Dynaudio Focus 600 XD Loudspeaker Review

In the first of a series of loudspeaker reviews, the Dynaudio Focus 600 XD full range tower is first up. This “all in one” full range, DSP optimized active 3-way floorstander, is a techno tour de force of computer aided design optimizations. In my review of the Vancouver Audio show, there were a handful of advanced loudspeaker designs that use Digital Signal Processing to optimize the performance of the offering. The Dynaudio’s were one of those speakers that I was curious to listen to and measure. In this review, I provide subjective listening impressions and objective measurement results.

Loudspeakers reviews are difficult to write as it comes down to personal preference, aside that everyone’s listening environments and equipment are different. However, this approach is designed so that my listening experiences and objective measurements results, as compared to industry guidelines, should translate almost 1:1 to virtually anyone’s listening environment. I show an example of this from half way around the world.

One goal is to characterize each loudspeaker against a set of industry guidelines, and build up a library of subjective listening impressions and objective measurements. The industry guidelines referenced in this article show a high degree of correlation to listening preferences and objective measurements, spanning 40 years of research and development. Covering not only the loudspeakers, but also the listening environment.

For the acoustic listening environment, as long as one is within the specs as described in the industry guidelines, then the review characterization and measurements should translate to one’s environment with a high degree of confidence. My stereo setup is offset to the left of center, along the long wall, firing across the short width of the room. In addition, not the most favorable room ratio and my room is quite lively. Still, with a little ingenuity, I am able to meet the industry guidelines for early reflections and reverb time and so can you.

600 XD Technology Overview

The 600 XD’s are outfitted with Dynaudio’s own custom designed and manufactured speaker drivers. A 28 mm silk fabric soft dome tweeter, 14 cm mid/woofer and dual 18 cm woofers in each speaker. The woofers are newly designed for this model with more powerful magnets, lightweight aluminum voice coils and specially developed magnesium silicate polymer cone membrane material.

The 600 XD features Pulse Width Modulation digital amplifier technology with each speaker directly connected to an amplifier module with a total of 600W per channel. There is no requirement for a separate DAC as the digital audio signal is converted to analog right before the amplifier. If one uses the analog input, an analog to digital conversion is performed and passed through the DSP. The internal DSP operates at 32-bit double precision and all internal processing is performed at 64bit/192kHz resolution.

A 3-way linear phase, 24 dB per octave digital crossover is used, with crossover points of 270 Hz and 3.1 kHz. The 600 XD is a time coherent optimized design. Additionally, Dynaudio employs a resonance compensating algorithm for identifying loudspeaker resonances using a minimum of 72 measurements. These measurement allow the speakers directivity to be traced so any system resonances are identified. The algorithm outputs a filter that pre-compensate the resonances.

Further, the 600 XD DSP XO’s are directivity optimized to balance between direct, early reflected and reverberant sound, ensuring a smooth in room spectral balance in an untreated room (like mine). The adjustable speaker position DSP, takes the acoustical effects of boundaries into consideration so that the effects are compensated for and eliminating most of the room’s influence on the bass response of the loudspeakers. Whether the loudspeakers are free standing, against a back wall or in a corner, adjusting the speaker positon DSP accordingly should result in a smooth bass response.

Multiple Connectivity Options

The Focus XD speakers are individually designated left and right on the rear panel and the vertically oriented LED display of the front top corner of each speaker will be visible on the inner side of the loudspeaker. The LED display offers visual feedback as to on/off status, analog or digital connection, as well as wireless Hub or Connection and input source.

Each speaker features digital and analog inputs, as well a digital output. Rotary control for the speaker position mentioned earlier, plus a treble adjustment. There is also a USB service connection that can be used to update the firmware.

According to the documentation, the ideal setup configuration for 24/192 resolution is achieved by connecting the Focus XD’s completely hardwired via the digital inputs and outputs. This requires two 75 ohm coax digital cables. The source digital output would be connected to the digital in of one of the 600 XD’s and Channel Mode set to operate as the Master. The 2nd coax would be connected to the digital out of the master and run to the digital in on the 2nd 600 XD, which is set to Slave. I did this using my Lynx Hilo’s S/PDIF digital out using Blue Jeans Belden 1694A.

