I have a lot of old Dayton-Wright ESL panels, maybe 30 or so. Each is about 6 wide x 16 inches high.
So, I've been thinking about some kind of continuous wall of sound*. But I've been wondering about how to think about channels. Obviously, it would be simplest to have just two channels and two amps (and/or bi-amp'ing as needed for subs and maybe tweeters in each channel).
The simplest way is to just have 16 panels on the left with the left channel and 16 on the right with the right channel. That's conventional.
Then I thought maybe better sound would be 12 on the left, 12 on the right, and the middle 8 would still be driven by the L and R amps, but some reversed in position so as to put a bit of R just left of centre and a bit of L just right of centre.
Next up in complexity would be a centre section formed by mixing L and R and feeding it to a third amp (or bi-amp). I have been using a single low-bass channel ("mixed bass") for the last 50 years.
My interests for the moment are purely in addressing the question making a great way to re-create serious music at home if you had a "blank sheet of paper" and a few dozen panels.
Any thoughts appreciated.
Ben
*yes I know, ESLs don't like walls. But for the moment, I am talking about channels, not room placement or other down-the-road questions like that, except as relevant to the larger issues.
So, I've been thinking about some kind of continuous wall of sound*. But I've been wondering about how to think about channels. Obviously, it would be simplest to have just two channels and two amps (and/or bi-amp'ing as needed for subs and maybe tweeters in each channel).
The simplest way is to just have 16 panels on the left with the left channel and 16 on the right with the right channel. That's conventional.
Then I thought maybe better sound would be 12 on the left, 12 on the right, and the middle 8 would still be driven by the L and R amps, but some reversed in position so as to put a bit of R just left of centre and a bit of L just right of centre.
Next up in complexity would be a centre section formed by mixing L and R and feeding it to a third amp (or bi-amp). I have been using a single low-bass channel ("mixed bass") for the last 50 years.
My interests for the moment are purely in addressing the question making a great way to re-create serious music at home if you had a "blank sheet of paper" and a few dozen panels.
Any thoughts appreciated.
Ben
*yes I know, ESLs don't like walls. But for the moment, I am talking about channels, not room placement or other down-the-road questions like that, except as relevant to the larger issues.
Last edited:
Thanks for your inquiry, but can we focus on channel and sound reproduction issues, please.
(The inert "welder's gas" allows the wide-spaced panels to run at bias voltages up maybe 15kv which results in high efficiency and also can get loud. Also, like a horn, heavy gas makes the one-meter-square films that seal the box (front and rear) into a large driving surface and also, instead of the SaranWrap driving thin air, it drives somewhat thicker welder's gas for a better impedance match. Lets the boxes be sealed from dust. Not everybody's idea of how to make ESLs. These cells work fine (maybe, better) without the gas and the box albeit at lower than out-of-this-world voltages.)
Ben
Last edited:
Phased arrays? (maybe with a wearable position sensor, so the sweet spot could follow you around)
what's your ceiling height?
I'd vote for a tall floor to ceiling line source. Probably toed in to face across the listening position (perpendicular to the face of the cells) for best results... wide-ish spacing.
That's the easiest and simplest method.
If you have extra cells, then set them up with a LP so that you get LF augmentation, maybe from about 250-300Hz down. As you probably know you can do the LF trick in the drive to the cells themselves.
Something like that, variations on this theme to suit the specific circumstance.
I'd vote for a tall floor to ceiling line source. Probably toed in to face across the listening position (perpendicular to the face of the cells) for best results... wide-ish spacing.
That's the easiest and simplest method.
If you have extra cells, then set them up with a LP so that you get LF augmentation, maybe from about 250-300Hz down. As you probably know you can do the LF trick in the drive to the cells themselves.
Something like that, variations on this theme to suit the specific circumstance.
Hi,
picking up the idea of an phased array. 😉
Mr. Walker has suggested a segmented wall-to-wall panel, resp. an array of slim panels connected via inductors, to form a delay line with constant delay per segment. Audio signal would be fed into each end.
