Electrostats and Dyamics
I have used the original Quad, Acoustat 2+2, and Martin Logan CLS as midrange reference transducers over the past 27 years. I like their lack of cabinet resonance and single-driver coherence; however, they all exhibit tremendous dynamic compression from a mechanical standpoint. At Cal Tech, we used a laser interferometry setup to measure the dynamic range as referenced to excursion potential vs. voltage. As the electrostatic diaphragms are stretched very tightly and have no elastic edge surround, they have a finite limitation on excursion potential, no matter how much more voltage you input. Initially, the diaphragm will move on a 1:1 scale, but rapidly, the excursion does not follow the input voltage, resulting in a very high degree of compression of the signal. This is the source of "lack" of dynamic range that most listeners pick up on. Most cone systems do not compress until very high volume levels are reached.
Albert Von Schweikert
According to this theory the electrostats will compress the dynamic shifts in the lowest octaves (<100Hz) due to excursion limits....
Why Steve M states the opposite?
(refering to ESL III and generally electrostats being able to reproduce dynamic shifts in an excellent way)
Thanks in advance
Michael
I have used the original Quad, Acoustat 2+2, and Martin Logan CLS as midrange reference transducers over the past 27 years. I like their lack of cabinet resonance and single-driver coherence; however, they all exhibit tremendous dynamic compression from a mechanical standpoint. At Cal Tech, we used a laser interferometry setup to measure the dynamic range as referenced to excursion potential vs. voltage. As the electrostatic diaphragms are stretched very tightly and have no elastic edge surround, they have a finite limitation on excursion potential, no matter how much more voltage you input. Initially, the diaphragm will move on a 1:1 scale, but rapidly, the excursion does not follow the input voltage, resulting in a very high degree of compression of the signal. This is the source of "lack" of dynamic range that most listeners pick up on. Most cone systems do not compress until very high volume levels are reached.
Albert Von Schweikert
According to this theory the electrostats will compress the dynamic shifts in the lowest octaves (<100Hz) due to excursion limits....
Why Steve M states the opposite?
(refering to ESL III and generally electrostats being able to reproduce dynamic shifts in an excellent way)
Thanks in advance
Michael
I´ve listened to ESL63 with Gradient subs and that was really nice, to short to really fall in love though.
Then I´ve listened to ML´s four times, different models, different rooms and they have never managed to put an ipression on me.. I don´t get it.
They sound hard, flat and unnatural to me. Sure, I hear something good, some kind of resolution but kind of synthetic at the same time.
You guys who swear by e-stats, how do you feel about SL´s dipoles?
Those I like big time.
/Peter
Then I´ve listened to ML´s four times, different models, different rooms and they have never managed to put an ipression on me.. I don´t get it.
They sound hard, flat and unnatural to me. Sure, I hear something good, some kind of resolution but kind of synthetic at the same time.
You guys who swear by e-stats, how do you feel about SL´s dipoles?
Those I like big time.
/Peter
ESL speaker sounds more coherence to me due to possible of just one cross over point. In dynamic speaker design, the cross-over filter can be fairly complex (2 or 3 cross over points) and can cause phase variation hence (to me) they don't sound as clean or clear comparing to ESL. In addition, the dynamic speaker also uses different drivers comparing to ESL with just a single membrane so the sound ESL produces will have the advantage.
The other side benefit is that the ESL panel does not have a box which will eliminate another interactive element of the system.
The key disadvantage of ESL speaker is that they don't play very loud with low distortion compare to the best of the dynamic speaker. However, in the same price point (less then 6K USD), I will lean toward ESL speakers over dynamic speakers but of course we should expect to find areas one is better than the others.
The bottom line is that the person pays for the products so it is his/her choice to pick whichever sounds better to purchase. Happy listening.
The other side benefit is that the ESL panel does not have a box which will eliminate another interactive element of the system.
The key disadvantage of ESL speaker is that they don't play very loud with low distortion compare to the best of the dynamic speaker. However, in the same price point (less then 6K USD), I will lean toward ESL speakers over dynamic speakers but of course we should expect to find areas one is better than the others.
The bottom line is that the person pays for the products so it is his/her choice to pick whichever sounds better to purchase. Happy listening.
Tinmanrtx;
"In addition, the dynamic speaker also uses different drivers comparing to ESL with just a single membrane so the sound ESL produces will have the advantage.
The other side benefit is that the ESL panel does not have a box which will eliminate another interactive element of the system."
Using one single membrane is IMO one big drawback with ESL´s, it´s what make them problematic in the high and low frequencies.
