I consider myself an equal opportunity offender in that I like absorption, reflection & diffusion equally...just that the right widget has to be in the right place.
You's my kinda dude Jeff!
I consider myself an equal opportunity offender in that I like absorption, reflection & diffusion equally...just that the right widget has to be in the right place.
When I was first exposed to this in a presentation by Alan Devantier (Harman/Infinity designer), I could not believe what I was hearing. I thought I knew acoustics and then I am sitting there gasping as I am told these things. It just made no sense that a reflection would be better than none. It take a 2-year journey of extensively reviewing the literature, sitting through a ton more presentations, arguing the points with others , reading Dr. Toole's book over and over again, and digging through 50 to 100 AES and ASA papers that I realized the amazing consistency of it, and lack of disclosure in public.
Amir in his article said:So what seemed like an open and shut case of eliminating wall reflections due to anomalies in frequency response of the room becomes much more complex when one considers how we hear sounds in our home listening spaces. It shatters “gut feelings” one might have about the problem and solution thereof. I don’t know about you but I am fascinated by all of this. It is not every day that we get to like some distortion and save money not trying to eliminate it! So complexity and deep understanding of the science does have its virtues.
What I find interesting is that for every Martin Logan or Soundlab owner that likes a speaker with a curved panel to beam the sound...
...you will always have a "science minded" person like Sanders who likes a flat, non-beaming panel - and will have a "science reason" to prove it works best.
Very very interesting post...I read it three times (as a non-techie) to get thru it and make sure I was following your description of sound waves and also the Sound Labs approach as well. I trust Microstrip's ears and know he has long enjoyed those speakers. Your description of their design approach is consistent with how people have described their sound relatives to 'live'.
I note that the new Gryphon Pendragon speakers are designed with a 6' ribbon for mids without any crossover, plus a series of tweeters of some kind...and then a separate set of tower subs with 8 cones that are self powered and actively crossed over. Also Gryphon mentions some of the characteristics of which you speak. They also specifically recommend setting them up in a very wide circular alignment where the curve of the circle uses the listener as the center of the circle.
Good stuff, Duke. I don't know Soundlabs (heard a pair once), but I've spent a lot of time listening to a variety of Martin Logans. If ever there was a speaker that screamed for room treatment to kill first reflections, MLs are it. The tonality changes radically, and not attractively, as you walk toward either side, so you know those first reflections are not going to mix well with the direct sound in the room.
I'm with Toole and Linkwitz; give me a speaker that's linear across the broadest possible field and it'll sound good in most rooms, more natural in all of them. But one thing sometimes gets forgotten; these kinds of speakers excite our listening rooms well. That makes for a more natural presentation, a mix of direct and reflected sound that is closer to what we are accustomed to hearing in life, but neither the speakers nor the room know anything about the concert hall or the recording studio. That ambience may or may not have been captured in the recording, but it can only be altered by the rooms we play them in, not brought closer to the original reality.
this is really great stuff. Thank you for taking the time. On the Gryphon Pendragon, could you provide a little bit more detail about why you think their particular hybrid approach is a good one? Specifically, in reference to their use of woofers and panel...the points was made in an earlier post about the differences in spl fall off over the same distance, and there was some concern that the panel spls will drop off more slowly than that of the woofers. What about the Gryphon approach do you like?
And do those reasons apply to other 4-box speakers like the old Genesis, Infinitys, and even today's newer Tidals?
sorry to bombard, but your explanations are clear and appear to come from good engineering (not that I'd know!)...but seriously, much appreciated.
In my opinion there are other radiation pattern shapes that can outperform "broadest possible field", but that's not a disagreement with Toole and Linkwitz over the basic principles - just an alternative way of applying them. And come to think of it, Toole is doing monopoles while Linkwitz is mostly doing dipoles, so even between the two we have different pattern shapes represented. I do controlled-pattern bipoles, which combine some of the characteristics of both, but consider my designs to take a back seat to the SoundLabs as far as ideal room interaction goes.
I wasn't clear; I didn't mean the broadest possible field of dispersion from the speaker, I meant the broadest possible linear field within the speakers' existing dispersion. Aren't Linkwitz and the Harman boys both going for good linearity throughout the field? My non-technical mind gets that; they're not suppressing reflections, they're using them, but making sure they sound more like the direct sound, more like real instruments would sound in the space. Where they part, I think, is in how, and how much reflection they're using, in how they're manipulating the illusion?
Aren't Linkwitz and the Harman boys both going for good linearity throughout the field? My non-technical mind gets that; they're not suppressing reflections, they're using them, but making sure they sound more like the direct sound, more like real instruments would sound in the space. Where they part, I think, is in how, and how much reflection they're using, in how they're manipulating the illusion?
And there, they are making a subjective choice, I think, because it is an illusion. Real instruments simply don't disperse like speakers, or even like each other. The radiation pattern of a cymbal, and how linear it remains as you move off-axis (there really is no off-axis for a cymbal) is radically different from an acoustic guitar, a trumpet, the whole drum kit. We really can't get all that close, unfortunately, but maybe we can understand that real music is played in real rooms, not padded cells, and let playback breathe a bit.
Glad to have you here, Duke. And please understand that all of the above is one big inquiry. I don't pretend to know; I'm just trying to draw some conclusions from what I hear.
Tim
Duke what is your opinion of tall main speakers plus sub or multiple subs (instead of array of woofers)?
Thanks for posting to correct my misperception.
By "good linearity", I take it that you mean something like "smooth net frequency response of the summed on-axis + off-axis energy" - is that close?
Yes, that's my understanding - one might say they are trying to work with the room instead of against it (though I don't think any speaker designer deliberately sets out to work against the room). I think they're barking up different sides of the same tree. In my opinion Harmon's wide-pattern-monopole approach is somewhat more speaker-placement-friendly, while Linkwitz's mostly-dipolar approach can get some nice delay on the backwave energy with proper placement. Neither one is really trying to bounce the first sidewall reflection off the opposite side wall - Geddes does that, and I think it has merit, so I do it too.
Agreed. All we can do is carefully consider (and where possible sample) the poisons available, and pick the most palatable one. However I do believe that the secret to audio nirvana (along with the secret to life, the universe, and everything) lies in doing a really good job off-axis.
Thank you very much, Tim! Please inquire away as much as you want. We're now talking about my favorite subject in the world of speaker design. I'm really glad to see Toole's work widely discussed here - imo that sets this site apart from most.
I mentioned Geddes (my mentor) earlier, and he has his own approach to getting the reverberant field right. What he does is employ aggressive, uniform radiation pattern control so that the speaker's pattern is 90 degrees wide (-6 dB @ 45 degrees off-axis) over most of the spectrum. He then toes the speakers in severely, such that the axes criss-cross in front of the listening position. This widens the sweet spot, and also bounces each speaker's first strong sidewall reflection off the opposite wall. This is desirable from the standpoint of how the ear/brain system processes reflections, and results in an increased perception of ambience and envelopment in a large acoustic space. Geddes likes to use an unusually live, diffuse room so that the spectrally-correct reverberant energy his speakers generate bounces around for a long time, and takes a long time to decay.
I would think this would completely bugger (the technical term) horizontal imaging!
Tim
(...)
I'm a major advocate of distributed multiple subs, either a la Welti or a la Geddes. Interestingly, they came to similar conclusions about the smoothing benefits of multisubs (imo smooth bass = fast bass) totally independent of one another but at almost the same time - Welti preferring symmetrical placement and Geddes asymmetrical, but both are very respectful of the other's work. I build a four-piece multisub system, using Geddes' original concept with his permission; he has since moved on to using three independently equalized subs.
Mine is definitely a one-listener set up -- near field. I occasionally watch movies on the system, and a center is not missed at all. Got to say I really enjoy the very precise pinpoint imaging you get from these kinds of speakers in a near field set up, even though I know it's not exactly what happens to audio in a room. It kind of replaces the sense of placement you get from being able to see where the instrument is in the plane. You've really got me curious about different setups, though. I may have to drag my speakers back out into one of the larger rooms in the house again and play around some....
Tim
Duke,
As far as I remember Geddes uses different eq and settings for each subwoofer, while the original Todd Welti approach used a single controller for all subs - this could explain in part why they differ so much in placement suggestions.
