The Waveform Loudspeaker Manufacturers Comment page 2
On being transported or having an orchestra transplanted, that is mythology. A speaker's only function is to reproduce pure tone, in that it renders the most plausible recreation of the recorded event as it was heard in the control room over the control monitors from the engineer's hearing and the producer's mandate. Sorry to burst the audiophile's bubble. Even when a perfect speaker is finally built, it may still only sound like a recording in your room, and uninvolving at that. It's only a stereo system—patronize the arts and your community woodworkers, build dedicated listening rooms with wood floors and hollow stud walls with open adjoining rooms as natural bass traps, as well as irregularly shaped cathedral ceilings, anything but uniform.
Dynamics are a function of flat on-axis, flat early reflections, and flat later reverberations. This is accuracy, which also produces near-perfect tone. Dynamics are a welcome byproduct of accuracy.
There are two ways people can go through life—on their feet or on their knees. I have no intention of pursuing a total redesign. Design did not stop in 1987. The lily has continued to be gilded in successive years at regular intervals. Witness the change to even steeper slopes, a tremendous aid to reliability and phase integrity.
I could have suggested an alternate tweeter equalization for your room early on, as one has been available since mid-1988, but you were never in the right space to hear this thing.
Smaller rooms do give a more accurate sense of recorded ambience, but only at a lower signal level. Immediacy and a sense of transport-me-from-chaos, as well as a feeling of spaciousness, can only be recreated by levels approaching the reference concert experience; that is best in a larger room. A loudspeaker is only a device, you can't have it both ways; let's get tone right first. When that happens, all else falls into place. Tone is the priority, and proper tone was #1 on the design list.
So many contradictions, Larry. How can a speaker be flat in musical presentation yet be admittedly more dynamic than anything you've ever heard? High-enders are so caught up in the hottest, newest, biggest, heaviest, most expensive, most number of boxes, most crossover components, the tish/boom cycle, detail, that when something comes along that doesn't draw attention to itself, it's called average-sounding. That puts me in the middle, exactly where I want to be. I take that as a compliment!
Nowhere do you discuss overall tonality of the speaker, the manner in which it makes recordings provide their own character. Instead you have taken an absolutist position. They're coherent, clean, well-balanced, but I hate them.
Or is it you hate what they represent? Something all the burnt-out dealers and audiophiles yearn for, the entire audio industry needs, and everyone is too afraid and protective to stand up and be counted for: standards. Well, here's Ötvös the Caucasian nigger, going through life erect and on his feet. Standards have given us CD, THX, DAT, SVHS, and a lack of them is withholding HDTV. We live our lives based on scientific applications of lessons gleaned from the laboratory to the field. Yet in 1989 it's still a free-for-all in most of audio; how apt, the phrase "dog eat dog." Until standards come, dealers like [Peter] McGrath will never be able to attract the music lover. You see, I read religiously, I just don't believe religiously.
You speak of ranges of frequencies "without due emphasis." If this is what neutrality or transparency is all about, then say so. You say they don't pull you in. In four years, including the first pair ever sold, all are still owned and enjoyed by their original owners. I'm always reading of Betas, Divas, and the other pretenders to the throne being put on the pre-owned market. Waveforms are a final purchase. Whether or not the design team entirely reached its goal is moot at this point. It appears as though the concept of a loudspeaker with a flat waveform output is anathema to the industry.
On the soundstage: Why no comment on how the speaker disappears, the totally holographic image they portray? Is this not the holy grail for so many sound buffs? Completely boxless.
Once again now, good tone begets wide dispersion, which begets some loss of localization, and complete loss in your 25' reverberant soundfield. May as well listen in mono. I regard imaging and localization as the residual artifacts of the recorded experience. You cannot have envelopment and localization at the same time, at least not from a stereo pair of forward-firing radiators. Precious little localization exists in the 100-piece-plus orchestra I've heard, and my lady and I attend often. Imaging and localization are totally engineered artifacts.
With respect for John's measurements, they beg a fundamental question. What is truth? Whose measurements are repeatable and verifiable from location to location? Whom do we believe? Should the audio community accept the measurements of journalists attempting to bolster the technical credibility of their subjectivity, using uncalibrated microphones and laptop computers with the latest trendy FFT programs, total capitalization for a few grand? FFT has always produced an excess of energy in the high frequencies and not enough resolution in the crucial midband.
