Paper tigers

September 15, 2018
 by Paul McGowan

There are perhaps as many ways to design a new product as there are new products. Our process is likely unique enough to warrant a few words.

When we start a new product the first phase is almost all done on paper, often the erasable whiteboard. These are hazy sketches perhaps of the chassis or the general internal layout. This might seem counter-intuitive as if we're designing how it looks before we actually have something that works.  But, in fact, we have done this kind of work for so many decades that we have a very good idea of how it's all going together. These first sketches help us understand if what we're proposing is even going to fit into the space it will eventually live in.

Then it's on to more and more detailed sketches: schematics, choice of components, wiring techniques, connector layout, etc. All this occurs well before we even start the actual design. Once the concept piece is fully modeled the physical process begins and from there (following endless meetings) prototypes are crafted and built. It is only then we start the long-term hands-on verification process in the lab as judged by our measurement equipment and then it goes into listening where we finally discover if all our hard work has paid off. Sometimes the listening room sends us back to the whiteboard.

Take for example our new line of loudspeakers. All the work that's gone into these has been done on whiteboards, paper, and computer modeling. From our new drivers to cabinets, these structures have never been built and are only starting down the path towards completion. Here, have a look.

Getting the speaker to this point has taken nearly a year from when Arnie Nudell and I started dreaming. We first imagined a line source like the IRSV they will someday replace. Then we imagined what kind of drivers would be in that line source and decided on Air Motion Transformers for the mids and tweets, servo woofers for everything else. As one group of engineers are working on designing those drivers to Arnie's specs another is imagining the housing for them, and still another the amplification, servo chain, DSP bass alignment. Will it all fit? Can we maintain the baffle width we need for proper dispersion? How will it be braced? Where will we place the heat sinks and once there will they provide the performance we need? And will the entire package be something people might want to put in their homes? Can mortal humans even move these beasts?

After all these considerations then, and only then, do we build a physical prototype to see how well we did and what changes will have to take place for them to measure the way we want. And then (and only then) do we go to the listening phase where everything we imagined is challenged and vetted. Does it do what we hoped?

Indeed, it's a strange and arduous process but once finished, there's hopefully a grand reward at the end when music comes forth in unhindered effortless beauty.

If not, we roll our sleeves up and start over.

Subscribe to Paul's Posts

28 comments on “Paper tigers”

  1. RIP Siegfried Linkwitz ...

    "One thing Linkwitz said stuck with me. I'd asked him about measurements and the often considerable difference between what's seen in measurements and what's actually heard in listening. 'What is important to the eye is not necessarily important to the ear,' he said. 'Why should it be? Nature makes sure each does its job and does its job perfectly. You get cues from the eye, but some things that look gross in the frequency response, the ear says, "I don't care".'

    "That from a guy who's more a scientist than an artist. Never forget those words when you look at measurements, in Stereophile or elsewhere. Measurements are useful tools, but don't let them hold you hostage."

    Michael Fremer's tribute to Siegfried

    1. Linkwitz's comment is philosophically linked to Helmholtz and Mendel. As much as I loved Linkwitz's work and talking to him, this is flawed thinking. Ears are not perfect machines, and DNA contains no blueprints for cognition.

      Rather, DNA contains rules for growing, wiring and programming neurons in response to quality and quantity of stimuli. We are only capable of hearing those coherent acoustic parameters we heard repeatedly during childhood. We can learn variations on the basic building blocks, but we can't learn constructions outside the Lego set, like screws, adhesives, welding and soldering.

      A better analogy is language. By age 13, your range of phonemes is determined and foreign languages will be adjusted to fit into your native phonemes. By 15 your grammatical construction is largely set. After 21, you are not capable of hearing and speaking like a native for example separating conversations in a bar.

      Further, acoustically trained human hearing is ten times better than a perfect machine! Professional musicians can hear time and frequency an order of magnitude beyond the "mathematically proven" limits of sonic analysis.

      I found myself banging my head against this trying to use spectral editing software. Izotope RX offers a choice of many FFT windowing algorithms, but all of them were "fuzzier" than my perception; and when I got rid of offending noises, playing the file gave me a headache and scrambled my hearing just like the YouTube compression algorithm.

  2. Your design approach, Paul, seems to be based on your longtime experience and on the know how about deficiencies with existing solutions/components; deficiencies which might be overcome with today’s technologies available. I think that is a most common approach which doesn’t address basic deficiencies of stereo reproduction. The latter has been addressed by Ralph Glasgal or Edgar Choueiri (Ambiophonics and BACCH). The arrangement of your line source drivers will result in the fact that the signal of the upper and lower driver will arrive at the listener’s ear much later than that of the driver in the middle. You accept this irritation of the ear-brain system because you follow the theoretical (!) goal of a cylinder wave being emitted. But which listening distances are required for getting a near perfect wave form?
    I would prefer a design based on a point source concept and a phase coherent match of all drivers resulting in the best step response possible. With conventional speakers (tweeter above midrange driver above woofers) I always hear a singer’s voice dropping in perceived height when the melody drops in frequency. Another strange experience is that the sound is much cleaner when listening seated right from the right speaker or left from the left speaker compared to the smeared sound in the sweet spot. Sometimes I think that there are so many irregularities in stereo due to non- identical speakers (too big driver and cabinet tolerances ?).