Also not only ESL´s have the (possibly) benefit of being boxless, there are boxless dynamic dipoles as well and IMO these compare very "favourably" to ESL´s.
All IMHO of coure.
/Peter
"In addition, the dynamic speaker also uses different drivers comparing to ESL with just a single membrane so the sound ESL produces will have the advantage.
The other side benefit is that the ESL panel does not have a box which will eliminate another interactive element of the system."
Using one single membrane is IMO one big drawback with ESL´s, it´s what make them problematic in the high and low frequencies.
Also not only ESL´s have the (possibly) benefit of being boxless, there are boxless dynamic dipoles as well and IMO these compare very "favourably" to ESL´s.
All IMHO of coure.
/Peter
I agree. The fact that ESL's are restricted in membrane materials is a limitation not an advantage. Look at the huge variety and research into material sciences used for cone diaphrams.
I also believe that mating the ESL panel with a dynamic sub is more difficult, and there are some high end ESL's which use a ribbon tweeter too - so still some potential xover points and mating issues to deal with.
Saying that I've only heard two different ESL's, both ML's, both jaw dropping experiences at their transient capabilities on first audition, but not so inspiring in other areas.
I also believe that mating the ESL panel with a dynamic sub is more difficult, and there are some high end ESL's which use a ribbon tweeter too - so still some potential xover points and mating issues to deal with.
Saying that I've only heard two different ESL's, both ML's, both jaw dropping experiences at their transient capabilities on first audition, but not so inspiring in other areas.
Yeah but no one of the cone materials can be as light as 3g of the whole diaphragm in an ESL panel of 120x50cm 3.8 micron.
Also there aren't any crossovers and there are some ESLs in order to be able to reproduce highs in a good way have separated panels for highs witout x-overs or eq.>... ER-Audio ESL III kit, Quad 989...
Also there aren't any crossovers and there are some ESLs in order to be able to reproduce highs in a good way have separated panels for highs witout x-overs or eq.>... ER-Audio ESL III kit, Quad 989...
Weight of the moving mass of the transducer is nothing that define high performance in itself. Bandwith, decay, dynamic range and distortion is what it is about and I have not seen anything that indicates that ESL´s would be better in any of these areas.
I would love to learn more though if I have missed something.
/Peter
I would love to learn more though if I have missed something.
/Peter
Actually Pan,
Mass of the moving element (driver cone or dome)is probably single most limiting factor of the sound transducers.
Bandwidth is limited on the low freq. with size of the moving
element and on the high frq. with mass of the moving element. That is why we have big bass drivers and small tweeters.
Regarding decay time, e-stats have advantage over
dynamic drivers because of their low mass and large area, air damping is enormous so decay time is minimal.
And we all know decay will produce distortions. This kind of "decay distortions" will be most noticeable with transients.
E-stats have low freq. reproduction limitations. That is why
dynamic is limited if you try to push them to work in this area. Simple solution is use of dynamic driver for bass reproduction.
Crossover point should be 300-500Hz and if so integration between e-stat and woofer is very good.
This way crossover point is far from critical freq. 2k-3k
and also resonant freq. of the e-stat. is no problem any more. Full range e-stat. is not necessary.
Narrow dispersion of the e-stats is good and bad thing. Good for less reflections of the walls and bad if you have more than one listener.
Another thing about e-stats, You have to live with them for
some time to fully appreciate sound. Listening in a audio store will not give you real picture. You need time to "decontaminate" from what ever you are used to listen before.
It is like cutting on spices and too much salt in your food to discover
real taste. Good steak does not need ketchup.
Bottom line is: we are diy-ers and to build dynamic system you need to spent ten times more money to get similar sound quality.
Best regards to all.
Sasha.
Mass of the moving element (driver cone or dome)is probably single most limiting factor of the sound transducers.
Bandwidth is limited on the low freq. with size of the moving
element and on the high frq. with mass of the moving element. That is why we have big bass drivers and small tweeters.
Regarding decay time, e-stats have advantage over
dynamic drivers because of their low mass and large area, air damping is enormous so decay time is minimal.
And we all know decay will produce distortions. This kind of "decay distortions" will be most noticeable with transients.
E-stats have low freq. reproduction limitations. That is why
dynamic is limited if you try to push them to work in this area. Simple solution is use of dynamic driver for bass reproduction.
Crossover point should be 300-500Hz and if so integration between e-stat and woofer is very good.
This way crossover point is far from critical freq. 2k-3k
and also resonant freq. of the e-stat. is no problem any more. Full range e-stat. is not necessary.
Narrow dispersion of the e-stats is good and bad thing. Good for less reflections of the walls and bad if you have more than one listener.