Alternatively, the Focus XD can operate with one 75 ohm coax and connection between the speakers is wireless. In this configuration, the Focus XD offers 24/96 signal processing. I tried that as well. There is also the capability to go full wireless digital connection using the optional Dynaudio Hub or Connect. I used the Connect that connects to one’s computer via USB, which acts like a soundcard to the software music player, in my case JRiver Media Center. The Connect will accept analog and digital sources, plus Bluetooth connectivity, a WLAN input, the ability to stream over Wi-Fi and the ability to control the Focus XD via a Dynaudio App. There is also a physical remote control that is handy for switching sources from the listening position. Finally, I tried the analog connection by using the Hilo’s analog output connected to 600 XD analog inputs.

Subjective Listening Impressions - Introduction

Before I begin with listening impressions, I want to establish a vocabulary for doing so from a frequency response perspective, as it will correlate to the objective measures later in the article. With kind permission from Grammy award winning Mastering Engineer, Bob Katz, allow me to present this chart:

What I really like about this chart is the ability to associate subjective descriptions of the sound being heard and correlate to an objective frequency range. For example, if a speaker sounds bright, it is likely to measure a bit higher in amplitude in the 2 kHz to 10 kHz range compared to the rest of the frequency range of the loudspeaker. Conversely, if the speaker sounds a bit dull, it usually corresponds to a reduction in amplitude in the 8 to 16 kHz frequency range. It can be a bit confusing, as the speakers could sound a bit dull due to excessive low frequency energy, which is counterintuitive. Please note this is not an all-encompassing approach, but rather establishing some common vocabulary that we can use to subjectively describe the sound we are hearing at the listening position and relate it to an objective frequency range.

As an ex-recording/mixing engineer, I used a similar chart working with musicians and artists to translate subjective descriptions into eq adjustments on the mixing console at the correlated frequency (e.g. give the kick drum more punch). Note that most studio control rooms have one or more set of monitors calibrated to a target frequency response that to most folk’s ears, is perceived as neutral or accurate at the listening position.

It is no different when reproducing music over a set of loudspeakers in a room. Ideally, if one’s preference is for accuracy, then the sound arriving at the ears should match as close as possible to the audio signal that is encoded on the digital media. We can’t change what was recorded, mixed, or mastered, as it is already encoded on the digital media. However, we can accurately reproduce it, or as close as one can get, given real world constraints of electroacoustic transducers and room ratios/construction.

A note on listening levels. Given that our ears are non-linear when it comes to frequency response, and sound pressure level, it is important to take this into consideration when evaluating a set of loudspeakers. Our ears sensitivity to bass frequencies, relative to other frequencies, decreases as the sound pressure level decreases. One sees on some equipment or software music players, a loudness compensation switch or control to be engaged at low listening levels.

In the professional audio industry, for monitoring, most studio control rooms have a sound pressure level (SLM) meter to monitor at what SPL one is listening to. Most research and industry guidelines suggest our ears frequency response is the flattest around 80 to 85 dB SPL, as measured by a SLM with C weighting and slow averaging on the meter.

Mastering Engineer, Bob Katz has written an excellent article on, “How to make better recordings in the 21st century – an integrated approach to metering monitoring and leveling practices”. Around 83 dB SPL is the level that most audio engineers listen in order to balance the mix, so that when played back at a similar level on one’s speakers at home, the frequency balance, phantom image, etc., is faithful to what was mixed and mastered in the control room, or as close as one can reasonably expect with all of the variables in play.

I spent tens years in a variety of studio control rooms as a recording/mixing engineer and our goal was to always produce the best sound possible. We followed the industry guidelines, as described in the next section, or as best as we could. We had a library of different music genres so regardless of what we were recording, we always had a reference we could compare the mix, at 83 dB SPL, to hear the balance. It was also like having the best stereo system; an acoustically designed room and calibrated sound reproduction system. When we weren’t recording/mixing, we were, you know, listening to the stereo man! This was a long way around to saying that I listened to music on the Dynaudio 600 XD’s on average at 83 dB SPL (i.e. reference level) at the listening position.

600 XD Listening Impressions

As mentioned, in the section on connectivity options, I listened for one week each with an analog input, digital input (coax), and digital input (USB) using the Connect, plus it’s wireless transmission. I was also able to switch listening modes (using the remote) between inputs. At no-time was I able to discern a sound quality change or degradation between all of the different source input options. All transparent to my ears. There never once was a wireless dropout of any sort that I could detect.

I listened to what I feel are good musical performances, recordings, mixes and masters of music I personally enjoy. I also watched movies, DVD and Blu-ray using JRiver Media Center, Netflix, and iTunes. I listened for days, while working from home, to hundreds of tunes from my collection, at many different sample rates and recorded sound quality. In all, about a month listening daily. Here is a picture my listening setup:

This room doubles as a music room for recording, mixing and mastering. With critical listening, I move the coffee table out of the way, throw a comforter over the drum kit, put the acoustic guitar away and muffle the strings on the electric. In this picture I was switching between headphones and the 600 XD’s, getting a sense of tonal balance comparison. The 600 XD’s are placed exactly where my reference speakers were placed, that are off to either side of the 600 XD’s.

Occasionally, I sit and play along with tracks of music I know. It is great to be able to listen to an acoustic guitar recording and play along with my Epiphone DR 500, which has a rich, warm tone to it, and compare playing live to the recording. It is amazing how close the recording can sound like the real acoustic guitar I am playing along with.

The 600 XD’s sound full range to my ears, with a tendency to be just a bit on the bright side, as compared to my calibrated reference (or preference). The sound quality is transparent. I did not hear any colorations at all. One physical give away for colorations is to gently tap on the speaker cone and listen for any box resonances and/or cone colorations – like lightly tapping on a drum skin. Also, tapping on the speaker enclosure itself, can give away any box colorations. I can say the 600 XD enclosures not only look great, but are well damped (i.e. inert) when knocking on the enclosure with my knuckles. Same with tapping on the speakers, well damped and neutral sounding, meaning no perceived coloration to my ears. The 600 XD amplification is silent. I can hear the faintest of hiss if I put my ear right up to the tweeter.

The first feature I immediately grew accustomed to, is when the speakers are in sleep mode and sense a signal present, automatically turn on to the last level they were set at. What a convenience! My previous tri-amped active speaker system had me turning on/off 6 separate amplifiers every time I wanted to listen to music.

I cranked up the level on the 600 XD’s for short periods of time at peak reference level (i.e. 105 dB SPL = really loud!) and did not hear any strain or distortion. One of my favs for peak testing is SRV’s Tin Pan Alley with a DR 18. Transparent keeps coming to mind. I suspect as a result of Dynaudio being able to (computer) control the design and manufacture of each speaker driver and apply sophisticated computer aided DSP algorithms to compensate for any additional cone or enclosure resonances.

The off-axis frequency response of the 600 XD’s is the best I have heard from a cone and dome loudspeaker. Here is what to listen for. Put on music that has a 16th note hi-hat groove and at the listening position, while focusing your attention on the sound level of the hi-hat in the mix, slowly move your head from side to side and listen for any changes to the tone quality or level of the hi-hat. The hi-hat is should remain at the same level in the mix. This can have a significant impact on how large the sweet spot is for any given speaker, especially at high frequencies.

In the case of the 600 XD’s the frequency response was noticeably smooth, no matter where I sat on my 6 foot wide, 3 seat couch. I must say I am impressed as this type of “constant or controlled directivity” is usually associated with waveguides, like those used on the JBL M2 for example. Or other DSP controlled directivity designs like the BeoLab 90 or Kii Three, which I have not heard, but would love to review. I can honestly say that is the first time I have heard that smooth of an off-axis response from a cone and dome speaker system. My reference system uses controlled directivity waveguides, so I have a direct, in-room comparison. For further reading on the importance of a loudspeaker’s directivity or polar response, I refer to Dr. Earl Geddes excellent article on, “Directivity in Loudspeaker Systems.”

I also notice the time alignment or time coherence, which represents all direct sound arriving at one's ears at the listening position at the same time. In a “typical” three system, with a passive XO, the tweeter is physically closer to the ear, than the midrange and followed by the woofer. That is also the order of the sound arrival at one’s ears at the listening position. As we will see, this is easy to measure, but it takes time for one’s ears to tune into what time alignment/coherence sounds like, especially if one has not heard a “real” time coherent speaker before.

My reference speakers are time aligned/coherent and I could immediately tell listening to the 600 XD’s that they had the same time coherence. Once one picks up what is sounds like, it is hard to go back. The main aspect of time coherence, from a listening perspective, translates into excellent transient response, as the bass, midrange and high frequencies are arriving at one's ears at the same time. I find it also greatly enhances the stereo phantom image.

Listening Impressions Summary

Full range, smooth frequency response, both on and off axis, with time coherence, and no audible colorations or distortions while playing at reference level. Plus the ability to integrate the speakers low frequency response into virtually any room. What’s not to like? Let’s see if my subjective listening impressions correlate with some objective, in-room measurements.

Objective Measurements - Introduction

Timbre is a very important consideration for accurate sound reproduction. Consider the Acoustical Society of America’s part definition for timbre: “Timbre depends primarily upon the frequency spectrum, although it also depends upon the sound pressure and the temporal characteristics of the sound.” That is to say that both frequency and timing response are key attributes to accurately reproducing music on a sound system that can reproduce the same timbre (i.e. tone quality) that is encoded on the media.

With that in mind, I will be measuring both the frequency and timing response of the 600 XD and comparing to known industry guidelines for neutral or accurate sound reproduction. I have calibrated my reference system using a custom designed, linear phase Finite Impulse Response (FIR) filter which aligns both the frequency and timing response to these industry guidelines. Let’s start with frequency response.

The following industry guidelines represent over 40 years of subjective listening tests and objective measurements for neutral or accurate sound reproduction at the listening position:

• Relevant loudspeaker tests in studios in Hi-Fi dealer’s demo rooms in the home etc. (PDF)
• Recommendation ITU-R BS.1116-3 (02/2015) Methods for the subjective assessment of small impairments in audio systems. (PDF)
• Listening conditions for the assessment of sound programme material, monophonic and two channel stereophonic. (PDF)
• The Subjective and Objective Evaluation of Room Correction Products.

Taking the recommended target frequency response for neutral or accurate sound reproduction at the listening position, from each of the 4 guidelines, and superimposing them on a frequency response chart:

As can be seen, 40 years of subjective and objective research correlates very well as the responses are all similar. For those wanting to know why a measured flat frequency response is not the preferred target or not heard as perceptually flat at the listening position, these two articles explain, “The Subjective and Objective Evaluation of Room Correction Products” and JJ’s AES presentation on, “Acoustic and Psychoacoustic Issues in Room Correction.”

If you were to check out one guideline, “The Measurement and Calibration of Sound Reproducing Systems”, is a free, open access Audio Engineering Society Paper. In Dr. Toole’s paper, turning to page 17:

My subjective preference is the “trained listeners” curve. I calibrated my sound reproduction system so that it is a similar frequency response as measured at the listening position. That target curve sounds neutral to my ears.

I am not the only one that likes that preferred frequency response at the listening position. Please allow me to share a short story to illustrate a very important point.

Ronald has built a very nice 25 driver full range line array:

Halfway across the world, I have a 3-way horn loaded (old) system:

Here is the frequency response of both Ronald’s system and my system, overlaid on a single chart, as measured at the listening position:

It is no coincidence that the measured frequency response is near identical. It is no coincidence that the frequency response curves match the subjectively preferred listeners curve to not only the Toole reference above, but also with the referenced industry guidelines, spanning over 40 years of R&D. There is a direct correlation between these objective measured frequency responses at the listening position and subjectively preferred target frequency responses for neutral or accurate sound reproduction.

Note that Ronald’s and my speakers are about as different as one can get; full range line array versus 3-way horn loaded. Plus our rooms are different, as are the room treatments, we have next to none. In fact, everything is different about our two sound reproduction systems, including the digital room correction systems we are using. The only common piece is we both use JRiver Media Center and listen to some of the same music. Yet, we are both listening to almost exactly the same frequency response at the listening position for a neutral or accurate response that correlates with industry guidelines.

One point I am illustrating here is that it is possible to correlate, in-room frequency response to accurate or neutral sound as perceived by one’s ears at the listening position. Reading JJ’s paper, based on why we hear what we hear in small room acoustics, instead of octave analysis, frequency dependent windowing (FDW) is used in the best acoustic measurement software and more closely resembles what we actually hear in the room. At low frequencies, both the direct sound from the loudspeaker and the room is analyzed, and as frequency increases, less of the room is analyzed, until at high frequencies, it is mostly the direct sound from the loudspeaker with no room sound at all.

The analysis result using FDW is a frequency response display that is pretty much WISIWIG. If the bass measures flat at the listening position, then it will sound perceptually even at each frequency in the bass region to one’s ears at the listening position. The downward tilt, typically starting at 1 or 2 kHz and reaching -6 dB to -8 dB is perceived as being perceptually flat to one’s ears at the listening position.

The variability is based on how dampened or lively the room is and how much high frequency energy one can take to the ears, relative to how much dynamic range compression is applied to the mastering. Too much can audibly alter the timbre if a heavy hand is used on the threshold, compression ratio, attack and release controls of the compressor/limiter, which the music is passing through during recording, mixing and mastering.

The other point I am making is, it’s not just Ronald and I listening to the same frequency and timing responses, or very close to it, but thousands of DSP’ers out there doing the same, even though everyone’s gear and rooms are different. The most neutral or accurate frequency response, as perceived by one's ears, directly correlates to the measured frequency response range in the Toole diagram above or within the four industry guidelines overlaid in the chart above, with a high degree of correlation and repeatability.

Using neutral response as the reference, we can now compare the speaker under test and describe the differences away from neutral both subjectively, using a common vocabulary, and objective measurements, as compared to industry guidelines. These are the differences that should translate almost 1:1 to virtually any environment, using neutral as the subjective and measured objective reference.

There is also a timing response target for the ideal loudspeaker, which we will get to, but first, now that I have a reference for a neutral or accurate frequency response, let’s have a look at the measured frequency response of the 600 XD’s and compare.

Objective Measurements – 600 XD

Frequency Response

Folks may be surprised to learn that most home based listening environments, unless unusually dampened or lively, will meet industry guidelines for early reflections and reverb time. As presented in JJ’s paper referenced above, it is the direct sound timbre that is the most important to our ears for both tone quality and sound localization (i.e. phantom image). While early reflections may have impact, there is an audible threshold spec in the guidelines that once below, does not affect how we perceive localization. Same goes for room decay or reverb time, there is a range that varies with room size, and most rooms will meet the guidelines. Meeting the guidelines means taking the room out of the review equation. With a few adjustments to my room I am able to meet the spec in the guidelines.

With the 600 XD’s in the same physical location as my reference speakers, here is the measured frequency response at the listening position, overlaid with my preferred target frequency response:

With no external custom filtering, this is a very good response in my live listening room. It correlates well with my subjective listening impressions of extended low frequencies, smooth response and just a wee bit more treble than my reference/preference target. The 600 XD measure down to 22Hz before starting to roll off. You can see that the hinge point for the gradual roll off starts around 1 kHz to 2 kHz, just like in industry guidelines. Just a little rise past 5 kHz relative to the target and response extends past 20 kHz.

Looking at the chart, and the red curve, which is the left speaker, is positioned towards the corner of my room, I should have turned down the speaker position DSP by one notch to reduce the slightly excessive low frequency energy, but it did not bother my ears :-) I am just pointing out the flexibility to be able to adjust the low frequency response of each speaker individually to accommodate scenarios such as mine.

If one take’s the frequency response average, from 100 Hz on up, it is +-3 dB tolerance which falls within the industry guidelines mentioned above. And about +-5 dB below 100 Hz. That is still way better than my reference, with no room correction as I was getting +- 7 dB or more swings just below 100 Hz. The Dynaudio results are pretty impressive for my less than ideal listening room.

What about off-axis response? To my ears, I could sit anywhere on the couch and it sounded completely smooth to my ears, no hi hat disappearing tricks:

I took 2 measurements of both left and right speakers. One 3 feet to the left of center listening position and the other 3 feet right of the center listening position. In other words, across a 6 foot wide sweet spot. As can be seen by the measurements, with the target as reference, sitting anywhere on my 6 foot couch, I am getting excellent tonal similarity and high frequency dispersion. Really well done.

Make no mistake, these are full range, very smooth sounding speakers that have virtually the same tone quality across a 6 foot wide sweet spot. Very impressive on and off axis frequency response.

Time Coherence – Step Response

What about the speakers timing response? First we need to establish a baseline understanding of what an ideal loudspeaker’s timing response should measure like.

In a series of articles on measuring loudspeakers, this is an excellent explanation on why we use a step response display to show a speakers time coherence instead of an impulse response display. The fundamental issue is that looking at an impulse response display of a multi-driver speaker system, it is dominated by the tweeter’s output. Relative to the tweeter, the midrange amplitude is a factor of 10 lower in amplitude and the woofer 100 times lower in amplitude. Therefore, it is next to impossible to see time alignment of the three drivers in a typical 3-way system using an impulse response display. However, the step response display, is more evenly weighted across the frequency range and therefore much easier to see system time alignment/coherence.

The Stereophile article then shows what a good and not so good step response look like. Check out Figure 11 for a visual on a good step response. These measured step responses provide the answer if a speaker is time aligned/coherent or not. Note most speakers are not. Here is an example of a typical 3-way speaker, showing the tweeter arriving first, midrange second, and woofer arriving last. I.e. not time coherent:

I should point out that no amount of eq will solve the problem of the tweeter response arriving at our ears first. We are so used to hearing this when listening to most multi-driver loudspeaker systems that our ears are accustomed to the treble arriving at our ears first, and we really don’t give it a second thought. However, once one hears a proper time aligned/coherent speaker system with the direct sound from each driver arriving at one's ears, all at the same time, it is difficult to go back.

Consider the ideal speaker, 20 Hz to 20 kHz frequency response with +-0.25 dB tolerance, essentially an extension of the amplifier, not altering the music waveform in anyway. Whatever is encoded on disc is reaching one's ears the same way it passes through an amplifier. However, most speakers do not reproduce 0 Hz (i.e. DC). There is always a low frequency roll-off (and high frequency roll off). In the example in the next paragraph, I am using 15 Hz as the beginning of the low frequency roll off.

In this filter design tool, I can design and model the ideal speaker using the specifications above. Here is what the frequency and corresponding step response look like from an ideal speaker with no crossover or room in the picture:

Here is a zoomed in view of an ideal speakers step response:

It is the vertical line in the step response (i.e. at location 0.125) that indicates time alignment. The decay of the waveform is based on the low frequency roll off, depending on if the speaker is a sealed or ported design, will change the shape of the roll-off.

Here is my reference with the overlaid ideal speaker step response at the listening position. The low frequency alignment of this enclosure is a 3rd order Butterworth. Remember, the ideal step response models a perfect loudspeaker, no crossover, and no room:

As one can see, my 3 way speakers are time aligned. I should point out that Ronald’s 25 speaker line array has a very similar time coherent step response, but he is lucky not to have to deal with crossovers. It is amazing how close one can get to the “ideal” speaker step response.

Here is the 600 XD, overlaid with the ideal step, which I believe is a 4th order Linkwitz Riley alignment (sealed enclosure) as measured at the listening position:

Looks very good. Note the vertical step indicating time alignment of drivers. The low frequency alignment is pretty good as well. As mentioned in the measured frequency response section, I could have turned the speaker position eq down one more notch which would lower the step decay more in line with the ideal step as there is just a bit too much low frequency energy. Which did not bother my ears

Finally, as in the frequency response measurement across a 6 foot area above, here are the step response measurements, for both left and right speakers, 3 feet to the left of the center listening position and 3 feet to the right of the center LP:

Aside from the expected amplitude level differences, we see the vertical step time alignment is still intact. Meaning the speaker drivers are perfectly time aligned, across a 6 foot listening sweet spot. Impressive.

One enabling feature that the 600 XD employs is using linear phase FIR filter for digital XO. As Rod Elliot says, “Firstly, it is important to understand that all analogue (and most digital) filters cause phase shift and group delay. The sole exception is the digital linear-phase 'finite impulse response' (FIR) filter.” See Rod’s article, “Phase, Time and Distortion in Loudspeakers.” Also, this article on “Thoughts About Crossovers” demonstrates this graphically.

As Rod concludes in his Phase Time article, “I originally started this article not to praise, but to debunk the theory that time alignment is the only way a speaker should ever be designed. Having done the research, run tests, and written the article, I confess that I must agree with many (perhaps even most) of the points made by the time alignment proponents. My overall opinion, based on the research for this article (primarily tests and simulations), is that time alignment is a very good thing, and perhaps all speakers should be designed this way.” Based on my own research, measurements, and listening tests, I can only agree.

I hope folks understand that the vast majority of speakers are not time aligned, and therefore are distorting the timing of the music arriving at one's ears. Assuming one prefers accuracy. I hear an audible difference, both in transient response and phantom image improvement, with time aligned and phase coherent loudspeakers. I am going to leave the subject of phase coherence for another article, but in multiway speakers, only a digital linear phase FIR filter XO does not distort phase as referenced in the two articles above.

Conclusion

The Dynaudio 600 XD sound smooth to my ears no matter where I sit on my 3 seat couch. The frequency response measurements correlate very well with what I am hearing. The transient response and stereo phantom image is excellent and correlates well with a time coherent loudspeaker. Transparent keeps coming to my mind. At no time did I hear any noise, hum or distortion from the speakers.

The ability to adjust the individual loudspeaker low frequency response, including accounting for boundary effects, is a great feature to be able integrate these loudspeakers into one’s listening environment. I adjusted the speaker position settings by ear first, before I took any measurements. I got pretty close, but could have made one more fine tuning adjustment.

The high frequency response, both on and off axis is smooth to my ears, and again, correlates very well with my in-room measurements. Really quite impressive for a cone and dome loudspeaker. The speakers are just a tad bright compared to my reference/preference, but can easily be adjusted. While there are several ways to achieve this, the Focus XD's DSP foundation and the available USB port on each speaker could facilitate a lot of possibilities. Per Dynaudio:

"Because we focus on this interdisciplinary, we can use the possibilities in digital signal processing to enhance the quality of our sound:

• by manipulating signals to offset irregularities in the speaker
• by improving certain frequencies
• by preventing loss of information in the signal path
• by continuously making better loudspeakers that reproduce sound as close to the original as possible."

- Jan Abildgaard Pederson, Dynaudio

Wouldn't it be great if Dynaudio could issue an individual user firmware upgrade, in which one can download a file, put it on a USB stick, plug it into the back of the speaker and upload the new firmware? Maybe even perfectly tailored to their own room? So, for example, imagine that rather than the 3 position treble switch having +-1 dB range, it could be "updated" via firmware to have a +- 3 dB range. That is if one prefers slightly less high frequency energy arriving at the ears (as do I). It doesn't seem to be outside the realm of possibilities, especially reading Dynaudio CTO Jan Abildgaard Pedersen's comments of DSP and the myriad possibilities: "With DSP, we can do things that are not possible with traditional technology."

The range of connectivity options should accommodate virtually any source and whether one wires the speakers or not. As mentioned before, I did not notice any loss in transparency, no matter what connectivity option I tried.

If one is looking for a small footprint, “all in one” active loudspeaker system that can be easily integrated into one’s listening environment and accurately reproduce music both in the frequency and time domain, the Dynaudio 600 XD does an excellent job.

Postscript

Whew, this was a long article. It took me quite a bit of time as the article includes both the review of the 600 XD’s and the approach for subjective listening impressions and objective measurements. I am basing my objective measurements on industry guidelines as a way to compare the speaker under test to a reference. Of course, it is one's preference in the end that counts. However, accurate would be a good place to start.

As an observation, the industry seems to focus on sample rate conversion algorithms, reclockers, de-blurring algorithms, etc., yet the majority of loudspeakers grossly distort the timing response of the audio signal arriving at one's ears. I look at loudspeakers as the “business end” of sound reproduction. It is, by orders of magnitude, the weakest link in any sound reproduction chain. Compare any speaker measurement to any component, electronic or software in the sound reproduction chain and it is plainly obvious, by orders of magnitude. I hope the industry focuses on loudspeaker technology to bring us more accurate sound.

I have another set of DSP enhanced loudspeakers that I am setting up for review. The next article should be shorter in length as the approach has already been laid out, so the focus will be more on describing the loudspeaker technology, subjective listening impressions and objective measurements results. With the repeatable approach it will be interesting to overlay the results as more loudspeakers are reviewed. And with that, I hope you are enjoying the music!

“Mitchco” Barnett.

I love music and audio. I grew up with music around me, as my mom was a piano player (swing) and my dad was an audiophile (jazz). My hobby is building speakers, amps, preamps, etc., and I still DIY today.

I mixed live sound for a variety of bands, which led to an opportunity to work full-time in a 24-track recording studio. Over 10 years, I recorded, mixed, and sometimes produced over 30 albums, +100 jingles, and several audio for video post productions in several recording studios in Western Canada.

This article covers a high level overview of accurate sound reproduction. For a more detailed perspective, I wrote this 327 page eBook that provides a step by guide to Accurate Sound Reproduction using DSP. Click on Look Inside to review the table of contents and read the first few chapters for free.