if anything it does to timbre, the character of sounds seems about the same, but it's like being in the (or "a") same room where the music is being performed instead of outside it.
A very "non-purist" approach, I admit, but it seems hard to get excited about being purist when speakers and rooms already mess with sound so much on their own anyway. I don't see myself taking out the ambience tweeters any time in the near future.
I put in some small wideband drivers (BMR types, wide coverage even in the top octave) facing the wall and ceiling behind them and drive them with a ~15msec delayed signal about 7dB down and from about 1500Hz up. Adds a nice ambience, doesn't hurt the imaging any that I can hear and significantly helps the depth illusion. Hard to say what if anything it does to timbre, the character of sounds seems about the same, but it's like being in the (or "a") same room where the music is being performed instead of outside it.
A very "non-purist" approach, I admit, but it seems hard to get excited about being purist when speakers and rooms already mess with sound so much on their own anyway. I don't see myself taking out the ambience tweeters any time in the near future.
This is a design that would provide both ambience and imaging (not the flattest frequency response though).
D-101S super-Swan
My question would be whether a theoretical point source would image perfectly? I assume it would because you're getting clean mirror image sound from both sides.
D-101S super-Swan
My question would be whether a theoretical point source would image perfectly? I assume it would because you're getting clean mirror image sound from both sides.
Last edited:
Has anyone got ideas as to what is the "best" omni reflector shape? This is what I mean omnidirectional speaker reflector - Google Search
The most common ones appear to produce a doughnut shape pattern, hardly what I'd call omnidirectional?
The most common ones appear to produce a doughnut shape pattern, hardly what I'd call omnidirectional?
Why? I would have thought dipole would be a better compromise between directivity and spaciousness (ambience)
But the "ideal"speaker is a pulsating sphere. Is it just that 360 horizontal is easier to accomplish?
.
First, I see no reason why a pulsating sphere is ideal - it has no directivity, it is omni-directional. I do see that having a virtual point source of sound is ideal, but one with doirectivity. A pulsating sphere does not do that. A waveguide presents an ideal spherical surface at its mouth that is emanated from a virtual point source of sound, but yet has directivity. To me, this is ideal. I do not want to excite a room to the maximum extent possible, especially not in a small room, the reflections are just too soon. I want to direct the sound to the listener so that there is a clean clear direct sound from which I can hear a very clear image. Too many early reflections mess that up. Minimizing the early reflections is necessary for good imaging. Absorbing them makes the room dead. A lively room with speakers with a high DI will have great imaging and good spaciousness. An Omni cannot do that, you have to choose one or the other.
Yes, I made a mistake, pulsating sphere is omni-directional. But it has non-flat frequency response. Limiting radius to zero, it becomes a point source, omni-directional and flat frequency response. However, absolutely unpractical re displacement needed and velocity that would be ultrasonic.
I am not sure why you say "it has a non-flat frequency response". If the source is pulsating with constant acceleration then it should be flat everywhere. If constant velocity or constant displacement it won't of course. Of course, constant acceleration can never be achieved in practice over a broad bandwidth, and therein lies the problem. But we do have EQ so given an unlimited amount of electrical power and such a thing could be done - almost!
What I say that I want is "a point source with high constant directivity." That's my ideal.
What I say that I want is "a point source with high constant directivity." That's my ideal.
Last edited:
My room is 14 x 20 feet with an 8 foot ceiling - pretty small. The speakers are about two feet from any surface at an end wall of the greatest dimension and toed in quite a bit. The imaging is solid at almost any seating position. Spaciousness is acceptable, but not pronounced. In general people love the system with studio work, but many classical pieces just come off as bad recordings since the room covers up nothing.
Pulsating sphere has a critical frequency w = Co/R, where w = 2pi*f, Co = 340m/s and R is the sphere radius. Radiating impedance (simplified) is radiating mass m in parallel with radiating resistance r, connected to common force F.
Normalised radiating impedance is Znv = jkR/(1 + jkR), Znv = A + jB
A = (kR)*2/(1 + (kR)*2)
B = kR/(1 + (kR)*2)
.... and the well known plots with kR as a parameter.
Normalised radiating impedance is Znv = jkR/(1 + jkR), Znv = A + jB
A = (kR)*2/(1 + (kR)*2)
B = kR/(1 + (kR)*2)
.... and the well known plots with kR as a parameter.
Thanks Earl, my room is smaller than that, about 13' by 15' by 7'6". I find however I have my speakers (mono/dipole) they sound best near field and well away from the front wall, probably due to it's small size. When dipole I have them facing parallel to side walls but when monopole they sound best crossing in front of me. I tend to prefer them dipole, particularly when listening to 50's-60's jazz.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.