This arrangement should result in two tilted-against-each-other wavefronts.
Source: Ron Wagner´s Book. fig 15.34.
jauu
Calvin
picking up the idea of an phased array. 😉
Mr. Walker has suggested a segmented wall-to-wall panel, resp. an array of slim panels connected via inductors, to form a delay line with constant delay per segment. Audio signal would be fed into each end.
This arrangement should result in two tilted-against-each-other wavefronts.
Source: Ron Wagner´s Book. fig 15.34.
jauu
Calvin
Calvin - I thought Walker's Quad ESLs used the segmentation to address beaming and frequency issues - successfully.
For sure, interesting notion to create a wall awash in sound and "corrected" or shaped by phasing the array along its length. Although not sure what kind of inductors could be used to control micro-micro farad ESL panels. In addition, phase is one of those things that seems real important to engineering theoreticians but doesn't seem to matter to human ears unless we are talking big perceptible delays.
Bear - from the point of view of the listener, the source, point, line, or otherwise isn't important, only the stuff that gets to their ears. And that can created in a variety of ways whether cone or film. With ESLs, a narrow source, as you suggest, is a wise move, unless you have a lot of panels to work with. An ESL line-source is also helpful if you imagine listeners can hear comb filtering.
Ben
For sure, interesting notion to create a wall awash in sound and "corrected" or shaped by phasing the array along its length. Although not sure what kind of inductors could be used to control micro-micro farad ESL panels. In addition, phase is one of those things that seems real important to engineering theoreticians but doesn't seem to matter to human ears unless we are talking big perceptible delays.
Bear - from the point of view of the listener, the source, point, line, or otherwise isn't important, only the stuff that gets to their ears. And that can created in a variety of ways whether cone or film. With ESLs, a narrow source, as you suggest, is a wise move, unless you have a lot of panels to work with. An ESL line-source is also helpful if you imagine listeners can hear comb filtering.
Ben
Last edited:
Can you show a pic of one of these panels?
I think you should think about how to create a flexible way to attach and increment the relative angle between panels, so that you can actually try all the different arrangements that we might suggest. (My first imagined layout is a 'conventional' one, but I can also get more crazy...)
I think you should think about how to create a flexible way to attach and increment the relative angle between panels, so that you can actually try all the different arrangements that we might suggest. (My first imagined layout is a 'conventional' one, but I can also get more crazy...)
Walker had many ideas of how to use ESLs to advantage besides the ringed delay network used in Quad's ESL-63. I believe the one Calvin is talking about is illustrated in the attached excerpt from one of his Wireless World articles. Basically an entire wall(floor-to-ceiling) replaced with a horizontally segmented ESL fed from the LHS and RHS and interconnected with delay networks. Ideally the result would be that no matter where you stood or sat in the room you would experience the same stereo image. Of course side wall reflections would complicate matters.
In this day and age, I would think rather than trying to build a HV delay network(requiring high value inductors capable of handling large voltages) one might consider driving each segment from its own amplifier. The input signal for each amplifier could be derived from a pre-amp level analog delay line fed from either and with the left & right stereo signal and with each end terminated with the line's characteristic impedance. Or a few multichannel DSP units would make short work of the incremental delay and summations required.
Attachments
I read - and I can not remember where or who - where it was said that the phased array of Mr. Walker will fail... much to my chagrin, since I always wanted to do this myself.
I do not recall the exact reason at this time that it would fail but I think it has something to do with what the opposing amplifier "sees".
Otoh, it could be done, money no object, by other means today.
_-_-
PS. fwiw one can hear comb filtering... or more properly, one may hear it.
I do not recall the exact reason at this time that it would fail but I think it has something to do with what the opposing amplifier "sees".
Otoh, it could be done, money no object, by other means today.
