Specific Questions
Room PlacementEquipment Recommendations
What makes a good loudspeaker?Specific Questions

Q4.1 A lot of hype is used to describe loudspeakers and I do not understand any of it. Is any of this stuff useful or is it all advertising techno-babble?

Answer: To a loudspeaker designer who lives and breathes loudspeakers the terms phase, amplitude, nyquist plot, and waterfall are as familiar as sky, tree and car to a normal person. So lets back up and take a fresh look at loudspeaker measurement. In my answer to this question I have restricted myself to the on-axis response, that is, what does the loudspeaker sound like when it is pointing at you.

In the early days of hi-fi, when two loudspeakers which "measured the same" sounded entirely different, the popular notion (even amongst early designers) was that loudspeaker measurements were completely meaningless. What they were measuring was the first and most simple characteristic of how loud it plays, which we call "amplitude". The amplitude is how loud the speaker is at different points on the musical scale. If a pianist plays every note at the same volume, then they should also sound the same volume through a loudspeaker. But that is only the first paragraph of a long story.

It was not until the late Richard Heyser wrote penetrating and lucid AES papers about the time behavior of loudspeakers, that it was even possible; in principle; to measure loudspeakers meaningfully. His great breakthroughs related to the effects of phase upon the sound of a loudspeaker. Basically his insight (in this specific area) was that the apparent position of a loudspeaker had something to do with its phase.

*footnote: Phase is the measure of whether one thing is in step with another. To illustrate this, imagine that the Salvation Army band is marching by in a parade. If you look at their feet, you will notice that the right foot of every man moves forward at the same instant. The feet in this band are in phase with each other. Contrast this with a crowd on a city street; there is no relationship between the feet - they are completely out of phase.

If the phase and the magnitude of a loudspeaker were not locked together in an intricate dance called "minimum phase", then the speaker would appear to be at several places in the room at once, stretched out in a line from the listener, through the speaker, and out to infinity. In order to prove this hypothesis, he designed equipment which could measure phase. The problem was, that measuring phase this way was about as easy as balancing three eggs one on top of each other. Partly to solve this problem, he devised "time delay spectrometry". I won't go into TDS here, but suffice it to say that this was a revolution in loudspeaker measurement. With TDS you could also see the time response of the speaker.

This brings us to the next important loudspeaker measurement which is called the Waterfall plot (or Cumulative Spectral Decay to scientists and engineers). This plot displays how loud the speaker is at all frequencies (called the Magnitude response by the same people) at each instant in time. These plots are shown one after each other, like cards in a deck. The effect is like looking at a cliff face. The steeper the cliff the better.

As time passed, the Maximum Length Sequence system was perfected and offered by DRA Labs in their MLSSA package. This allowed engineers to process their measurements in all sorts of interesting ways. One plot, which shows a huge amount in once glance, is the Nyquist plot. In this plot, the loudness and the phase are displayed like the hands on a clock. The longer the hand, the louder the speaker and the hour the hand is pointing to is the phase. In this system, 3 o'clock is zero degrees, 12 o'clock is 90 degrees, and so forth. As the frequency rises the hand traces out a line on the clock face and this is the plot you get. A ideal loudspeaker would show a straight line going from the centre out to some distance and then squiggling around in a tight little bunch for all frequencies (So far as we know the only conventional loudspeaker which does this is the Duntech Prince).

So in summary if the manufacturers had the nerve all they would need to show you is the waterfall plotm and the Nyquist plot and you could tell virtually everything you need to know about on axis response.

Q4.2 You claim that you build the "worlds most accurate loudspeaker". What is so good about that? (a related question is, "What is your competitive advantage over other speaker manufacturers?")

First, let us note that the loudspeaker is not a musical instrument in itself. The most important function of the loudspeaker is to let you hear a recorded musical performance without headphones. This is deceptively simple, since the concept of hearing a performance has more to it than just having the notes hit your eardrum. What you want, is to feel like you are at some kind of natural venue listening to the actual performer. In order for this illusion to work you have to be able to hear the emotional message the musician is trying to convey to you, and to imagine the space the performance took place in. For you to hear the performer, every nuance of pitch, volume and tempo must be presented to you without modification. So from a qualitative standpoint, a loudspeaker must not add or subtract anything from the performance.

So far so good, but how could we claim that our speaker is the most accurate? Well, in our opinion the subjective concept of 'accurate' stated above, has an exact restatement in technical terms. An accurate loudspeaker must reproduce the form of the electrical waveform fed into it as an acoustical wave of the exact same shape. To actually achieve this in a real loudspeaker is a titanic undertaking. Not only do waves of a certain level have to be reproduced, but very small waveforms and very large ones must be reproduced in exactly the same way.

