A monopole placed in a room corner excites all room modes maximally - thereby delivering maximal energy into the room.
A dipole placed in the same room corner excites all modes minimally – delivering minimal energy into the room.
A monopole placed in the middle of the room excites the first degree room modes (half wavelength between walls) minimally.
A dipole in the same position excites them maximally – except the modes perpendicular to the dipole axis.
Rotation of a monopole in the room doesn't change anything.
Rotation of a dipole in the room can excite or attenuate a room mode relative to room modes perpendicular to it.
This should explain why measurements of a dipole and a monopole at the same position in the room can't decide on the best aptitude for either M or D for the room. Even if one can find different optimal placements for M and D in the same room, they need not be the best placements in another room of different size and proportions.
Rudolf
A dipole placed in the same room corner excites all modes minimally – delivering minimal energy into the room.
A monopole placed in the middle of the room excites the first degree room modes (half wavelength between walls) minimally.
A dipole in the same position excites them maximally – except the modes perpendicular to the dipole axis.
Rotation of a monopole in the room doesn't change anything.
Rotation of a dipole in the room can excite or attenuate a room mode relative to room modes perpendicular to it.
This should explain why measurements of a dipole and a monopole at the same position in the room can't decide on the best aptitude for either M or D for the room. Even if one can find different optimal placements for M and D in the same room, they need not be the best placements in another room of different size and proportions.
Rudolf
"Velocity" here refers to the "particle velocity" or the velocity of motion of the particles in space. This velocity can be any value from zero and up. The velocity of sound is the speed at which the pressure disturbance moves and is not the same as the velocity at which the particles in space oscilate.
This is exactly like voltage and current in a circuit. The pressure is like voltage and the current is like the particle velocity. We know that the electrons in a wire don't move back and forth very fast, in fact the actual motion about the mean is actually pretty slow, but the electric field that drives the electrons moves at near the speed of light.
And thanks John - sensibility is finally creaping into this discussion. But it's all "theoretical" and has no relavence to "beliefs" based on "what I hear".
Last edited:
What is the connection of the pressurisation ability and the modulation reproduction ability ?
Sorry to say that, but the connection is:
We are still lacking a controlled, fair and equal comparison, that shows categorically a better modulation reproduction ability of the dipole. Either by theory or by measurement.
Rudolf
How do you setup this EQ. Do you start with one sub and EQ all major peaks (leaving the dips alone), then add a second sub and again EQ the peaks, add a third sub...
Or do you use something proprietary?
And I think that a reliable connection between "modulation production" and listener perception is also lacking. I fail to see why a LF playback wants a a high rate of modulation production (or whatever the measure is). I could give hypothetical rationalizations to say just the opposite.
I have software that does this, not completely automated, that would be complex. This software is proprietary. But the technique is basically what you said - start with the sub farthest from the listener and EQ it to blend in well with the mains (no crossover is used, but the mains generally don't have a very low cutoff). Then add in the next closest sub and optimize. Finally the third, if there is one and so on. With the software, which shows the results in real time, I can converge on a good solution in 15-20 minutes.
This service is only offered to my customers however because I need some value added to sell subs, which are basically commdities and all the same in my book. I think that you can see that I really don't care what kind of sub it is, the technique works with anything. The controller is a DCX2496 or sometimes a MiniDSP.
A room setup this way is incomparable to one in which the sub is simply placed in the room, or one in which the LFs are handled by the mains. After some 40 years of studying the LF sound in small rooms and doing sound system setup, I have concluded that this is the only way to consistantly get great bass in any room. But still, some rooms are better than others, I can see this in the variety of examples that I have done. My room, which has very high LF damping, still comes out among the best. With a lot of LF damping not much is required from the subs, but some control always helps.
Im certainly not up to speed with the Drs in the house, but this thread got me thinking.
Perhaps i repeat others concepts BUT:
Why not turn the ENTIRE room into a bipole?
If you humour me for a minute, imagine front and rear in wall IB subs, wired push pull.
How would the room behave?
With DSP i imagine it would be simple to tune with delay and or phase rotation. Probably im hand waving frantically and its vague at best, but the thought just occurred to me.
Perhaps i repeat others concepts BUT:
Why not turn the ENTIRE room into a bipole?
If you humour me for a minute, imagine front and rear in wall IB subs, wired push pull.
How would the room behave?
With DSP i imagine it would be simple to tune with delay and or phase rotation. Probably im hand waving frantically and its vague at best, but the thought just occurred to me.
I think John K has a similar suggestion (like the push-pull dipole idea) on his website: DP_woofer_room
A DBA is slightly different, since it has a time delay between the front and back array corresponding to the path length between the two arrays. The front array generates an approximation of a plane wave which is then actively dampened by the rear array when it arrives at the rear wall. Thus it creates a pure travelling wave.
It's expensive and relies very much upon room shape and acoustics in order to function properly, which are major drawbacks. It's not practical in most rooms (although it is very cool
).
A DBA is slightly different, since it has a time delay between the front and back array corresponding to the path length between the two arrays. The front array generates an approximation of a plane wave which is then actively dampened by the rear array when it arrives at the rear wall. Thus it creates a pure travelling wave.
It's expensive and relies very much upon room shape and acoustics in order to function properly, which are major drawbacks. It's not practical in most rooms (although it is very cool
).
It is not lacking but you are being ignorant. And it is not production but reproduction that we want.
The perception is everything. It is very easy to just listen to the modulated signals. If you are not able to DIY you can even go and buy Linkwitz's toneburst CD:
Toneburst Test Signal CD
Modulated wavelets are good because you can also play them and listen how they sound with the real speakers in a real room.
If you compare the steady state FFT responses posted earlier, it is near to impossible to tell the perceptual differences from them. And there is a huge perceptual difference between the monopole and dipole when playing signals with modulations. It has become obvious the steady state FFT freq response is useless in describing this.
The modulated wavelet analysis on the other hand clearly shows how the signal behaves, and indicates the big difference between monopole and dipole.
Finally, and most importantly, it correlates very well of what is being perceived.
- Elias
I am wondering what exactly do you mean with "modulated wavelets". Do you mean those cosine shaped bursts of Linkwitz?
No, it's not useless, but there are better methods of visualize the same data. That's what you are doing with your wavelet CSD.
But it is not hard to find important differences in your frequency responses. Below 40 Hz both work nearly the same. Then there's a mode at 45 Hz, still both with same level. But the next mode at 50 Hz is less excited from the dipole. Above the monopole seems to produce a null at the measurement position, while the dipole seems to work very uniform. Then, at 65 Hz, there's another very deep null from the monopole, a hint to a long decaying mode*. Above 65 Hz, both are nearly the same again, with the dipole having a slightly deeper null at 85 Hz. And all this is also visible in your CSD, although the modes are not clearly visible. That's because the used wavelets have to much bandwidth to show it. Reducing the bandwidth will give a better distinction of each mode, while sacrificing time resolution.
I would not damn the frequency response visualization. It is well enough for showing the modal behaviour of a room, and much better suited when designing EQs oder absorbers for single modes.
* I once experimented with notch filters placed on nulls, and it was quite successful. The frequency response was much more uniform across different places in the room. The perception at the listening position was not much affected.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.