Another thing about e-stats, You have to live with them for
some time to fully appreciate sound. Listening in a audio store will not give you real picture. You need time to "decontaminate" from what ever you are used to listen before.
It is like cutting on spices and too much salt in your food to discover
real taste. Good steak does not need ketchup.
Bottom line is: we are diy-ers and to build dynamic system you need to spent ten times more money to get similar sound quality.
Best regards to all.
Sasha.
"Mass of the moving element (driver cone or dome)is probably single most limiting factor of the sound transducers."
the mass dictates high frequency capability sure, but I don't think Peter said anything to the contrary.
"Good steak does not need ketchup."
But a good curry does need spice. How broad is the bandwidth of your palette.
"Bottom line is: we are diy-ers and to build dynamic system you need to spent ten times more money to get similar sound quality."
I don't understand? it sounds like you're saying that us DIYers are spending ten times more on EDL's to get similar sound quality as ESL's?
the mass dictates high frequency capability sure, but I don't think Peter said anything to the contrary.
"Good steak does not need ketchup."
But a good curry does need spice. How broad is the bandwidth of your palette.
"Bottom line is: we are diy-ers and to build dynamic system you need to spent ten times more money to get similar sound quality."
I don't understand? it sounds like you're saying that us DIYers are spending ten times more on EDL's to get similar sound quality as ESL's?
Hi Sasha!
"Mass of the moving element (driver cone or dome)is probably single most limiting factor of the sound transducers."
I know some about drivers, my point is that even though ESL`s and ribbons may have lower mass than domes that does not translate to any kind of superiority. Lowest distortion is still from dynamic tweeters and they can reach up to 100k, that´s enough in my book. ;-) And BTW, if a ESL membrane weigh 3grams (as someone stated) that is really heavy compared to domes and ribbons.
"Bandwidth is limited on the low freq. with size of the moving
element and on the high frq. with mass of the moving element."
In real life inductance of voice coils and transformers comes into the equation not to mention break up and various membrane/cone/dome resonances. And BW is not limited by size down in frequency. It´s a weight/suspension thing.
"That is why we have big bass drivers and small tweeters."
No, that is more becasue of the amont of air that needs to be moved to approach realistic levels, and about dispersion. Now how was it , has ESL´s small membranes?
"Regarding decay time, e-stats have advantage over
dynamic drivers because of their low mass and large area, air damping is enormous so decay time is minimal."
The waterfall graphs that I have seen has not been pretty of ESL´s. Have you missed that? In theory it may seems like that, but the absence of a suspension and sufficient elasticity/flexibility makes the ESL membrane very resonant. Please point me to a measurement that highlight the superiority to ESL´s.
"And we all know decay will produce distortions. This kind of "decay distortions" will be most noticeable with transients."
And that is the reason I believe many do not like ESL´s.
"Crossover point should be 300-500Hz and if so integration between e-stat and woofer is very good.
This way crossover point is far from critical freq. 2k-3k"
What is it that makes you feel that crossing at 2-3k is bad/critical?
Crossing in the middle of the fundamental range of voice and most acoustic instruments is something that most authoritys suggest that you avoid. This is both due to phase distortion and maybe mostly the sudden change in dispersion (unless same technology is used for both LP and HP drivers).
"Narrow dispersion of the e-stats is good and bad thing."
Yes, it CAN be a + in the lows and mids, however the HF is not something stats excel at.
"Good for less reflections of the walls and bad if you have more than one listener."
Less sidewall reflections, yes. However the backwall is a big problem with dipoles.. I know I have used dipoles for a long time myself.
"Another thing about e-stats, You have to live with them for
some time to fully appreciate sound. Listening in a audio store will not give you real picture."
I recognize accurate (good IMO) sound instantly.
"You need time to "decontaminate" from what ever you are used to listen before.It is like cutting on spices and too much salt in your food to discover
real taste."
Absolutely not. You would need to know what I have been listening to first, and second I use live sound as a reference when evaluating audio components.
"Bottom line is: we are diy-ers and to build dynamic system you need to spent ten times more money to get similar sound quality."
Not anywhere close, and no matter, bang for the buck was never part of this discussion.
Cheers!
/Peter
"Mass of the moving element (driver cone or dome)is probably single most limiting factor of the sound transducers."
I know some about drivers, my point is that even though ESL`s and ribbons may have lower mass than domes that does not translate to any kind of superiority. Lowest distortion is still from dynamic tweeters and they can reach up to 100k, that´s enough in my book. ;-) And BTW, if a ESL membrane weigh 3grams (as someone stated) that is really heavy compared to domes and ribbons.
"Bandwidth is limited on the low freq. with size of the moving
element and on the high frq. with mass of the moving element."
In real life inductance of voice coils and transformers comes into the equation not to mention break up and various membrane/cone/dome resonances. And BW is not limited by size down in frequency. It´s a weight/suspension thing.
"That is why we have big bass drivers and small tweeters."
No, that is more becasue of the amont of air that needs to be moved to approach realistic levels, and about dispersion. Now how was it , has ESL´s small membranes?
"Regarding decay time, e-stats have advantage over
dynamic drivers because of their low mass and large area, air damping is enormous so decay time is minimal."
The waterfall graphs that I have seen has not been pretty of ESL´s. Have you missed that? In theory it may seems like that, but the absence of a suspension and sufficient elasticity/flexibility makes the ESL membrane very resonant. Please point me to a measurement that highlight the superiority to ESL´s.
"And we all know decay will produce distortions. This kind of "decay distortions" will be most noticeable with transients."
And that is the reason I believe many do not like ESL´s.
"Crossover point should be 300-500Hz and if so integration between e-stat and woofer is very good.
This way crossover point is far from critical freq. 2k-3k"
What is it that makes you feel that crossing at 2-3k is bad/critical?
Crossing in the middle of the fundamental range of voice and most acoustic instruments is something that most authoritys suggest that you avoid. This is both due to phase distortion and maybe mostly the sudden change in dispersion (unless same technology is used for both LP and HP drivers).
"Narrow dispersion of the e-stats is good and bad thing."
Yes, it CAN be a + in the lows and mids, however the HF is not something stats excel at.
"Good for less reflections of the walls and bad if you have more than one listener."
Less sidewall reflections, yes. However the backwall is a big problem with dipoles.. I know I have used dipoles for a long time myself.
"Another thing about e-stats, You have to live with them for
some time to fully appreciate sound. Listening in a audio store will not give you real picture."
I recognize accurate (good IMO) sound instantly.
"You need time to "decontaminate" from what ever you are used to listen before.It is like cutting on spices and too much salt in your food to discover
real taste."
Absolutely not. You would need to know what I have been listening to first, and second I use live sound as a reference when evaluating audio components.
"Bottom line is: we are diy-ers and to build dynamic system you need to spent ten times more money to get similar sound quality."
Not anywhere close, and no matter, bang for the buck was never part of this discussion.
Cheers!
/Peter
Okay guy's,
Too many things we disagree on so no point arguing and it is too late for me anyway.
I guess, it comes down to personal taste. I have heard both kinds and there is no contest for me.
If you like what you hear go for it. I always prefer natural sound over slam and ultrasound.
Regards.
P. S.
If Pink Floyd mixed "The dark side of the Moon" on QUAD it is good enough for me.
Too many things we disagree on so no point arguing and it is too late for me anyway.
I guess, it comes down to personal taste. I have heard both kinds and there is no contest for me.
If you like what you hear go for it. I always prefer natural sound over slam and ultrasound.
Regards.
P. S.
If Pink Floyd mixed "The dark side of the Moon" on QUAD it is good enough for me.
Yes I agree with sasha, as it is a matter of taste...
I have heard many Esls that I didn't like and same goes for the dynamic loudspeakers, on the other hand there are some ESLs that really are superb, and some dynamic too (with the last ones costing >6000$)
As I said before 3 g membranes (three) should not be compared with the tweeter only but also with the other mid/bass and bass woofers (Mms of 15-70g) that share the frequency range.]
And in the ESL kit I will be buying soon one of the three panels that is for high frequencies is just 0,16g
(remember that the membranes are about 10-20 thinner than humans hair!)
Thanks by the way for the opinions!
I have heard many Esls that I didn't like and same goes for the dynamic loudspeakers, on the other hand there are some ESLs that really are superb, and some dynamic too (with the last ones costing >6000$)
As I said before 3 g membranes (three) should not be compared with the tweeter only but also with the other mid/bass and bass woofers (Mms of 15-70g) that share the frequency range.]
And in the ESL kit I will be buying soon one of the three panels that is for high frequencies is just 0,16g
(remember that the membranes are about 10-20 thinner than humans hair!)
Thanks by the way for the opinions!
IMO, I think electrostatics are *the best without question* – in the *midrange - *for material that doesn’t need high SPL with moderate sized drivers - especially including eg voice. That’s where the *magic occurs.
In treble, ribbons are about as good.
I hope to employ them *only as the midrange, in the centre channel of a mixed HT/ music system, where L & R are the SL Phoenix, where dynamic compression is not so much of an issue.
I expect much better *midrange resolution from electrostatics than the dipole Phoenix Scan Speak mid-basses.
Many say that the L, C & R should be the same. But the centre channel in HT is mostly voice. For dialogue, I want maximum clarity of *voice from the centre (more so than the L & R, which I have chosen for best general music reproduction). I think to have the centre channel better for voice, is better.
I was considering using a Manger centre mid, which may be about the same quality as an electrostatic, but they are expensive, and you can DIY electrostatics http://sound.westhost.com/project105.htm now with Part 2, so an electrostatic is better value.
Cheers
In treble, ribbons are about as good.
I hope to employ them *only as the midrange, in the centre channel of a mixed HT/ music system, where L & R are the SL Phoenix, where dynamic compression is not so much of an issue.
I expect much better *midrange resolution from electrostatics than the dipole Phoenix Scan Speak mid-basses.
Many say that the L, C & R should be the same. But the centre channel in HT is mostly voice. For dialogue, I want maximum clarity of *voice from the centre (more so than the L & R, which I have chosen for best general music reproduction). I think to have the centre channel better for voice, is better.
I was considering using a Manger centre mid, which may be about the same quality as an electrostatic, but they are expensive, and you can DIY electrostatics http://sound.westhost.com/project105.htm now with Part 2, so an electrostatic is better value.
Cheers
Of course it´s a matter of taste, but that does not change the science.
I believe one big reasons why many like ESL´s is due to the dipole radiation which means higher direct/room sound ratio.
Michaelpage26,
the lowest distortion tweeters I´m aware of is Accuton D20/D30 and D20 weighs 0.09g.
rick57,
"IMO, I think electrostatics are *the best without question* – in the *midrange - *for material that doesn’t need high SPL with moderate sized drivers - especially including eg voice. That’s where the *magic occurs.
In treble, ribbons are about as good."
Ribbons are the drivers with lowest distortion in midrange. Dynamic drivers rules in bass and highs... talking distortion now.
"I expect much better *midrange resolution from electrostatics than the dipole Phoenix Scan Speak mid-basses. "
That´s possible, however with Excel or Accuton as mid drivers in dipoles you might be surprised. In the end it´s much about integration and power response of course.
Please do not get my posts wrong here now. I do not suggest that ESL´s can not be good speakers that suits many listeners and rooms, however the reason why this is the case may be open for argument. The technical explanations from ESL fans simply does not fit the reality.
Cheers!
/Peter
I believe one big reasons why many like ESL´s is due to the dipole radiation which means higher direct/room sound ratio.
Michaelpage26,
the lowest distortion tweeters I´m aware of is Accuton D20/D30 and D20 weighs 0.09g.
rick57,
"IMO, I think electrostatics are *the best without question* – in the *midrange - *for material that doesn’t need high SPL with moderate sized drivers - especially including eg voice. That’s where the *magic occurs.
In treble, ribbons are about as good."
Ribbons are the drivers with lowest distortion in midrange. Dynamic drivers rules in bass and highs... talking distortion now.
"I expect much better *midrange resolution from electrostatics than the dipole Phoenix Scan Speak mid-basses. "
That´s possible, however with Excel or Accuton as mid drivers in dipoles you might be surprised. In the end it´s much about integration and power response of course.
Please do not get my posts wrong here now. I do not suggest that ESL´s can not be good speakers that suits many listeners and rooms, however the reason why this is the case may be open for argument. The technical explanations from ESL fans simply does not fit the reality.
Cheers!
/Peter
I also love an ESL's resolution, but that's not the full story. On another note, I don't remember seeing hardly any ESL's at the hifi expo a couple of months back at Heathrow. Hundreads of thousands of pounds of equipment but without an ESL in sight. The biggest names in the game choose to market their product using dynamic drivers. why? 
Pan:Of course the Mms of the Accuton D20 has a really small vallue of 0.09 g, but this tweeter costs 2,700Euro/Each.
Also how can you compare a dome with a big membrange for the highs that measures 1.5"x48" that is just 0,16g?
In addition those ESL III from Er-Audio have really low sensitiviy (85db/w) and low Max Spl (100db) with big space between the stators, so there's enough Xmax in order dynamics not be compressed and have very good bass and extension (sub needed though)-as Hifi world Review states... They have big panel surface 0.55 m^2 that is about 50-60% bigger than QUAD ESL 63 and most ML designs
And one final thing: as I have said before you can't compare the mass (2.8 g) of the esl membrane that handles 30hz-23000hz with the sum of Mms of the three at least cones in a full-range electrodynamic loudspeaker (that always overcomes the value of 70g)... If you can't accept this then, I don't know how to give you another point of view
Anyway, everybody finds his audio nirvana sometime...