To give you a point of comparison consider a fairly normal loudspeaker like the millions produced each year world wide. This normal loudspeaker has a wafer thin enclosure, cheap drivers, iron cored inductors and cheap electrolytic capacitors in the crossovers, and all the drivers are mounted on the same surface on the front of the loudspeaker. When you measure them, their response is all over the place because the components have been taken out of a pallet load and stuffed into the box. Quality assurance consists of connecting the finished speaker to an oscillator to see if it makes a noise. So what's wrong with that? Firstly, the cheap drivers sound different at loud levels than they do at quiet levels. They also modify the sound of everything put through them by adding their own quack or honk. Their magnetic circuits and suspensions are non-linear, and produce spurious sounds (called "distortion"). The iron cores in their inductors saturate magnetically on loud sounds, and so the speakers don't go any louder beyond a certain point, whether the music is louder or softer at that point. The cheap capacitors overlay a dark thick pudding texture to the highs, and the misaligned crossover produces either a near complete absence or vast overabundance of sound at certain frequencies. The inattention to detail means that one speaker will behave differently from another. Finally, putting all the drivers on the same wafer of wood means that the sound from each driver will arrive at your ears at a different time , and the front baffle will put out sound on top of this mess as well since it is ringing like a bell.

To build a perfect loudspeaker you need to ensure the drivers are of the highest quality, and are arranged so that the sound reaches your ears at the same time. The crossovers must be absolutely transparent and must ensure a smooth response, on and off axis. The magnitude and phase should be locked together in a minimum phase configuration, and diffraction from the front baffle should be minimised. This is making sure that the desirable sound from the drivers, does not bounce off the face of the cabinet, producing a ghost sound after the main sound (This is like the ghosting you get on a TV when an airplane flies over your house). Finally, the cabinet itself must be inert and must not radiate sound itself (Note that just lowering the vibration will not necessarily fix the problem. The sound radiated from a vibrating surface, can, and usually does have a different spectrum from the vibration energy, since some of the vibrational modes have a different coupling efficiency to the air).

Every one of our speakers and their components are tested from start to finish. Everything we make is accurate within the quoted limits, the difference between models, is their ability to deliver higher SPLs (loudness) and their amount of bass extension.

Q4.3 Are your speakers compatible with home theatre/home cinema?

The goal of cinema sound, is to help you suspend belief and feel that you are actually there with the actors on the screen. This is the same goal to which the high end of audio aspires to. A very good music reproduction speaker will make a very good cinema speaker. You should feel confident in using any of our speakers for home cinema with the following proviso. Most of our loudspeakers are not magnetically shielded so keep them at arms length from the TV. You will be astonished how good movies and even television shows sound through a pair of Duntech speakers.

Q4.4 Is it advisable to install a subwoofer?

Answer: If you want more bass from your system use a subwoofer. Our loudspeakers are designed to provide accurate reproduction of all frequencies including the bass. For cinema soundtracks, the bass level can be very high since by their nature these soundtracks are completely synthetic constructions made up of different snippets of recordings. In addition, throughout your life as a movie goer you have been preconditioned to want a lot of bass when viewing a film. Basically our expectations of movie soundtracks, based upon our experiences in the cinema, are out of sync with the kind of sound actually produced in real life. If you feel like you are missing out on something, then by all means, get a subwoofer.

Q4.5 What are the benefits of a centre channel?

Answer: For non-discrete surround sound like Dolby Pro-Logic there is no benefit to a centre channel, provided your speakers are Duntech speakers. The quality of imaging produced, is so compelling that the centre channel actually degrades the sound, in our opinion. Use the phantom centre channel setting and save your money in this case.

For discrete surround sound like DTS, Musicam or Dolby Digital 5.1 the centre channel provides a place for the discrete centre channel sound to come out. If the film is skillfully mixed then this centre channel information will actually improve the experience. Buying a centre channel is basically a concrete expression in your confidence in the recording engineer's skill and discretion.

Q4.6 What do the power ratings on a speaker mean?

Answer: For the most part the power ratings on a speaker mean that you can write "500Watts" on an in store display to help you sell speakers. This is probaby a rather unsatisfying answer, so I will elucidate.

The ratings originally arose from a genuine desire to let the consumer know if the amplifier he had would blow up the speaker. The problem with this line of reasoning is found in this riddle: "Which amplifier will blow up your speakers first; the 1000 Watt amp or the 10 Watt amp? Answer: the 10 Watt amp because it will clip first.". Unless you are using a tone control, you will most likely blow up your tweeter, because the amplifier can't deliver at the volume levels you want. When you do this, the top of the waveform gets sliced off, as if it was clipped by garden shears. The result of all this, is the production of a huge amount of high frequency information, which is why the tweeter is the one to fry.

Q4.7 How powerful should my amplifier be?

The factors involved are, how sensitive is your speaker and how big is your room. (And for some people, how loud do you like your music.) Most Duntech speakers will produce 90 dB sound pressure level at 1 metre for 28.3 Volts of input and 110 dB for 28.3 Volts of input. To actually get 110 dB continuously, you would need a 200 Watt amplifier (into 4 ohms) with enough reserve to reproduce the peaks in the music correctly. If you were in a factory and were exposed to 110 dB they would be required, by law, to send you home after 1 minute 52 seconds. So a 200 watt amplifier with adequate reserve is all you would ever need in a normal room (This is a little simplified, since the reverberant field level is less than the 1 metre level, but you get the gist). At the other end of the scale a 30 Watt amplifier could give you 101dB, which is not as loud as real music gets, but in a small room might sound like plenty. Strangely enough, even a 10 Watt Single Ended Valve (Tube) amplifier can sound quite satisfying, and since this kind of amplifier is so benign in its clipping characteristics, it won't blow up your tweeters either.

From a practical standpoint, there are not very many excellent 30 Watt amplifiers out there with the reserve capacity to do the job, so go for 70 to 100 Watts unless you have a really big room or like your music dangerously loud.

Q4.8 What is dangerously loud?

Answer: Rub your thumb and index fingers together near your ear while listening to music. If you can't hear the fingerprints rubbing together at any time then it is probably too loud.

Q4.9 What is impedance and why should I care about it?

Answer: Impedance is a measure of how much the speaker impedes the flow of current through it. A normal impedance these days is 4 ohms. A low impedance is 2 ohms and a high impedance is 12 ohms. A low impedance speaker lets a lot of current flow, and if your amplifier can't provide that torrent of electric current it will clip and overheat. Clipping is explained above, in Q4.6. Overheating can be disastrous if there is no thermal cutoff in the amplifier. The reason why huge tank-like amplifiers exist is so they can drive low impedance speakers. Most amplifiers will happily drive a 4 ohm speaker, but watch out for bridging. When an amplifier which happily drives 4 ohms is bridged, it may not be able to provide the current for 4 ohms. In such a case, you should use the second amplifier in a bi-amped setup, and remove the shorting links from the back of your speaker.

Q4.10 I have been looking for a Duntech dealer with no success, what can I do?

Answer: We do not have distributors in every country. In situations where a customer has no distributor in their country, we reserve the right to service these customers directly.

Q4.11 I can spend $500, and get a whole system with CD Player, amplifier, radio, speakers, and a voucher to get a free CD. Why should I spend $2500 on a pair of speakers alone?

Answer: The speakers in your $500 system will make some kind of sound, but it won't be music as we know it. The value in a pair of speakers, is that they turn your favourite CDs and LPs into a musical event. In addition, Duntech loudspeakers are handcrafted to a very high standard.

The design of the speakers from the baseplate to the top is directed at producing the most accurate sound as possible, in a cabinet which is a beautiful piece of furniture in its own right.

Q4.12 Does it matter what kind of recordings I play on my system?

There is a saying in engineering circles which goes like this, "Garbage in, Garbage out". Your system can't make a bad recording good. The best you can hope for is to get everything out of a poor recording which is there. The good news, is that most master tapes which are used to make recordings are excellent compared to what you are used to. This means that a company like Mobile Fidelity, can come along and remaster a recording in a way which captures all the excellent sound on the master. Just listen to the Modern jazz quartet in 1964 on MFSL-XXXX. You wouldn't believe that it is at least forty years old.

Q4.13 The Duntech company has been making speakers since the 1960's and the first Sovereign came out in 1982. Aren't these speakers old technology?

First lets note that the Sovereign of today is entirely different from the first one produced. The measurement techniques let us "see into" the performance of the speaker in a way which was impossible to do back then. The drivers themselves have been improved over and over by the manufacturer. The crossover components are vastly improved over those in the past.

Secondly, our speakers are designed correctly. We are still using the pulse coherent principles because they are still correct. The designs of Duntech loudspeakers are not swayed by trends and fashion.

Q4.14 Some people say Duntech speakers don't produce enough bass, why is that?

If there is no bass in a recording then there should be no bass in the playback. Some speakers seem to produce some bass out of thin air, even while playing a soprano singer. This is because stray bass sounds in the recording are being blown up by the boosted bass.

The other problem is that many speakers rely on a pure ported output. The correct way to do it, is to use a combination of ported and sealed output, which Duntech have found is the only way to create satisfying and accurate bass (which is why the entire Gemstone range and the Prince use this operating principle!)

What happens in a pure ported enclosure, is that the sound is produced by both the bass driver and the port. The sound from the port is delayed in time because the sound is actually the result of a plug of air in the port "bouncing" on the air in the enclosure. This takes some time to happen and the lower the frequency the greater the delay (in technical terms it is called "group delay"). This delay produces a muddying of the bass. For many people, unused to real music in a real venue, this becomes the "normal" sound of bass, and when they don't hear it from a speaker they assume there is not enough bass.

Q4.15 This criticism of ported bass would apply to your speakers too wouldn't it?

Many of our speakers don't have porting at all and use a sealed enclosure. To get very low bass this way you need a really big enclosure. Which is exactly what we do in most of the Classic series.

If you want to make a smaller speaker and still have full range, then you need to produce bass another way. This is usually done with a ported arrangement tuned to a rather high bass frequency like 50Hz or 60Hz. The group delay effect, which produces a thick pudding like bass, is worst at the resonant frequency of the port, so this kind of design is fatally flawed.

In our Gemstone range we have a sealed driver and ported driver working together over the same range. The ported driver (which is very subtly assisted by the port) and the sealed driver work together. The sealed driver is very fast, and supplies bass energy at the same instant as the other transient information in the music. The ported driver fills in the low end, and subjectively the speaker goes very low without any sacrifice in sound quality.