[ TNT | Who we are | HiFi Shows | Factory tours | HiFi Topics | Listening tests | Tweakings | Inter.Views ]

Inter.view with Matthew Polk, co-founder of Polk Audio

[Italian version]

TNT-Audio >
Probably there's no need to briefly introduce you and your Company to the audiophile community, since Polk Audio is one of the best known loudspeaker brands in the Planet but, for sure, some newbie this side of the pond may need to know how your biz started back in 1972.

Polk >
I met my future partner George Klopfer the first day of classes at The Johns Hopkins University in Baltimore. I was a Physics major and an audio hobbyist, sort of a nerd actually. I was the kind of kid that would drive local audio store proprietors nuts because I spent a lot of time talking about and listening to audio gear - without buying anything! George majored in History.
He spoke 4 languages fluently and had gone to school in Rome. A very sophisticated guy. For some reason we became great friends probably because he was sort of an audio nut also.

We both graduated from Hopkins in 1971 and I continued at Hopkins as graduate student in Marine Biology (Don't ask why. It seems silly now.) George moved to New Jersey where he proceeded to get fired from a succession of jobs. George is a brilliant guy but he's fundamentally unemployable.
You could say he was born to be his own boss. Meanwhile I found out that I hated Marine Biology. So, when George came back to Baltimore while between jobs I was receptive to his suggestion that we start our own business. About that time friends were asking my advice on putting together a sound system for a Bluegrass Convention they were sponsoring.
George, who is a terrific cabinet maker, offered to build the system if I would design it. I agreed and, voila!, we were in business. Unfortunately (or perhaps fortunately) our friends couldn't afford to pay us for the system.
So, George and I became part owners of the system and put the Polk Audio name on it. (At that time George designed the original logo that we still use today) Since we didn't have any money for new amps or mixers we used lots of old monaural Macintosh and Marantz tube amplifiers and monaural preamps which were available very cheaply on the used market in those days.
People who heard the system were very impressed. And, in fact, it sounded damn good. Almost immediately we picked up some customers who wanted us to build similar systems for them. I just wish I still had some of those tube amps.

For lots of reasons pro-audio wasn't the right business for us. Plus, my real interest was high-end home audio. The early 70's was an interesting time for our industry. Especially in the UK there were some great sounding, very sophisticated speakers being produced like the Spendors and original Quad ESL's, most of which were out of my price range.
These systems had great mid and vocal performance but were also pretty difficult for us Americans to live with.
They wouldn't handle any power, had crappy bass and blew up if you looked at them sideways. In the US we had a whole bunch of basically terrible sounding loudspeakers from manufacturers whose names I won't mention. But, true to what the US market wanted, they went loud and produced gobs of bass.
I was impatient to have it all. I wanted the sophistication, great midrange and soundstaging of the European speakers with the bass, dynamics and sheer fun of the American speakers. I also wanted to do it for a price that most people could afford. So, right from the start, our mission was to put high-end sound into the homes of regular people. Although I personally enjoy the high-end as much as any foaming-at-the-mouth audiophile, the real challenge is to make good sounding stuff that real people can afford and can live with. That's still our mission today.

Now, in our efforts to achieve this goal we've had to come up with new ideas from time to time. For example, we were one of the first to use small, quick drive units to drive a much larger passive radiator. The idea was to take advantage of the speed and detail of the small driver combined with the ability of the larger passive radiator to produce deep bass with good efficiency.
Our first really successful model, the Monitor 7 (introduced in 1974), used a 6.5 inch drive unit coupled to a 10 inch passive radiator with a 1" inch soft dome tweeter. Even today it's tough to beat the Monitor 7 for musicality and involvement in a speaker that can still go loud and produce genuine deep bass. Of course, modern materials and technology have made it possible to go much further.

TNT-Audio >
If I understand well, almost any piece of a Polk Audio speaker is designed and manufactured inside your plants, including drivers and cabinets.
Despite of this expensive procedure your loudspeakers are known to have an extremely high quality/price ratio: What's the secret?