  3. Hello Paul
    Either they sound good or bad. You can send me a set and we'll measure them through or invite and we will come to the United States and listen to your new speaker.

    More love from Denmark

  4. I have been working in aerospace for four decades, and we still begin preliminary design of most of our aircraft on 2D small and large marker boards.
    The only significant changes to the process have been the ability to make an electronic copy of our scribbling at the touch of a button, and the the absence of the some times intoxicating smell of the markers.

    As with all PS products, visually Paul ‘s new speakers have clean and simple lines where form follows function. I like!

  5. I love building loudspeakers... it's the tweaking that's difficult. I play them for the nineteenth time, then go over to any musical instrument -doesn't matter what it is- play three notes, then blankly stare at the loudspeaker like an old man waiting for the nurse to come by with the pills, drooling on my shirt.

  6. Hello Paul,
    We all agree that the new speakers should be fine-tuned for sound first and foremost.
    Here's another consideration if you can: Weight and moveability. Not all of us have multiple employees around to help shift speakers 1" this way and that over a few weeks or months for optimization, as you do at PS. My back is not what it used to be! In fact, this is why I just ordered a P12 and M700s. I can lift them!

    Also, a base with straight lines rather than artful curves would help in adjusting and aligning the 2 speakers. When I was young (like you still are) I did not think about it, but my back is not what it used to be. Here's an idea: slightly smaller straight-line base with an artful curved base cover that slides easily on when in position, and off when it is time to adjust position again. Also, is there some kind of material you could put on the bottom of the stands that would allow us to adjust position by sliding on wood floors, rather than having to pick them up each time?

    What about some kind of grab bar on the back? Also, what about a fully recessed binding post "well" area on back, so cable connections do not stick out, beckoning children's attention? Perhaps some kind of small plastic panel cover could slide in place to cover them, open at the bottom for the cables. Remember that most of us will put them in our living rooms, away from the wall.
    Okay, I've annoyed you enough with my user requests. Thank you for all your team does. - Jeffrey Saunders

  7. Will these loudspeakers have an active or passive crossover? Why are you making that choice?

    The comment about their size and weight is significant. They look difficult to move, let alone scooch around for fine tuning placement. Will they be easy topple? A removable protective grill cover over the bass drivers is something I'd recommend including because they are at the perfect height for small hands to leave their imprint.

    Will this design end up being priced mostly for billionaire oligarchs and super-villians? Or will there be a model affordable for mere mortals? The super-villain market is very specific in that they only play operatic arias on it while imbibing expensive scotch whiskey while telling us about their plans for world domination.

    1. Over in the Forums, the numbers that are being used for the 3’s and 2”s are somewhere under $10k (shooting for $6k), and $20k, respectively. A guess for the 1’s would be $50k, but maybe Paul has a better estimate at this point.

    2. These speakers have passive built in crossover for the mids and tweets, with active crossover and built-in amplifiers for the midbass drivers (on the front) and the side mounted woofers.

      This particular model, the AN-1, will be heavy. Real heavy. We guess around 500 pounds. If you look closely in the picture you'll note a silver protrusion on the base. This is the wheel lock so these speakers can be moved around with lifting.

      Bass drivers will have perforated metal grilles to cover them.

      The AN-1s are definitely going to be priced only for super villains and oligarchs. We don't expect to sell more than one pair per month. Their packing crates will likely cost us more to build that our entry level speakers just to give you an idea of these beasts. Definitely not for everyone.

      We will have two lower priced models, with the lowest being $10K for the pair.

  8. JAS has a point I also immediately thought of, the weight "issue".
    Over the years I heard many wonderful speakers, a lot of which were way too heavy to handle on your own.
    In audiophile country, to get the "right" sound, a lot of experimenting (positioning) is needed until you're finally satisfied.
    And I don't wanna call one of my neighbors every time I wanna move my speakers. They still like me and I wanna keep it that way.
    So, for me no speaker of 120 lbs. But maybe the small (does PSA know this word when it comes to speakers ?) ones....
    BTW., I never understand why people are so nervous about little children in combination with delicate audio stuff.
    I had little kids once (now grown-ups), and first thing I did is teach them (first sign language and later words) not to touch anything of daddy's audio toys.
    When they grew up, as a teenager they had their own "stereo" in their bedrooms.
    Yes I know, for some it should have been their own "mono", but my kids never complained.
    At that age it was the room they never left anyway, except for food (mostly chips), listening to their music and learning (??).
    So, maybe a bit strange these days, but my kids just did what I told them, or stayed out of the way.
    No worries for me. And they still love me, I think.

  9. The first thing I do when I need to engineer something is study the problem and learn as much about it as I can. I try to define the problem as completely and specifically as I can. I try to measure it based on my understanding of it. If I can't get the right tools to understand it then I have to invent the tools first. I can't solve what I can't understand. I look at how other people solved the problem, how did they see it, what approach did they use, what can I learn from them studying what they did that worked and what they did that didn't work. I look at the problem from both ends. I look at the final results I want and I look at what I have to work with. I use a systems approach, not focusing on individual elements until later. I study the forest first, later I look at the trees. If you don't understand the forest but only focus on one or a few trees you don't know enough.

    In this problem I start at the end studying sound fields I want to create. At the beginning are the recordings I can buy. I try to find the best path in the maze that connects the two. That's the plan to duplicate the forest using trees and any other elements I need which may have been overlooked. In this problem there are two aspects that cannot be overlooked. One are the variables inherent in the recordings and the other is the crucial role listening room acoustics plays. These must be incorporated in the solution if it is going to work. Once the understanding is gained, the details of each tree may not be quite as important to get the desired results. Focusing only on the trees is IMO doomed to fail. Only blind luck using that approach has any chance of success. Luck is not what engineering is about. It's about applying scientific principles to contrive a desired outcome.

  10. We went to Paris a few years back for the opening of the Louis Vuitton building by the great American architect, Frank Gehry. He is a real genius. We've seen quite a few of his buildings, even stayed in the only hotel he designed (it's mad). There was a big retrospective exhibition about Gehry at the Pompidou Centre and his creative process. It was truly fascinating, including a film that Sidney Pollack made about him a while back. It's here.
    I bought the book about the design of the Louis Vuitton building. You can see some images here, he starts with mad crazy sketches of the overall aesthetic and form, then progresses to drawings, mock-ups and eventually construction.
    The exhibition had a lot of these early sketches going back to the 1980s - rough and crude - but they follow through to the final building.
    One of Gehry's sayings, mentioned in the film, is that there is nothing new, it is only technology that is new. The image above is a box designed to propagate sound waves. Those have been around a long time. One day someone will come across a new way of doing that particular trick and we will be amazed. Apple used to do that trick, with the iPod in particular. It's one of the largest companies in the world and it's be years since they came up with something new. And as many people are aware, there is nothing much new about Apple's products because they are very similar to designs by Dieter Rams that preceded them by 30 or 40 years.

  11. Siegfried worked at Hewlett Packard starting in the 60s along with Brian Elliott, both music enthusiast’s possessing creative minds who applied their engineering skills to audio and became speaker designers whose system’s were full-range active dipole designs.

    Few of us have access to an anechoic chamber and Brian being a measurement freak, after listening to a music presentation was always curious to measure the ‘listening environment’ to learn of any additions or subtractions the given room contributed to the overall sonic presentation.

    Both inspired individuals who placed passion over commerce, they took their hobby seriously and applied their individual genius to advancing the art of music reproduction in the home.

    A fun read and a window into the mind and soul of Fritz ...

  12. Paul, I am excited about your creation, and I know you will not take everyone's input, but I would like to just offer a constructive
    viewpoint and you can do with it what you will. I'm sure your loudspeaker will go through many cabinet refinements, but one thing I would certainly like to see go are the two dust catching cut outs on the side. I'm sure if they go, you will be back to the low frequency drawing board, but I have always resisted speakers with powered subs, or passive radiators on the sides of beautiful cabinets. The black grills on the sides of some current speaker designs is bad enough, but having cut outs that are so aesthetically distracting, I'm just asking, is there another way to do away with any cut outs on the sides of your speakers?

  13. Since I started in 1998, I have built 28 speaker cabinet designs. Of those, two were scrapped and four prototypes were modified, so 79% of my initial paper design prototypes met the design goals and 93% of the concepts worked with at most a few table saw cuts.

    This is even more remarkable because 18 of these models were novel designs based on an arcane model of hearing with far more stringent specifications than the standard model of frequency response and IALD. I was designing for not merely timbral accuracy (on-axis and free field room response), but also temporal, transient, dynamic and spatial accuracy, and lowering Doppler inter-modulation distortion.

    AFAIK, Bruno Putzeys is the only other speaker designer working to this many parameters. The PS Audio team is only one short, and that omission makes sense from a marketing perspective. My speakers make 99+% of recordings sound worse by interfering with the fake stereo illusion of pan pots. They have a glaring hole in the middle with mixed multi-mike recordings; but they are more accurate with physically spatial recordings (unprocessed near coincident pair).

    Likewise, the Kii III is a great studio monitor because every knob in the studio is clearly audible. This informs the trade-offs of every decision to twiddle, but made the demo tracks torturous for ears that only hear acoustic music and unprocessed audio chains.

    I suspect that the design of the Kii speakers were also the result of detailed simulations to reduce prototyping expense.

Leave a Reply

© 2022 PS Audio, Inc.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram