Yeah it'd be cool to compare to some other speakers, like the C1 or something (though if you did exactly those @heeman would probably have a coronary worrying about it).
Yup!!
Flint Acoustics: Rocketman Loudspeaker (one off build)
- Thread starter Flint
- Start date
Yup!!
I finally accurately recreated the cumulative spectral decay measurements with the SVS MBS01, which is a wonderful speaker I have truly loved listening to for years.
Under the same conditions as I tested the new Rocketman speakers, I got this chart:
As you can see, there is a considerably greater level of ringing reaching out past 4.5mS. This is mostly due to the quality of the drivers and the use of a passive crossover. Being fair, the 6 inch mid-woofer in the SVS can be obtained for about a third of the cost of the midrange in my Rocketman speakers. The tweeters are likewise in different leagues with the tweeter in the Rocketman selling for about four times more. Based on the drivers alone one shouldn't expect both speakers to behave equally.
Add to this the shape of the of the MBS01 only moderately trying to control diffraction (but they did make some effort contributing to their better than average sound), and the crossovers storing and releasing electrical energy and exaggerating ringing and decoupling the drivers from the amp, and you get what appears to be a significantly worse looking chart.
When I go on and on about passive crossovers being a huge weakness in common loudspeaker designs, this measurement is a major and obvious reason for my ranting. When I talk about a speaker's ability to sound accurate, low distortion, and providing very high resolution, this is one of the main measurable characteristics I am talking about.
So, go with active speakers, use good drivers, design excellent baffles and enclosures, and you'll get stunning results.
Under the same conditions as I tested the new Rocketman speakers, I got this chart:
As you can see, there is a considerably greater level of ringing reaching out past 4.5mS. This is mostly due to the quality of the drivers and the use of a passive crossover. Being fair, the 6 inch mid-woofer in the SVS can be obtained for about a third of the cost of the midrange in my Rocketman speakers. The tweeters are likewise in different leagues with the tweeter in the Rocketman selling for about four times more. Based on the drivers alone one shouldn't expect both speakers to behave equally.
Add to this the shape of the of the MBS01 only moderately trying to control diffraction (but they did make some effort contributing to their better than average sound), and the crossovers storing and releasing electrical energy and exaggerating ringing and decoupling the drivers from the amp, and you get what appears to be a significantly worse looking chart.
When I go on and on about passive crossovers being a huge weakness in common loudspeaker designs, this measurement is a major and obvious reason for my ranting. When I talk about a speaker's ability to sound accurate, low distortion, and providing very high resolution, this is one of the main measurable characteristics I am talking about.
So, go with active speakers, use good drivers, design excellent baffles and enclosures, and you'll get stunning results.
I've gotten some requests from people who have read my older speaker reviews to post a dispersion chart like I used to post rather than try to explain an entirely new chart style. This would make it easier to compare the speakers I've measured and posted about. So... here are those charts.
First I will show the measured frequency response with the microphone 1 meter away at tweeter height from directly on-axis (0 degrees) to directly to the inside of the enclosure (-90 degrees) in 15 degree increments:
Chart: Frequency response taken at 15 degree increments inside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
And for outside the cabinet:
Chart: Frequency response taken at 15 degree increments outside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
The response curves stay pretty even right through the crossover range and up to about 8kHz where they start dropping off more rapidly as the frequency gets higher and the mic is moved further off axis.
The same data as above is more understandable if I normalize all the curve to the zero degree (on-axis) measurement. So here are those charts:
Chart: Normalized frequency response taken at 15 degree increments inside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
Chart: Normalized frequency response taken at 15 degree increments outside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
First I will show the measured frequency response with the microphone 1 meter away at tweeter height from directly on-axis (0 degrees) to directly to the inside of the enclosure (-90 degrees) in 15 degree increments:
Chart: Frequency response taken at 15 degree increments inside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
And for outside the cabinet:
Chart: Frequency response taken at 15 degree increments outside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
The response curves stay pretty even right through the crossover range and up to about 8kHz where they start dropping off more rapidly as the frequency gets higher and the mic is moved further off axis.
The same data as above is more understandable if I normalize all the curve to the zero degree (on-axis) measurement. So here are those charts:
Chart: Normalized frequency response taken at 15 degree increments inside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
Chart: Normalized frequency response taken at 15 degree increments outside of 0 degrees on axis. Black = 0 degrees, blue = 90 degrees
Now, I have already been asked about why I bothered to measure the off axis response both inside and outside of zero degrees (on-axis)?
Since the tweeter and midrange drivers are positioned vertically off centered, the edge diffraction and reinforcement of the baffle will be different from the inside to the outside. As such, some amount of various will be present in the inside dispersion pattern versus the outside pattern.
To exemplify what I mean, I took the response curves from off axis by 30 degree inside and outside and normalized one to the other to clearly show the difference between the two:
Chart: Normalized frequency response taken at 30 degrees inside and outside, black = outside, red = inside
As you can see, there is about a 0.5dB variation between the two up to about 8kHz where it gets hinkier by as much as 1dB and completely different above 15kHz.
To further show the difference between inside and outside, I made the same chart but using the data from 60 degree off axis, inside and outside:
Chart: Normalized frequency response taken at 60 degrees inside and outside, black = outside, red = inside
Since the angle to the baffle is greater, the difference in distance to the edges is more exaggerated, so the affects of the baffle on dispersion are greater. In this case the difference is more than 1dB throughout most of the range of the measurement where they go back and forth on which his louder at certain frequencies.
This difference between inside and outside off axis performance is both a good thing and a bad thing. It is good in that a negative impact on the sound in the room caused by the left and the right propagation is reduce because the potential peak or null at a given frequency isn't repeated identically on both sides, making it 6dB more impactful. It is bad in that you want the reflective sound in the room to be generally balanced to ensure good room response and balanced early reflections in an untreated room.
In the case of these speakers, I took many steps to reduce edge diffraction, so the negative effects are greatly reduced, then what small effects are still there are slightly misaligned to reduce any one defect to be exaggerated from dual identical effects occurring.
If you study the two normalized chart from the post above, you'll notice in the 2nd chart for outside off-axis curves there is a little peak at 8,900kHz in the 15, 30, and 45 degree curves which isn't present in the inside curves in the chart above it. This is a true study in the nerdy shit I go to in order to get the most from the drivers I am using AND to fill the room in the best way I can for average users in average homes.
Since the tweeter and midrange drivers are positioned vertically off centered, the edge diffraction and reinforcement of the baffle will be different from the inside to the outside. As such, some amount of various will be present in the inside dispersion pattern versus the outside pattern.
To exemplify what I mean, I took the response curves from off axis by 30 degree inside and outside and normalized one to the other to clearly show the difference between the two:
Chart: Normalized frequency response taken at 30 degrees inside and outside, black = outside, red = inside
As you can see, there is about a 0.5dB variation between the two up to about 8kHz where it gets hinkier by as much as 1dB and completely different above 15kHz.
To further show the difference between inside and outside, I made the same chart but using the data from 60 degree off axis, inside and outside:
Chart: Normalized frequency response taken at 60 degrees inside and outside, black = outside, red = inside
Since the angle to the baffle is greater, the difference in distance to the edges is more exaggerated, so the affects of the baffle on dispersion are greater. In this case the difference is more than 1dB throughout most of the range of the measurement where they go back and forth on which his louder at certain frequencies.
This difference between inside and outside off axis performance is both a good thing and a bad thing. It is good in that a negative impact on the sound in the room caused by the left and the right propagation is reduce because the potential peak or null at a given frequency isn't repeated identically on both sides, making it 6dB more impactful. It is bad in that you want the reflective sound in the room to be generally balanced to ensure good room response and balanced early reflections in an untreated room.
In the case of these speakers, I took many steps to reduce edge diffraction, so the negative effects are greatly reduced, then what small effects are still there are slightly misaligned to reduce any one defect to be exaggerated from dual identical effects occurring.
If you study the two normalized chart from the post above, you'll notice in the 2nd chart for outside off-axis curves there is a little peak at 8,900kHz in the 15, 30, and 45 degree curves which isn't present in the inside curves in the chart above it. This is a true study in the nerdy shit I go to in order to get the most from the drivers I am using AND to fill the room in the best way I can for average users in average homes.
Oh.... I also wanted to point out, in case you missed it in the middle of the paragraphs above, all the measurements are from 905HZ to 20,000Hz because the longest I can adjust the gate for the measurements to prevent room reflections from having an impact on the measurements limited the capture range to a low end of just over 900Hz. If I were to take these outdoors and put them in a more anechoic space, I could go much lower frequency and get meaningful data.
However, the dispersion characteristics really matter most in the 1kHz to 7kHz range the most as that's the range human brains use most instinctively to place sounds in relation to our ears. I proverbial "twig snaps in a forest" is really the sweet spot for our ability to pinpoint the source location of a sound. It is based on the absolute distance between our ears and the amount of sound blocking our face does between our ears. People wearing glasses have a lesser ability to pinpoint locations based on sound alone. This is why really serious critical listening purists will often take off their eyeglasses and hats (if they are wearing hats) in order to hear the speakers and music the best way possible.
However, the dispersion characteristics really matter most in the 1kHz to 7kHz range the most as that's the range human brains use most instinctively to place sounds in relation to our ears. I proverbial "twig snaps in a forest" is really the sweet spot for our ability to pinpoint the source location of a sound. It is based on the absolute distance between our ears and the amount of sound blocking our face does between our ears. People wearing glasses have a lesser ability to pinpoint locations based on sound alone. This is why really serious critical listening purists will often take off their eyeglasses and hats (if they are wearing hats) in order to hear the speakers and music the best way possible.
Here's a comparison of the polar plot of the Rocketman speaker versus the new Paradigm Premier 800F tower speakers.
To be clear, the software, mic, room and test conditions used to create these curves are not identical, so it isn't a perfect Granny Smith Apple to Granny Smith Apple comparison. It is more like a Delicious Red Apple to Honeycrisp Apple comparison. One obvious difference is the scale in my chart goes from bright red to full black where the other plot goes from bright red to dark blue, making my plot more extreme in its results.
Note that the Paradigm curve isn't as smooth shows the impacts of edge diffraction with the evenly spaced ridges of yellow and orange across the test range. Also, the major null at 13,000Hz appears to be a speaker baffle issue and not a driver issue given the shape of the dispersion pattern. The Rocketman has a null looking dip at 6,000Hz, but in the normalized version of the polar plot that dip in response isn't reinforced by the same anomaly and is more likely just a by product of the tweeter's natural response dip. So, the comb-filter series of dips in the response of the paradigms treble output is very likely to be caused by the baffle shape, size, and the tweeter's lack of absolute directionality control.
Fullrange Driver Polar Plot
I also found some other speakers' polar plots, and I thought I'd share an extreme example.
Here's the Rocketman polar plot next to the polar plot for the MarkAudio SOTA Cesti T tower speaker which has a 2" fullrange driver and two 4" woofers:
Again, remember the conditions or software are not identical, so this isn't an absolutely perfect comparison.
Note that the full-range driver has the traditional issues you'd expect from any large driver in operating at higher frequencies (though this is an amazingly good driver). However, the cabinet baffle is perfectly square edged and the drivers are mounted dead center. You can see the very clearly solid comb-filtering at narrow gaps throughout the range of the measurements. This is NOT good. The frequency response of the driver is fraught with that comb-filtering across the critical midrange. All of the diffraction is complete noise added to the sound which causes the response aberrations. Instead of the direct sound hitting your ears as it is produced by the driver, there are tiny out of phase echoes arriving within milliseconds which is flat out time-domain based added noise. Not good.
CAVEAT
These diffraction issues are not proof of a speaker sounding terrible. Many great sounding speakers have terrible issues with diffraction, and in some cases the signature sound of a speaker comes from the diffractions. However, when it comes to absolute clarity and resolution - the things headphone lovers often talk about as being the advantage of cans over speakers - all the added signals from diffraction is one of the main causes of the loss of clarity and resolution. Of course, the room plays a role as well, but as long as the room reverb and echo is arrives at your ear at least 15mS after the initial direct sound arrives, the perception is that the room sounds are not from the source. In other words, our brains can tell an echo is an echo as long as it is about 15mS after the initial sound. Everything which arrives before that, or before about 10mS, our brains assume are part of the original sound.
