Some of D edwards questions or statements, and Duke's responses:
D edwards: How can a speaker with a marginally controllable fullrange effects/reverb system win a clarity contest?
Duke: This is a good question, and gets right to the heart of the matter.
Clarity can be degraded by resonances in the driver, enclosure panels, or enclosure internal airspace; by resonances within the room; by reflections (in particular strong, distinct ones - the technical term is “specular”); by frequency response aberrations; by severe distortion of the initial waveform; and by diffraction. There are probably other mechanisms that don’t come to mind right now.
In this discussion we’re focusing on the backwave energy of the SoundLabs, but note that they don’t have any kind of box resonance; they don’t generate floor and ceiling bounce reflections; they have a very smooth frequency response; they are inherently time and phase correct; and they are directional enough that they don’t generate diffraction artifacts. They do have “drum-head” resonances within the individual cells, but the size of these cells is staggered to spread out these resonances so they don’t sum to an audible peak.
Now note that all loudspeakers generate off-axis sound (and therefore reverberant energy). If reflections can degrade clarity, then one could argue that all else being equal the speaker with the highest direct-to-reverberant sound ratio is likely to have the best clarity. Wouldn’t you agree?
A few years ago I measured the output of a point-source-approximating speaker (6.5 inch two-way) and a SoundLab M-1 at a distance of 1 meter, and again back at 8 meters, in a living room. I used pink noise, with NO time gating, so that the room’s contribution would be fully included.
Anechoic theory predicts that the output of the point source speaker would fall off by 18 dB going from 1 meter back to 8 meters (three doublings of distance). Anechoic theory predicts that the output of a line source speaker would fall off by 9 dB over that same distance. So if the in-room SPL falls off by less than anechoic theory would predict, that difference must come from the reverberant field contribution. Wouldn’t you agree?
The point source speaker’s SPL fell off by 11 dB over that distance, which is 7 dB less than anechoic theory would predict. So the reverberant field was contributing 7 dB back at 8 meters.
The SoundLab’s SPL fell off by 4 dB over the same distance, which is 5 dB less than anechoic theory would predict. This means the point source speaker’s reverberant field was contributing 2 dB more than the SoundLab’s was, even INCLUDING the backwave! That translates to about 60% more reverberant field energy from the point source speaker.
This proves that SoundLabs generate less reverberant energy in the room than a conventional speaker does. Now I’m an advocate of diffusing that backwave energy so that you don’t get a specular reflection, and once it’s diffused it doesn’t have the characteristics that would audibly degrade clarity. (Helmut Hass, “The effect of a single echo on the audibility of speech”, Journal of the Audio Engineering Society, 1950's [as I recall]).
D edwards: Psycho-acoustics? please, thats a pretty broad stoke.
Duke: A late-arriving, diffuse, slowly decaying reverberant field has been shown to be desirable from a psychoacoustic standpoint. (Pisha and Bilello, “A Live-end/Dead-end room, Part I”, Audio magazine, mid 80's.)
D edwards: Our disconnect is you cannot articulate any quality experiences with the Manger, then that is where our miscommunication is going to remain.
Duke: Whether or not I articulate any quality experiences with the Manger has nothing to do with this discussion. I’ve heard the Manger three times that I can remember. How in the world does that relate to a discussion of a completely different speaker? I’m not engaging in a “my speaker can beat up your speaker” debate.
D edwards: I don't want to talk negatively about Sound Labs and other Dipoles but they have issues which hurt them in areas like clarity, even though some fake clarity very well, its just that, a fake. You're a speaker expert I'm sure I don't have to explain how its done.
Duke: Sorry, but I’m not that much of a speaker expert - how do you fake clarity? Maybe you mean by tipping up the high frequencies? I’ve seen realtime unsmoothed wideband frequency response measurements of the SoundLabs, and they have a gently downward-sloping frequency repsonse curve - no artificial treble boost. The measurement was not time-gated, and it was the smoothest unsmoothed curve I have ever seen by a considerable margin.
D edwards: A multichannel surround system clearly will demonstrate that although pleasing this fixed dipole "reverb" clouds and homogenizes the image of a dipole speaker like the Sound Lab. You just want to overlook the detrimental aspects of slapping a fullrange signal off some drywall and assume psychoacoustic will explain this very un-ideal situation into a positive, I can't imagine anyone not thinking this is very very optimistic.
Duke: I've addressed the drywall issue (use diffusion). In my opinion the theoretical ideal would be an anechoic room with all of the reverberant energy you would have heard in the original venue reproduced by dedicated surround channels. Maybe one day we’ll have the recordings and systems to go there. Also as mentioned earlier, I advocate diffusing the rear channel energy - especially if it’s going to arrive within less than 10 milliseconds of the direct sound (Richard C. Heyser, review of Quad ESL 63, Audio magazine, early 1980's).
