I am not sure that I can explain. Its interesting data and results, but I am not sure that I do know what is going on.
I certainly would not consider times on the order of 20 us as relavent, or do you mean 2 ms?
My point is simply home much the room changes any and all characteristics of the source at LFs. Free field characteristics of a LF source just do not have much relavance to its use in a real room.
I certainly would not consider times on the order of 20 us as relavent, or do you mean 2 ms?
My point is simply home much the room changes any and all characteristics of the source at LFs. Free field characteristics of a LF source just do not have much relavance to its use in a real room.
In post #444, the second graph from the bottom titled "Raw Impulse responses Monopole top, Dipole bottom" was actually a bit confusing to me. For the monopole the region from the impulse upto .02ms looks crowded, but for the dipole this is clean. However after .02ms there is pronounced ringing in the dipole response; much more than the monopole response.
Does this mean that the dipole is cleaner before .02ms compared to the monopole. However after .02ms the monopole is cleaner....is this the reflected soundfield because of room reverberation.....but if the room is the same, so should be the reverberation time. In any case after .02ms why is there more overhang in the room for the dipole than the monopole.
Please explain. Thanks.Actually this is only comparative graph which is distinctly different b/w the monopole and dipole.
We don't know anything about Barleywater's system so hard to draw conclusions.
It looks like though that the time axis is in seconds. Then the dipole resonance at 130 Hz would match the time scale ringing.
I think the 130 Hz peak can simply be a H-frame cavity resonance.
- Elias
I don't understand how the corrected impulses in post 444 appear to be perfectly symmetrical. The system would have to be linear phase as well as flat response for that to happen. even assuming the unequalized woofer is minimum phase, amplitude correction would still leave it as a minimum phase band pass device which should show a casual impulse with some post peak oscillation. It doesn't make sense to me.
Just reading from the graph, the reading in the x-axis is ".02" and the unit labelled as "hms". I am not sure if this means .02 seconds or 2 milliseconds OR .02 milliseconds
The room responses after equalization is eerily similar with NO difference
Even with the same diameter drivers, equalised to same LF extension, listened at moderate SPL, to even nearfield (2-3 Meters) there is that slight overhang in sealed units that makes it sound like a "speaker" which is not apparent in a dipole(H or Uframe) making it sound that last bit different ... or "natural'. No scientific logic/evidence "yet" discovered....Wonder if it is the lower group delay in the dipole which is small yet audible
Reverberation does not need to be "scientifically re-discovered". When you hear it from around, coming from different directions, it is reverberation that gives you an information about the space in which the music sounds. But when you hear it from some point source it is "speaker" sounding, while room reverberation still exists, and sounds like "Speaker In The Room". You hear 2 different volumes, one (room) from around as reverberation, another from speaker like "speaker sound". But if to blend them together in space around head and ears it sounds more natural. It is exactly what dipoles do, and why they sound more natural.
Just as bad is what appears to be some DC content in the corrected impulses. That's not possible for an acoustic pressure response, especially for a dipole. Something isn't right.
This is what this thread is all about...What EXACTLY (and how ) do dipoles do to make them sound more natural ( at least to some...who go to great extents at higher costs to expand their dipole systems and never come back to monopoles again).
This is what this thread is all about...What EXACTLY (and how ) do dipoles do to make them sound more natural ( at least to some...who go to great extents at higher costs to expand their dipole systems and never come back to monopoles again).
Let me just say here that the majority of my customers last speakers were dipoles, so your claim is simply not true, its just your perception.
This is what this thread is all about...What EXACTLY (and how ) do dipoles do to make them sound more natural ( at least to some...who go to great extents at higher costs to expand their dipole systems and never come back to monopoles again).
It is one half of the truth: dipoles in untreated rooms sound as if more natural. That does not mean that they more naturally reproduce records, so room treatments are applied. But if to treat the room this advantage goes away, and distortions caused by higher excursion start dominating.
Post #444: No DC. Corrected response view is short window. Measurement was made with exponential sweep 2-2205Hz, 10485760 samples long 44.1kHz sample rate. Drive voltage was 26V RMS. Sample rate conversion to 4410Hz sample rate (1048576 sample) was used for IR recovery and spectral display.
Correction filters are FIR, results are linear phase. Correction applied to continuous signal results in highly coherent passage through system. Recording of band pass filtered square waves look much like test signal.
Early reflection 2ms is seen in raw monopole IR. Extended tale of raw dipole response is primarily dipole path length and associated cavity behavior.
Regards,
Andrew
Correction filters are FIR, results are linear phase. Correction applied to continuous signal results in highly coherent passage through system. Recording of band pass filtered square waves look much like test signal.
Early reflection 2ms is seen in raw monopole IR. Extended tale of raw dipole response is primarily dipole path length and associated cavity behavior.
Regards,
Andrew
I'm having a bit of trouble with the plots of #444 also. The 3 EQs are essentially the same even though the unEQed responses are considerably different. Not much correction going on. (Or is this the response with correction??) I can see that the EQ time domain (is this EQ alone or woofer plus EQ?) is essentially a linear phase filter with the sin x/x ringing. Still, a linear phase correction convolved with the minimum phase woofer should look more like the previous time response just with more added pre-ringing.
Some more explanation would help.
David S.
Some more explanation would help.
David S.
Last edited:
One thing I've wondered about 3 distributed dipoles vs 3 distributed monopoles:
If I understand correctly each dipole can be modeled as two monopoles with a distance D, and that if you distribute them and use EQ the in-room response can be made equal or close enough to equal.
What I wonder however is if this also is true for bass leaking outside the room? I listen fairly quiet and so reduced SPL won't be that much of a problem if I use them as dipoles, the main focus is to reduce noise for the neighbours.
If I understand correctly each dipole can be modeled as two monopoles with a distance D, and that if you distribute them and use EQ the in-room response can be made equal or close enough to equal.
What I wonder however is if this also is true for bass leaking outside the room? I listen fairly quiet and so reduced SPL won't be that much of a problem if I use them as dipoles, the main focus is to reduce noise for the neighbours.
I would think that sound leakage would be a function of sound power of the sources. That is certainly the way that room acoustics people model it. We also know that the directivity difference between a dipole and a monopole is 4.8 dB (meaning a dipole has 4.8 dB less power response for the same axial SPL. Viewed another way the monopole is 4.8 dB louder on axis for the same radiated power).
This gets us back to the universal question of whether we set the systems up with the same axial SPL or the same power response? I tend to think we would (at LF) set for about the same power response and then there would be no difference for the neighbors.
David S.
This gets us back to the universal question of whether we set the systems up with the same axial SPL or the same power response? I tend to think we would (at LF) set for about the same power response and then there would be no difference for the neighbors.
David S.
Room acoustic properties (either mean alpha or room constant) and radiated sound power would define the SPL in the room (the transmitting room.) TL or wall transmission loss would determine the energy loss to the adjacent room. Finally the room acoustics of the adjacent (receiving) room would determine the SPL in that room.
If flanking paths exist it gets more complex but that is generally how the calculations go. I'm not aware that Schroeder frequency is ever considered but most acoustical tests are ideally done in large rooms with added diffusion.
David S.
If flanking paths exist it gets more complex but that is generally how the calculations go. I'm not aware that Schroeder frequency is ever considered but most acoustical tests are ideally done in large rooms with added diffusion.
David S.
John - correct, the leakage would be independent of the type of source used to create the SPL. This assumes that the SPL's in the room are adjusted to be the same. A monopole can excite a very low resonance that a dipole cannot and it is these very low tones that tend to be the leakiest. Hence, in practice, the higher efficiency of the monpole would tend to have more leakage SPL.
Regarding post #444, I too was bothered, initially, but I have concluded that the results shown could be possible, so I have nothing on which to conclude that they are not correct. If an adaptive DSP was setup to create a bandlimited linear phase result, then it could do that, exactly as shown, for any source.
In that case, it is an extremely interesting result, completly proving the point that EQ can overcome any differences in source types. LFs in a room are not about the type of source.
Interesting discussion!
I've always wondered about the transmission of sound at low frequencies. At higher frequencies (above the Schroeder-frequency) I assumed it would be mainly a function of radiated power, but at lower frequencies I would assume standing waves play a role: at frequencies where there are standing waves, the pressures at the room-boundaries are the highest and thus walls would flex more and there would be greater sound-transmission. Of course this is all ignoring the mechanical/acoustical properties of the room boundaries. Could anybody fill me in?
I've always wondered about the transmission of sound at low frequencies. At higher frequencies (above the Schroeder-frequency) I assumed it would be mainly a function of radiated power, but at lower frequencies I would assume standing waves play a role: at frequencies where there are standing waves, the pressures at the room-boundaries are the highest and thus walls would flex more and there would be greater sound-transmission. Of course this is all ignoring the mechanical/acoustical properties of the room boundaries. Could anybody fill me in?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.