To be clear, the software, mic, room and test conditions used to create these curves are not identical, so it isn't a perfect Granny Smith Apple to Granny Smith Apple comparison. It is more like a Delicious Red Apple to Honeycrisp Apple comparison. One obvious difference is the scale in my chart goes from bright red to full black where the other plot goes from bright red to dark blue, making my plot more extreme in its results.
Note that the Paradigm curve isn't as smooth shows the impacts of edge diffraction with the evenly spaced ridges of yellow and orange across the test range. Also, the major null at 13,000Hz appears to be a speaker baffle issue and not a driver issue given the shape of the dispersion pattern. The Rocketman has a null looking dip at 6,000Hz, but in the normalized version of the polar plot that dip in response isn't reinforced by the same anomaly and is more likely just a by product of the tweeter's natural response dip. So, the comb-filter series of dips in the response of the paradigms treble output is very likely to be caused by the baffle shape, size, and the tweeter's lack of absolute directionality control.
Fullrange Driver Polar Plot
I also found some other speakers' polar plots, and I thought I'd share an extreme example.
Here's the Rocketman polar plot next to the polar plot for the MarkAudio SOTA Cesti T tower speaker which has a 2" fullrange driver and two 4" woofers:
Again, remember the conditions or software are not identical, so this isn't an absolutely perfect comparison.
Note that the full-range driver has the traditional issues you'd expect from any large driver in operating at higher frequencies (though this is an amazingly good driver). However, the cabinet baffle is perfectly square edged and the drivers are mounted dead center. You can see the very clearly solid comb-filtering at narrow gaps throughout the range of the measurements. This is NOT good. The frequency response of the driver is fraught with that comb-filtering across the critical midrange. All of the diffraction is complete noise added to the sound which causes the response aberrations. Instead of the direct sound hitting your ears as it is produced by the driver, there are tiny out of phase echoes arriving within milliseconds which is flat out time-domain based added noise. Not good.
CAVEAT
These diffraction issues are not proof of a speaker sounding terrible. Many great sounding speakers have terrible issues with diffraction, and in some cases the signature sound of a speaker comes from the diffractions. However, when it comes to absolute clarity and resolution - the things headphone lovers often talk about as being the advantage of cans over speakers - all the added signals from diffraction is one of the main causes of the loss of clarity and resolution. Of course, the room plays a role as well, but as long as the room reverb and echo is arrives at your ear at least 15mS after the initial direct sound arrives, the perception is that the room sounds are not from the source. In other words, our brains can tell an echo is an echo as long as it is about 15mS after the initial sound. Everything which arrives before that, or before about 10mS, our brains assume are part of the original sound.
TitaniumTroy
Well-Known Member
Very interesting Flint, think you could find polar plots for the M2 or JBL 4367? I was really impressed with the imaging of Revel's new ProformaBe speakers.
Well, I am done tweaking, measuring and confirming my design to be about as perfect as I can get from these drivers.
I hope to sell these soon. There is some interest, but no commitments. If anyone here knows of a audio nut who might buy these, let them know. I will provide my contact info if you don't already have it.
I'll be posting more measurements such as THD, Impulse, and other things as time goes buy and I process the data.
I hope to sell these soon. There is some interest, but no commitments. If anyone here knows of a audio nut who might buy these, let them know. I will provide my contact info if you don't already have it.
I'll be posting more measurements such as THD, Impulse, and other things as time goes buy and I process the data.
In addition to measuring obvious stuff like frequency response and polar patterns, I also captured THD at 94dB SPL (1.5M) and Impulse response.
I like to measure THD at 94dB SPL from half the distance to the listening position, which is typically 3M, which translates to at SPL of 88dB at the listening position. I could measure from the listening position, but reflections can interfere with the results, so this is a compromise.
For the Rocketman speakers, the THD under these conditions is shown in this chart:
The THD is very low down to 30Hz, but there is an uptick in the bass/midrange area centered around 280Hz where it gets as high as 3%. This was consistent in all the measurements which suggests the weakest performance spot with these speakers is in that range. That said, the rest of the usable output range has harmonic distortion of less than 1% most of the time. This is VERY good for any speaker.
I will plot some other speakers to give you some comparative references to better respect what this means in terms of performance.