Or do we lend credence to acousticians and scientists, serious academics, physicists, and intellectuals who've bitten the bullet, paid their dues, received a doctorate, are asking fundamental questions about how we hear, what are the measurable parameters necessary for great sound? People who don't need to protect their livelihood based on what they sell, be it a magazine or a loudspeaker; people who base their findings on tedious repetitive procedures in order to establish real-world truths. These same people work out of labs capitalized well into the millions of dollars. I bought a philosophy, a methodology, a system of knowledge, a way of looking at the world if you will, to which I can account for all your perceptions minus the traditional psychobabble. Do you really think you can attract music lovers to high end, financially able people with fine living rooms, and then tell them they must have these ¾"-thick wires running all over their Persian rugs? The system works fine with ordinary coax and 16- or 14-gauge wire. I will not pander to the connector crazies.
About accentuated ticks and pops, one certainly does hear all the high frequencies with the Waveforms. I've become a little sick of high-end systems with no record noise. Violins sound like violas. It seems that if high-enders don't understand a brightness playback anomaly they simply roll it off. Out of mind, out of sight, out of earshot.
Now to this subject of brightness. Why is the on-axis rise so much more objectionable than all the long-forgotten rising-coil cartridges you've been a part of promoting over the years? Did you even think to try a [Shure] V15V with a preamp with a flat RIAA curve? You didn't ask me. They do exist, I use one at home. This first impression was unfortunately for me a lasting impression, ie, your so-called objective measurements. Sort of like the stick in the water and how it, the water, refracts light. The stick appears bent, therefore it is bent, because my eyes don't lie and I'm not stupid. This is ontological reasoning at its worst.
There is an intentional 3dB boost, not 4dB, in the DEC between 16kHz and 20kHz. It starts to rise slowly at 10kHz. No, the speakers were not broken, and yup, that's how they sound, bright on brightly recorded material, warm, rich, and luscious (wet Harry) on CDs like London's renderings of Charlie and his MSO. Why recordings are bright I shall save for another dissertation. Sorry, you have the old measurements for the earlier 12–18dB/octave effort. Even with the updated system, in none of the documentation we've produced is there a 10dB rise. If you insist on performing elementary room measurements, then be prepared for reflections to color your chart recorder pickup. The closest is in the power-compression curves, which show some tweeter compression at 110dB—big deal.
Brightness comes from a boost in the presence region, 2500Hz to, say, 8kHz. A boost at 16–20kHz is responsible for sparkle and air only. That's all that's up there. Apologies to all those wanting an increased sampling rate to enjoy on the mediocre speakers which die at 16kHz. In a reverberant echo-chamber like yours, a rise of 1.5dB at 60–75° off-axis in the tweeter, which we currently have, in a speaker which is ±1dB elsewhere coupled with bright recordings and seated too far back, can appear psychoacoustically like a 10dB increase, especially since you boys really screwed up my bass. This is what transparency is all about, the 2% flaw overshadowing the 98% correct adjustment. The speaker is not bright in moderate normal rooms, here it zings and sings. Try it, you'll like it.
What you heard is the Achilles heel of the forward-firing format taken to its logical extension. It does not and can never have flat sound power. Off-axis reproduction is always rolling off and, tweeters being smaller and more able to spray energy around, come on in this speaker audibly in large rooms. We are still working on this. You don't have to know everything. This is also a function of interference caused by multiple drivers. Imagine how hot those speakers are, with drivers physically separated on their baffles by great distances. This is why our cluster is close together and overlapping.
And John, the tweeters can be placed anywhere as long as they produce the desired result; ie, from an acoustic standpoint, and not from a marketing position. This is a sad fact of life I've had to find out. This means that there is, somewhere down the road (remember, it's the journey and not the destination that maintains bliss), another dynamic loudspeaker with even more drivers than the M1, and utilizing sophisticated DSP to control interference and time-related room decay. The question begs, who out there has the commitment to build this one right as a technical feat without compromise? Step forward and let me get back to my cellulose and lignin. All the companies I know who could are too market-driven now to do it right. This is why Waveform Research was founded in 1985, to fill the gap, and we're still here as steady as the rock of Gibraltar, ten years ahead of our time.