I remember Amir posted about a new JBL Synthesis system for optimal subwoofer setting but was proprietary.
Actually the curved panels of the Martin Logans and SoundLabs are intended to reduce beaming.
The curvature of the Martin Logans is a continuous curve, and I'm not sure what angle it covers, but perhaps in the ballpark of 20 to 30 degrees. Because it's a continuous curve, it behaves like a slice of an expanding cylinder: As the diaphragm moves forward, it is tensioned. As it moves backward, the tension is relaxed. The tensioning of the diaphragm as it moves forward puts stress on it. The solution is to make the curve gentle, and to use a very strong diaphragm material.
The SoundLabs use a faceted-curve diaphragm made up of many flat vertical facets. The facets are dimensioned and angled so that they give a wide continuous coverage with no venetian blind effect in the audible range. SoundLabs have a pattern ranging from 45 to 90 degrees wide, depending on the model, and so they actually produce an unusually wide sweet spot.
Not to take anything away from Roger Sanders (who actually invented the continuously-curved diaphragm electrostat), but there is no lack of science in the design of the SoundLabs, nor presumably of the Martin Logans. And the flat panel that Sanders uses beams significantly. This gives fantastic imaging for the person in the sweet spot, at the expense of both imaging and tonal balance for people outside the sweet spot.
Of particular relevance to the topic of this thread is the radiation pattern of the SoundLabs. First a bit of background, which I think will be consistent with Toole: A diffuse, relatively late-arriving, spectrally correct, well-energized reverberant field is perceptually desirable. The less spectral discrepancy between the first-arrival sound and the reflections, the better. Following this paradigm, Toole and/or his associates have designed monopole speakers with very wide, uniform patterns, like the Revel Salon series.
The SoundLabs approach this ideal from a different angle. First of all, the radiation patterns of their 90-degree models may well be the most uniform in the business. They start out as a figure-8 at low frequencies, and then as we go up in frequency and the panel's directional characteristics take over, we have a 90 degree pattern front-and-back. So there is very little discrepancy between the pattern at low, middle, and high frequencies. I do not know of another technology that does such a good job of this. You can listen to SoundLabs from outside the room, with no line-of-sight to the speakers, and except for a little bit of high frequency rolloff (because short wavelengths are more easily absorbed as they bounce around a room), the tonal balance is correct. The tonal balance is reminiscent of listening to a live band from outside the room the band is playing in.
SoundLabs like to be positioned out in front of the wall a bit, and I normally recommend 5 feet or more if possible. This imparts a good 10 milliseconds or so delay before the backwave energy arrives at the listening position, which is a longer time delay than we get from most speaker's early sidewall bounce in most rooms. The SoundLabs beam like giant lasers in the vertical plane so the early floor and ceiling bounces are eliminated, and if we toe them in fairly aggressively we avoid the early near sidewall bounce as well (and the first strong sidewall bounce of the left speaker is off the right side wall, and vice versa, which is desirable). Now the typical near sidewall bounce is usually beneficial (according to Toole), but in exchange the toe-in not only bounces the frontwave off the opposite side wall, it also aims the backwave off to the side, so that now much of the backwave energy will be coming in from the forward-side quadrant of the listening position, but with a nice long time delay. I believe that this is a more desirable presentation of early reflections than what we get with conventional speakers, and more closely approximates what we experience in a good recital hall. Of course the SoundLabs are not right for every room, set of listening preferences, or budget - but they are an interesting design to analyze.
So imo in the design of the SoundLabs we find a great deal of attention being paid to exploiting the psychoacoustics of room reflections, which happens to be the topic of this thread.
Duke, brilliant stuff again. Like the Classics in my library, I can re-read your posts and learn new things every time.
Do you recommend (a) any, (b) some, (c) a lot, or (d) no room treatments for speakers like Soundlabs? for point source speakers? And why?
I know MBL doesn't recommend any room treatments, but I would imagine sticking some in corners (wall/ ceilings/ floor) intersections (vs. all over the room) would have beneficial bass effects while preserving the ambiance the designer worked so hard to create. Thanks!
Steve Williams Site Founder | Site Owner | Administrator | Ron Resnick Site Co-Owner | Administrator | Julian (The Fixer) Website Build | Marketing Managersing |