I like to measure THD at 94dB SPL from half the distance to the listening position, which is typically 3M, which translates to at SPL of 88dB at the listening position. I could measure from the listening position, but reflections can interfere with the results, so this is a compromise.
For the Rocketman speakers, the THD under these conditions is shown in this chart:
The THD is very low down to 30Hz, but there is an uptick in the bass/midrange area centered around 280Hz where it gets as high as 3%. This was consistent in all the measurements which suggests the weakest performance spot with these speakers is in that range. That said, the rest of the usable output range has harmonic distortion of less than 1% most of the time. This is VERY good for any speaker.
I will plot some other speakers to give you some comparative references to better respect what this means in terms of performance.
Last edited:
I also measured the impulse response from 2M (ideally I'd measure at the listening position, but I discovered reflections made that data pretty useless).
This is a very similar measurement as Stereophile's Step Response chart used in their speaker reviews. In this case, the start of the waveform from each speaker is arriving almost exactly at the same time. My calculations and other measurements suggest that at the listening position the arrival time is almost ideal, so this is a very good result.
You can see the first major reflection arriving at about 2.6 mS, so the data to about 2mS is very clean.
I am very pleased with this because I designed the cabinet to align the drivers so a passive crossover would work fairly well. In this test there was no digital delay on any of the individual drivers. Happy!
This is a very similar measurement as Stereophile's Step Response chart used in their speaker reviews. In this case, the start of the waveform from each speaker is arriving almost exactly at the same time. My calculations and other measurements suggest that at the listening position the arrival time is almost ideal, so this is a very good result.
You can see the first major reflection arriving at about 2.6 mS, so the data to about 2mS is very clean.
I am very pleased with this because I designed the cabinet to align the drivers so a passive crossover would work fairly well. In this test there was no digital delay on any of the individual drivers. Happy!
I made some THD measurements of two other sets of very different speakers in order to get an idea of what we are dealing with.
When SPLs adjusted for 88dB at the listening position 14 feet from the speaker, the huge Line Array speakers I built which have sixteen 4.25" midwoofers per channel have astonishingly low distortion in the woofer range, but as the transition is made to the tweeters the distortion goes up considerably. Part of this is the nature of having multiple drivers reproducing the exact same signal and experiencing phase interactions, but part of it is the tweeters have higher distortion at these listening levels.
However, the tiny 2.5" fullrange speakers I built from recovered speakers out of a table radio, the THD was much higher, even when turn down the target SPL at the target listening position to accommodate the expected listening levels for such a small speaker (the graphic has a typo, that should be 84dB SPL @ 1/2M). Yes, when the test tone sweep was running, you could distinctly hear the distortion in the lowest bass range clear as day, as you'd expect.
When SPLs adjusted for 88dB at the listening position 14 feet from the speaker, the huge Line Array speakers I built which have sixteen 4.25" midwoofers per channel have astonishingly low distortion in the woofer range, but as the transition is made to the tweeters the distortion goes up considerably. Part of this is the nature of having multiple drivers reproducing the exact same signal and experiencing phase interactions, but part of it is the tweeters have higher distortion at these listening levels.
However, the tiny 2.5" fullrange speakers I built from recovered speakers out of a table radio, the THD was much higher, even when turn down the target SPL at the target listening position to accommodate the expected listening levels for such a small speaker (the graphic has a typo, that should be 84dB SPL @ 1/2M). Yes, when the test tone sweep was running, you could distinctly hear the distortion in the lowest bass range clear as day, as you'd expect.
And just to brag about my top end Statement speakers, here's their distortion measurement using the same scale, but measured from the listening position with a reference loudness of 88dB SPL.
I couldn't fit all the data into the same window the difference between the root signal being tested versus the harmonic was so darn low. Basically, with a listening position level of 88dB SPL, the THD is well under 0.1% from 80Hz to the upper limit of the test gear, and under 0.07% for most of the range above 400Hz. This just reinforced my choice of drivers and design for them.
Here's the same data but with the scale adjust so that each Y-Axis division is 10dB rather than the 5dB divisions used in all the previous charts:
I have never in my lifetime measured or even heard of any loudspeaker with lower distortion than these. The closest I've seen were the original Quad ESLs from the 1960s which in the midrange had distortions levels below 1% at 85dB SPL, but they struggled to play anywhere close to as loud as we demand from our speakers today.
I couldn't fit all the data into the same window the difference between the root signal being tested versus the harmonic was so darn low. Basically, with a listening position level of 88dB SPL, the THD is well under 0.1% from 80Hz to the upper limit of the test gear, and under 0.07% for most of the range above 400Hz. This just reinforced my choice of drivers and design for them.
Here's the same data but with the scale adjust so that each Y-Axis division is 10dB rather than the 5dB divisions used in all the previous charts:
I have never in my lifetime measured or even heard of any loudspeaker with lower distortion than these. The closest I've seen were the original Quad ESLs from the 1960s which in the midrange had distortions levels below 1% at 85dB SPL, but they struggled to play anywhere close to as loud as we demand from our speakers today.
After some advice and quick review of federal guidelines, here are the final marketing specifications for the Rocketman Loudspeaker System:
Specifications
* Options for passive crossovers, speaker grills, and in-home setup and tuning available
Specifications
- Tweeter = 28 mm hand-coated Silk-Dome
- Midrange = 12 cm polymer composite cone with oversized 54 mm voice coil
- Woofer = 28 cm high power sandwich W-Cone
- Impedance = 8 Ohms compatible
- Effective Maximum Output = 113dB SPL 1M per speaker (w/ appropriate amplification)
- Frequency Response = 29 - 20,000Hz
- Bass Limit = 21Hz (-10dB)
- Crossover* = Hand-tuned electronic active 3-way crossover
- Minimum Amplifiers = 6 channels of Amplification with >50W per channel into 8 Ohms
- Recommended Amplifiers =
- Woofers 2 channels @ >250W / 8 Ohms,
- Midranges & Tweeters 4 channels @ >125W / 8 Ohms
- Finish = Red Stained Natural Oak
- Overall Dimensions = 40 x 16.5 x 19 inches (HxWxD)
* Options for passive crossovers, speaker grills, and in-home setup and tuning available
I had a guest over today who builds speakers himself, mostly horn loaded models paired with low power S.E.T. tube amps. He was eager to hear the Rocketman speakers as he had followed my posts on FB about them and was intrigued. He had a working knowledge of speaker design and was glad to see someone following well-established principles for good sounding design rather than building solely for aesthetics and forcing highly compromised acoustical design to sound decent.
He brought his own music, and to my delight it turned out to be almost entirely real acoustic instruments and singing rather than over produced synth-heavy aural cotton candy that most "audiophiles" use to demo speakers. His selection was good, I asked for the source of all the tracks as I may pick up some of them.
After about an hour with the Rocketman speakers, he stopped and had lots of nice things to say, and made clear what he saw as limitations. He was right on the money, too. Yes, the dynamics of the Rocketman speakers is not as impressive as large horn speakers, but inversely they are delicate, detailed, and unquestionably natural and realistic sounding overall. He was especially impressed with the clarity in the bass range, the stereo imaging, and the resolution.
He did not know about he big Statement speakers in my main listening room, which he later fell in love with after listening for even longer. But when I offered up the selling price for the Rocketman speakers, he was shocked. He was sure I was asking at least $10,000 for them. Maybe he can help me find a buyer!
Anyway, we hit it off and may form a formal area audio group in the coming months.
He brought his own music, and to my delight it turned out to be almost entirely real acoustic instruments and singing rather than over produced synth-heavy aural cotton candy that most "audiophiles" use to demo speakers. His selection was good, I asked for the source of all the tracks as I may pick up some of them.
After about an hour with the Rocketman speakers, he stopped and had lots of nice things to say, and made clear what he saw as limitations. He was right on the money, too. Yes, the dynamics of the Rocketman speakers is not as impressive as large horn speakers, but inversely they are delicate, detailed, and unquestionably natural and realistic sounding overall. He was especially impressed with the clarity in the bass range, the stereo imaging, and the resolution.
He did not know about he big Statement speakers in my main listening room, which he later fell in love with after listening for even longer. But when I offered up the selling price for the Rocketman speakers, he was shocked. He was sure I was asking at least $10,000 for them. Maybe he can help me find a buyer!
Anyway, we hit it off and may form a formal area audio group in the coming months.
For the rest of the story on these speakers, check out the thread started by the new owner of these speakers: