Measured monopole and dipole room responses

[Elias]

Originally Posted by Rudolf

I have done that with my earlier dipole speakers. Significant altering of modes will only start if you rotate the dipole at least ~45° against the mode axis. You don't want to switch off single modes, but probably want to attenuate one mode and raise another mode at the same time. Inevitably that led me to a 45° toe-in of the LF dipoles - every time. And it did change the direct sound significantly!
At 45° an ideal dipole should be 3 dB attenuated compared to 0°. If you keep the mid/high part of the dipole at 0°, you have a matching problem. JohnK discussed that at Power matching
Again the easiest way to solve it is to rotate the mid/high part of the dipole too - which has some advantages in itself. In the end I never had a convincing reason to rotate the dipole LF part independently from the upper part.

Rudolf

I've also done few tests with dipole rotation in the past, and my experience is somehow similar to yours. While rotation can help, why not, in the few cases I tried it the optimum rotation was, be surprised or not, almost the same as the direction of the main speakers i.e. aimed to the listening area.
So at the end it became feasible to just have a fixed panel.


- Elias

[gedlee]

Originally Posted by Elias

I made it with Octave, how else
- Elias

There are lots of ways to do this. I did it in FORTRAN more than 30 years ago and more recently a much bigger model in MathCAD. Octave is hardly the only choice.

Are the actual impulse responses posted somewhere. I'd like to see what they look like.

[markus76]

Originally Posted by Elias

Here's the FFT frequency responses of the same impulse responses as before.

This is what you base your opinion upon that dipoles sound better than monopoles? Those responses are both really, really bad. Put a sub directly behind your listening position at shoulder height and get a frame of reference what good low frequency reproduction sounds like.

[Wavebourn]

Originally Posted by gedlee

And some prefer to hear music the way that it was recorded - you know "Hi-Fi"!

That's why I am going to add some book shelves instead of curtains. Too many curtains for a living room, they cover memory foam mattress toppers on walls.

I would not explain differences of sound of monopoles and dipoles in the room by different frequency response. Dipoles rather somehow blend reverberation of rooms with reverberation reproduced by stereo speakers making kind of more realistic soundstage. But more realistic does not mean closer to real soundstage that was recorded.

[Elias]

Originally Posted by markus76

This is what you base your opinion upon that dipoles sound better than monopoles? Those responses are both really, really bad. Put a sub directly behind your listening position at shoulder height and get a frame of reference what good low frequency reproduction sounds like.

That was just to show that in an irrelevant metrics they can look whatever they do. Steady state unwindowed FFT frequency response does not correspond how we perceive music signals which are constructed of temporal modulations.


- Elias

[Elias]

Originally Posted by gedlee

Are the actual impulse responses posted somewhere. I'd like to see what they look like.

What is there to see. Humans do not perceive impulse responses.

[markus76]

Originally Posted by Elias

That was just to show that in an irrelevant metrics they can look whatever they do. Steady state unwindowed FFT frequency response does not correspond how we perceive music signals which are constructed of temporal modulations.


- Elias

If the magnitude response would be flatter then there wouldn't be as many time problems either. Magnitude response is not irrelevant, it's intimately connected to the time domain.

Nobody has said that problems in the time domain are inaudible. Up to this point all you're showing is a very bad monopole response and a very bad dipole response - caught between a rock and a hard place.

[markus76]

Originally Posted by Elias

What is there to see. Humans do not perceive impulse responses.

Err, your wavelet analysis is based on impulse responses, no?

[soundaatma]

Seems that this thread is dying once again without reaching any conclusions about the best among : sealed vs cardioid vs dipole, for optimum realistic bass reproduction in acoustically large rooms.

I guess we have to wait for John K to carry out his independent comparisons using Elias's software and methods.

[gedlee]

If you are waiting for a conclusion that everyone agrees to, that is never going to happen.

I have reached my conclusion. I posted it awhile back. Not what anyone wanted to hear, but hey, I not here for fanfare.