D edwards: I think what you failed to take into consideration is these comments were in reference too the Manger driver, a vastly superior drive element to the technology used in the Sound Lab. And because it appears you have little or no idea what the Manger can do you take my comments as a big insult to the Sound Labs, well it wasn't, because the Sound Lab has to do very many things extremely well to simply lose out to the Manger on clarity.
Duke: I do not comment on your claims about the Manger’s performance. As I said, I’m not engaging in a “my speaker can beat up your speaker” debate.
D edwards (starting out by quoting Duke): "To the ear, the backwave of a dipole is not 100% distortion; rather, it is reverberant field energy. If reverberant field energy were undesirable, concert halls and recital halls would resemble anechoic chambers"
Complete false analogy and misunderstanding of how the backwave is working and the difference between reproducing sound through speakers and listening to live sound.
Duke: I’ve already responded in part, but if you’ve ever seen the microphone placement at a recording of a live performance in a large hall, you’d know that the microphones are positioned to pick up a much higher ratio of direct to reverberant sound than what people in the audience would hear. The recording is made with the expectation that some reverberant energy will be added by the playback room. And in the interest of recreating a natural sound field, that reverberant energy should come from all around, not just from the two loudspeakers (which would be the case in an anechoic environment). (Jens Blauert, “Psychophysics of Human Sound Localization”, a psychoacoustics text).
D edwards (starting out by quoting Duke): "SoundLabs are exceptionally good at getting the reverberant field right, something that live voices and instruments routinely do but few loudspeakers emulate."
Can you answer how they get the reverberant field right, or is this just a quantification of your experience. How can something that makes everything sound 7 feet tall be getting the reverbent field right? Maybe you just like the extra reverb on your music....singing in the shower type thing?
Duke: The ear derives directional cues (including image height) almost entirely from the first arrival sound, not the reverberant field. This is very basic. You might want to look up something called the "precedence effect".
Set up correctly, SoundLabs do not make everything sound like it's seven feet tall. If the panels are exactly vertical, the image height will be at ear height, whether you are sitting or standing. This is because the shortest path (first arrival sound) will be to that part of the panel that's at ear height.
A live instrument in a good hall gets the reverberant field correct, wouldn’t you agree? What then are the distinguishing characteristics of that reverberant field?
1. There is a considerable time delay between the first-arrival sound and the onset of reverberant energy.
2. The reverberant field is highly diffuse
3. The reverberant sound started out with the same spectral balance as the first-arrival sound, and the only modification to this spectral balance has been done by the room’s natural acoustic characteristics.
4. The reverberant field decays fairly slowly.
Now some of these characteristics are largely room-dependent, but number 3 is highly speaker-dependent and in a given room number 1 is controlled by the speaker’s directional characteristics (in the case of the SoundLabs, the time delay imposed by the distance off the back wall delays the arrival of the backwave energy that much longer than would have been the case if it had been included in the frontwave).
D edwards (starting out by quoting Duke): "A spectrally correct reverberant field is conducive not only to natural timbre, but also long-term fatigue-free listening. I can explain why this is so if anyone is interested."
Correct timbres relating directly to the frequency response of your drywall....no need to explain. Theory is nice but reality is difficult and non-linear.
Duke: Actually, I think there is a need to explain at least a little. The ear derives timbre from both the direct and reverberant sound. At normal distances in a semi-reverberant room, most of the energy that reaches your ears is reverberant energy.
The listening fatigue part would take me a long time to explain. I’ve done so in other threads, so will simply post a link:
http://www.audioasylum.com/forums/hug/messages/101458.html
D edwards (starting out by quoting Duke): “Note that a line source speaker like the SoundLabs generates a much higher ratio of direct to reflected sound energy at the listening position than does a typical direct-radiating point source speaker, even factoring in the backwave.”
Over what listeing window or to be specific gate time? Over a very short window yes, but open the window wider and here comes the 2nd blip....what to do with all that maybe correlated maybe not correlated signal?
Duke: See the measurements I posted above, which were not time-gated. They demonstrate that the relative reverberant field contribution is greater with conventional speakers.
I apologize for not posting exact dates and/or page numbers in the sources I've cited, but I'm going by memory and not writing a thesis paper here.
In an earlier post, D edwards wrote:
"Maybe you should measure your Sound Labs and see what they really do like I have, then maybe we can talk about it."
I've talked about some of the measurements I've made on SoundLabs. So now let's hear about the ones you made - tell me "what they really do".
Duke