Polk >
In the early years we were able to offer a lot of performance at relatively low prices for a couple of good reasons. First, we kept our overheads very low. George, myself and a few of our friends all lived together in a huge beat up old house in Baltimore that also served as our factory, offices, labs, etc. Of course we didn't have much choice.
We had very little money and it was the only place we could afford. Second, our accounting in those days wasn't very good and we didn't really know what it cost us to make the products. We were lucky we didn't go out of business.
Lastly, and most important, we were very clever in using simple system designs built around high quality components and making sure that we could use the same components in many different products. In fact, until 1990 we used nothing but 6.5 inch drivers in all of our home products.
We also tried to manufacture as many of our own components as possible ourselves. As you can imagine we became very efficient at making very high quality 6.5 inch drivers.

Obviously we don't work out of our house anymore, not that my wife would put up with that, and our accounting is a lot better, but we still have the same philosophy of simple product designs that rely on high quality components.
We also manufacture many of those components ourselves much more efficiently than we could buy them. We build most of our own drivers and tweeters and recently began manufacturing our own cabinets.

Now, just because you make something yourself doesn't automatically mean that it's either better or cheaper. Our decisions about what to make ourselves and what to buy from outside vendors are made very carefully. We try to focus on the things that will give us the greatest "design leverage" both in performance and in cost.
Drivers and tweeters are the most difficult items to maintain consistently high performance in production. In addition, the specific characteristics of the drivers and tweeters, more than any other components, determine the direction of our system designs.
It would be very difficult, perhaps impossible, to get anyone to make them just the way we want them. That's why we make most of our drivers and tweeters. We may spend a little more on these components but the added performance of the components allows us to keep the total cost of the system lower.

Another example is our cabinet production. A typical cabinet supplier must have enough flexibility in their manufacturing to accommodate requests from many different customers. This raises their costs and means that there are many types of cabinet designs that they won't be able to produce cost efficiently and perhaps not at all.
However, we have set up our cabinet production to produce the types of cabinets we need very efficiently and with very high quality. This is how we are able to supply wood veneered cabinets at the same price that other companies charge for vinyl.
In addition, we have the luxury of designing production processes that allow us to make things in our own facility that no one else can produce at all, such as the RM-7500 satellites.

Of course, we buy many things that would make no sense for us to manufacture ourselves. It's hard to see where we would add value by making our own screws or stamping our own steel parts. We focus on things where we can add value.

TNT-Audio >
You have developed several techniques to improve the performance of your products, such as the "Dynamic Balance" etc. Which are, according to you and your vast experience in the field, the weakest spots of a loudspeakers project?

Polk >
Loudspeaker systems really are more than the sum of their parts. So, it's hard to point to any one thing as being the biggest problem. However, Modal resonance or "cone breakup" in drivers is still the greatest limitation in driver performance.
And, as we know, the drivers (or tweeters) are probably the most important component in the speaker system. Eliminating cone resonance isn't possible but they can be reduced or moved to a range where their impact on performance is small.
The trick for doing this in a practical way is to match all the different material characteristics in the drive unit or tweeter so that the mechanical energy supplied by the voice-coil flows smoothly through the mechanical parts of the driver and is smoothly converted to acoustic energy without being delayed by resonance.
This "Dynamic Balance" of materials is a difficult concept which took us years of research at Johns Hopkins in their laser lab to work out. When a material, like a loudspeaker cone, resonates it stores the mechanical energy being fed into it and releases it over a longer time as sound. A bell is a good example and looks a little like a loudspeaker cone.
You hit it once and it rings for a long time. But, using the technique of dynamic balance you can practically eliminate this tendency to ring by carefully matching the material of the cone to all of the surrounding parts. If this is done correctly, when the driver cone is "hit" by energy from the voice-coil it travels through the cone once and into the surrounding structure where it is absorbed.
As a result the response of the cone starts and stop quickly over a broad range of frequencies giving it the ability to reproduce much greater musical detail and at higher output levels without complaint. This is particularly noticeable with well recorded, powerful female vocals.

Nevertheless, the driver and tweeter are only one part of the loudspeaker system and the tough part of speaker design is its iterative and interactive nature.
By that I mean if you make one little change in one part of the design other factors have to change too. I think that's one thing that shocks do-it-yourself speaker builders or people who like to modify existing designs.
Common sense says that if you improve one component in the speaker system, the result is automatically better. If you use a "better" driver the system should sound better.
If it were that easy, I'd be out of a job. The fact is it takes enormous patience, a lot of experience, and the ability to believe what your ears are telling you to create an all-around great sounding system.

TNT-Audio >
Recently we have been asked to judge several DIY speakers project during the National DIY speakers contest organized by one of the largest drivers manufacturer here in Italy (Ciare drivers). Besides subjective sound quality of each design (which influenced the final score for less than 60%) we were asked to judge some electrical measurement (ambience frequency response and impedance) and the quality of the construction.
I'd like to ask how do you see the relation between measures and listening tests. In other words, could you please tell us the role electrical measures and listening test play in a Polk Audio design?

Polk >
Great sound is the bottom line, but as a designer you'll never get there without good measuring equipment and knowing what to do with it. I've seen plenty of average sounding systems that had great frequency response measurements.
But, I've never seen a great sounding system with major problems in the measured frequency response. Good measured performance seems to be a prerequisite for sounding good but there is much that is difficult to measure or that we don't fully understand that makes a speaker sound good. Once we have a prototype that seems pretty good I like to work with a book of measurements made on slightly different configurations of the prototype.
Then I sit down and listen. I try to guess what type of change in the design would push the sound more towards my goal.
As I make each change I take a look at the set of measurements we made for that configuration and begin to build a picture of the measurable changes that will take the design where I want it to go. I make notes on the frequency response charts about how that configuration sounded as I proceed towards the final design.
Of course, it's still a subjective process but this very disciplined approach creates a documented design path for each system. It also gives us a design data base that shows us how to solve similar problems on future projects.

My one piece of advice to anyone designing a loudspeaker, whether a professional or hobbyist, is to have a sonic goal in mind when you start. Use your measurement tools to work towards that goal and don't fall in love with what your equipment is telling you.

TNT-Audio >
You manufacture several satellites + sub combos. Could you briefly summarize the pros and cons of such an approach?

Polk >
As with anything there are pluses and minuses. There are some things that these small, coffee-cup sized satellites do better than most full-size speakers.
For example, the small baffle minimizes diffraction. The better satellites can "disappear" while delivering an amazingly precise sonic image. They also use smaller drive units so that cone resonances are pushed to frequencies so high that they don't matter. Nevertheless, the laws of physics cannot be violated and those small drivers don't provide the lower midrange weight to properly anchor male vocals or the upper bass punch and texture that really makes some music come alive.
Also, let's not forget that blending from the satellites to the subwoofer can be hit or miss even for a well designed system if the customer puts the woofer in a bad place. It is because of the subwoofer blending problems and the "hole" in the lower mid/upper bass that many audiophiles don't like sub-sat systems.

But, let's keep in mind that sub/sat systems weren't designed for audiophiles. They were designed for real people. Placement flexibility is the reason most customers are attracted to these sub-sat systems. Very few customers have either the right room or the willingness to place a full-size speaker system in just the right place for best performance. I'm sure you've seen plenty of examples of badly setup systems.
We did an experiment once where we had some of our people pretend they were looking for a new house. Whenever a real estate agent took them on a tour of a house we had them take pictures of how the stereo or home theatre equipment was setup. At first we all had a good laugh. People had stereo speakers in different rooms, even left and right speakers sitting on top of each other facing into a corner. Then, to our horror, we realized that these were real customers! Our customers!
And this is how they were using our products. Regardless of the quality of their full-size speakers, almost anyone of these customers would have been better off with a good sub/sat system. Why? Because they would've been able to easily place the small satellites where they would sound pretty good as compared to the horrible sound they were getting with their poorly placed full size speakers.

Frankly, I know quite a few audiophiles who wish they could get the performance they want from something as convenient as these small sub/sat packages. Now, I'm happy to say that there is real hope that they can have what they want. The sub/sat systems we are making now are far, far better than those we were making just a few years ago. Research into small driver technology has yielded small driver designs with much greater output capability in the lower midrange and upper bass.
We have also made tremendous progress in the design of powered subwoofers and crossovers which smoothly and invisibly blend with the improved satellites. But the most promising solution we have found is to design the sub/sat speaker system as part of a fully integrated all-in-one system such as our RM Digital Solution-1 system.
In a system like this all of the electronics, amplification and surround processing is designed and optimized specifically for the characteristics of the speakers.
The idea of active speakers with optimized amplifiers is not new but it is only recently that we have recognized the tremendous performance benefits you get by applying this technique to a sub/sat system and particularly to a multi-channel surround sub/sat system.
For example, in the RMDS-1 because it is a fully integrated system we are able to optimize bass management to complement the performance of the specific speakers in the system for all signal types, from plain old stereo to 5.1 Digital.
In addition, by using optimized active electronics we can greatly extend the dynamic range of those little satellites. So, while we may not have completely eliminated all of the performance complaints audiophiles have about sub/sat systems with the RMDS-1 we are getting very close. That could be why I've been getting calls lately from some of my audiophile friends asking about the RMDS-1.

TNT-Audio >
Your Company is getting more and more involved in Home Theater products. Besides the obvious (shielded drivers etc.) which are, if any, the main differences between an audio-only and a home-theater oriented speaker design?

Polk >
The requirements for a good home theatre system are the same as for music, just more demanding. Other than using dipoles for the rear channels (which you wouldn't even have in a stereo system) the biggest difference is bass...
People just seem to like more bass with movies than music and movie producers make sure they get what they want. This mean that speakers used for movies have to be more robust and go louder in the bass than speakers dedicated to music but they still have to do everything else right too. In the future, with DVD audio, we'll need all of this capability for both movies and music.

Also, the fact is that most people have only one high quality system. It's our opinion that music is more important to customers in the long run but we recognize that the desire to have a home theatre may be the reason they are buying the system.
We think it's just good common sense to make the extra effort to make sure the system is good enough to handle both.

TNT-Audio >
It seems, correct me if I'm wrong, that Polk Audio products are getting much more exposure (and tests) on Home-Theater oriented mags. Are Pure-Audio mags losing interest in your products or....is it Polk Audio that has lost interest in HiFi mags? :-)

Polk >
Well in the US at least there are more magazines, editorial pages and web pages aimed at home theater than stereo so the odds are in favor of seeing more Polk home theater system reviews than stereo system reviews.
We certainly have no policy to avoid Hi-Fi magazines or web sites. In fact I think the pendulum has swung a bit too far away from music in our industry and I hope it swings back. I believe we are about to submit a stereo pair of speakers to TNT-Audio for review.

TNT-Audio >
Could you please tell us about new exciting products you are about to release?

Polk >
We just released more new models at one time than in any time in our history. This spring we released 10 new front/stereo models (the RTi Series products), three new center speakers, three new subwoofers, 2 new sub/sat systems and 1 new surround speaker. Whew! I think we deserve a rest after that, don't you?

We are also building a new factory this year so we'll be taking it easy on the manufacturing people by holding down the number of new models this year. However, in a few days we'll be shipping the new DS-2 system- a follow-up product to the RMDS-1 that continues to do very well for us.
The DS-2 is the first all-in-one integrated home theater system to use full-size speakers. The same benefits of optimizing the signal path that we talked about earlier apply equally well to full-size speakers. The sound of some of the DS-2 configurations is really spectacular.

We are building these Digital Solution products because we believe that there are many "convenience" oriented customers who would like to have an exciting, high performance system that they can live with.
We have found that, for whatever reason, many of these people are finding the process of choosing a system very difficult. I mean, we are in this industry and sometimes it's confusing to us.
The average customer has no chance of really understanding what they are buying let alone of setting up correctly. But, we feel that these people deserve the chance to have a high performance system and we intend to do our best to give it to them.

We also believe that there are many "performance" oriented buyers who would like a more convenient HT system but don't like the performance sacrifices they would have to make. Let's face it, manufacturers of most "all-in-one" systems seem to have decided that if a customer wants convenience they couldn't care less about quality or performance. We strongly disagree.
With the RMDS-1, and now the new DS-2, we are serving the needs of quality-minded customers who want a high quality system but haven't found anything that balances their desire high-performance with simplicity of choice and operation.

Long term, we think that speaker/electronics integrated packages will make more and more sense for the sake of both performance and ease-of-use. So that's something we will continue to work on. We're also looking at ways of making it easier to add sound to more rooms of the home and generally move toward our original goal of putting high-quality sound into the homes of regular people.

Courtesy by Matthew Polk for TNT.

Copyright © 2000 TNT-Audio - http://www.tnt-audio.com

[ TNT | Who we are | HiFi Shows | Factory tours | HiFi Topics | Listening tests | Tweakings | Inter.Views ]