All enthusiasts know that work never stops on a good design. Audiophiles like to think of themselves as perfectionists, believing they can walk into a store and purchase a new device to cure their most recent cause of consternation. Try being a craftsman, working with natural materials and all their inherent limitations, pick up some cutting tools, go for the vanishing glue line, balance inline with outline.
Tweaks at this level of performance are more evident in the blind tests we employ, for in the field they are soon swamped by gross differences in recorded menus and bloody rotten rooms.
With respect for two dear friends and the best supporters and dealers a new kid could have, I guess Keith [Yates] and Evelyn [Sinclair] have such terrible hearing that over the years of living with the speaker they became so lost rather than involved in its musical presentation that they omitted telling me of these shortcomings.
Now for the boosters club. Most designers know that high frequencies, being short in length, are extremely directional vertically. When mounted on a slightly sloping baffle, 8" above the reference ear level reset in 1987 at 36" down from 39" in 1986, the very highest become rolled off when designing with one mike height in the listening window. A new speaker has emerged on the scene, a favorite son, with an even steeper sloped baffle and a tweeter mounted at least 18" above seated level, no boost in the DEC, it has none. Imagine how rolled off the air will be with it, and it is. It is a given fact that people will be seated while listening to music, usually slumped in an easy chair, and I decided to ask Paul to recalibrate the speaker. The boost is flat on-axis in our lab, drooping slightly one decibel toward 20kHz in the window. For those readers not familiar with the NRC window, we are long since past the rudimentary one-axis measurement to determine sound quality. This is only one of ten measurable parameters employed. The window, another parameter, is a computer-optimized average of 256 step tones spatially averaged to avoid clutter, taken at 0°, 15° left, 15° right, 15° up, and 15° down. The other eight parameters will be discussed in lesson #17.
The position of the woofer close to the floor has little to do with existing room modes since low frequencies are so long and, being omnidirectional, would eventually set up standing waves anyway. Oh goddess Athena, where are you in my moment of need? Many good designs already approved for sale by the magazines have an even closer coupling: Snell AIII, Janis, and surely Velodyne, etc. That comment is uninvestigative fairytailing.
Regarding Larry's sweet seat, aka listening chair: Tell me how you equalize a system for someone making subjective measurements from a rocking chair? I showed at 80Hz a 14dB difference between the forward and rearward extension of this wooden devil.
Re. time alignment: Multiple arrival times are a function of reflections and reverberations in the room. They cannot as yet be controlled; however, there are people working on that too. Show me any two competing designs employing so-called time-aligned crossovers with similar tonal balance. There are none.
I'm sorry, but your measurement methodology has no allowance for room contribution. This is one reason among many why the best designs are now all begun in the chamber, and it's analog too—subtle irony there.
On your finding of a high-Q resonance at 157Hz in James's subwoofer, it isn't worth the paper it's written on. To let you in on another trade secret, the woofer to mid crossover point is skewed to allow for the naturally rising parabolic curve of the sub. It is actually rolling off beginning at 75Hz, the mids rolling on at 150Hz. At 18dB/octave the sub output is already down well below audibility at 157Hz. Besides, high-Q resonances are relatively inaudible anyway. So much for taking the woofer up too high. Me and my big mouth; let the cat out of the bag, didn't I Paul?
So now that all of you know so many of the little things that make the Waveforms so great, let's see how fast you can catch up, make yours dynamic too, and get those music lovers in the store for Peter's sake, but get rid of the wise-guys. For every secret that I tell you here, there are 20 more that you won't know about. You see, I couldn't care less about commercial success. "History will absolve me." Our measurements are on record for all time, and when another speaker comes along as coherent, well-balanced, clean, and dynamic as the Waveform, it will likely be as uninvolving too, but it will measure similarly if we do the taping with our yardstick.
Back to the bass: The manufacture of the 2235H is more consistent than the robotics-produced Philips ribbon (oops, another secret). The distortion characteristics are well known among serious acoustic minds, as is its astronomical price, but it's worth it and that's why I use it. The ears are far less sensitive to distortion in the bass, a function of the Fletcher Munson curves, and you have documented this with the levels enjoyed. Not everyone is sucked in to employ "faster," more transient-perfect 8" subs with their very high levels of distortion. Transient response is a function of frequency and has nothing to do with driver size reproducing the same frequency as long as both do it accurately; ie, flat.
You contrast the Waveform as superior to the Beta in many areas, yet Arnie's midnight creation is still Class A and the Waveform is dogshit. 25' back for the Waveform, 12' back for the M1, how far for the Beta, some comparison, what fairness and objectivity, and with different levels, too.
And now this. How about an objective subjective most-accurate-speaker-in-the world contest? You boys do all the choosing and the listening, but first you have to have your hearing run through an audiogram. Game? Waveform Research will sponsor this. Bring your two Class A speakers, grab a WAMM and IRS while you're at it. The challenge is that the Waveform loudspeaker will clean up in the overall fidelity ratings. But it must be blind, with levels matched and screened to prevent creeping psychoacoustics. Let's make it interesting, Larry, and put a small gentleman's wager on it, say $50,000 US. That shouldn't be too much for someone who recently turned down 5 million bucks for Stereophile. Winner take all, and this time you publish the results of your visit to the NRC. It may prove embarrassing to the review, but what's a little money lost between friends? Probably in another lifetime—eh?
During the past 24 months of the review ordeal, two books of interest to audiophiles have sustained me. The first is Edward Herman and Noam Chomsky's latest work, Manufacturing Consent. The second and vital work is The Power of Myth, by Joseph Campbell with Bill Moyers. Surely Mr. Campbell was one of the greatest American freethinkers in recent times. He mentions how "passion is what destroys reason, and in politics the principal passion is greed." I urge all self-respecting readers to obtain these works and apply the lessons learned to audio.
Returning one final time to this adjective "uninvolving." Recently John Jensen of VIFA visited with me. He asked for orchestral music with combined voice and instrumentation. His comment was, "Amazing you don't have to strain and lean forward in your chair to discern all the different parts. They're all there and in perfect harmony with each other." If uninvolving means the stereo system has disappeared, then isn't that what the music lover wants? Perhaps this is why high end doesn't sell to music lovers, their systems are too involving.
To close now, I trust that sometime in the future, when, Larry, both you and John see the digital dynamic light, we'll all become friends—there is much we can teach each other. Your magazine has improved immensely over the years. After all, when it's time to go, all one should leave behind is "good memories, good friends, and good genes," and some of us would say good speakers, too!—John Ötvös, Woodworker
JA responds: As always, I must point out that it is not my policy to respond to "Manufacturer's Comment" letters. However, John Ötvös raises a number of points in his cry of anguish concerning political and ethical issues that do require clarification.
Larry's comments on the Waveform's sound in the November issue have no connection with the opinions of Peter Aczel, Harry Pearson, or any other reviewer. They are based solely on the speaker's sound in his room.
There is no relationship between the outcome of a review in Stereophile and whether or not the manufacturer advertises. This is a cheap shot from an angry man.
Having auditioned the Waveforms, Tom Norton subsequently declined to review them. He is writing a follow-up.
The bass level was adjusted to give a measured flat response. The 25' listening distance was not LA's normal choice but was actually suggested by Mr. Ötvös during his visit, as was the near–room-boundary speaker placement. Regarding Mr. Ötvös's puzzlement over Larry's reluctance to request the revised crossover with the less exaggerated ribbon tweeter level, a) both Larry and I pointed out the top-octave boost during his visit, only to be reassured that, yes, this is how the Waveform is intended to sound and measure in the high treble, and b) no mention was made at the time that such a revised crossover existed.
Our measurement microphone is a calibrated B&K 4006. The crossover's 4dB boost in the top octave was measured with an industry-standard Audio Precision System One. The only-too-audible rise in the top octave was apparent both in the anechoic FFT measurements, taken with the B&K microphone, and with the Audio Control Industrial SA3050A real-time analyzer, using its calibrated microphone. The room effects on the measurement postulated by Mr. Ötvös are therefore irrelevant.—John Atkinson