_-_-
PS. fwiw one can hear comb filtering... or more properly, one may hear it.
Being the last kid on my block to go stereo, I experimented some. I still have a University transformer that does tricks with a central speaker and satellite wings. Betcha not too many folks can remember what a "Bauer" circuit does? If you understand what it is all about and why and why not, no need to read any more of this post.
Yes, a large and very deeply mathematical literature on stereo imaging following Blumlein. Too deep by far for me follow. And whazzat "kunstkopf" recording theory, if I have the name right?
My impression is that the basic issues have to do with human perception and which cues matter to our brain and which don't.... which obviously is a question not answered by math. That's because there aint no way to put Carnegie Hall in your living room.
For sure, there are walls of sound that sound like great wonderful walls of sound. A great way to create a great wall of sound is to set up your room carefully, match a pair of drivers carefully, and then put a sound-transparent but light-tight curtain between you and the speakers. Nothing leads to better stereo.
Looking at Walker, am I right to say he has the timing theoretically right but loudness across the wall is all screwed up? If arrival of sound is all that matters, he's notion would be perfect. If loudness is all that matters, just no good. Depends on human perception principles (which, if I recall, is a little of each, depending).
Ben
Yes, a large and very deeply mathematical literature on stereo imaging following Blumlein. Too deep by far for me follow. And whazzat "kunstkopf" recording theory, if I have the name right?
My impression is that the basic issues have to do with human perception and which cues matter to our brain and which don't.... which obviously is a question not answered by math. That's because there aint no way to put Carnegie Hall in your living room.
For sure, there are walls of sound that sound like great wonderful walls of sound. A great way to create a great wall of sound is to set up your room carefully, match a pair of drivers carefully, and then put a sound-transparent but light-tight curtain between you and the speakers. Nothing leads to better stereo.
Looking at Walker, am I right to say he has the timing theoretically right but loudness across the wall is all screwed up? If arrival of sound is all that matters, he's notion would be perfect. If loudness is all that matters, just no good. Depends on human perception principles (which, if I recall, is a little of each, depending).
Ben
Last edited:
The difference in loudness and timing present in the recorded left and right stereo signals would be retained.
Looking at the image you can see that feeding the segemented wall ESL results in two waved fronts approaching the listener from equal angles, one left one right. This would simulate a set of stereo point sources positioned infinitely far away. The angle of separation dictated by the delay time chosen.
Certainly looks like even with the line properly terminated on each end the driving amplifiers would need to be able to deal with signals coming from the opposite end. It would be prudent to experiment at line-level first, or fire up the SPICE simulator.
Member
Joined 2003
With that many panels I would be tempted to try a AMT configuration.
It would be very simple to do.
There have been a few discussions on the subject but nobody has taken the time to build the many panels it takes to try it.
It is one of the things on my list of things to try sometime list.
jer 🙂
It would be very simple to do.
There have been a few discussions on the subject but nobody has taken the time to build the many panels it takes to try it.
It is one of the things on my list of things to try sometime list.
jer 🙂
"Retained"... I thought where you are sitting (say, near one corner near the R speaker) and Walker's timing change, meant that neither is retained? Possibly the timing "bends" the wave front so it works as an "infinity" source as long as you are sitting in the middle, as far as the math goes.
Again, since Blumlein first did the stereo math, math (and hence the whole stereo enterprise) hasn't been the problem for sitting in the middle or with headphones (kind of). In a way, Walker's scheme contradicts (or alters) Blumlein.
Anyway ignoring the math niceties as I am sure we are entitled to do in real rooms, multi-track recording, and contrived recording tricks, my mixed-in-the-middle panel layout does something similar by spreading a little L over towards the right side and vice versa.
Does this help produce a sound that would get higher approval from Toole when you are sitting in the middle (my original post)? Or sitting anywhere else? Is it a good idea?
Ben
By "AMT" you mean round (or spherical!) speakers in this context?
For sure, going back to Toole, after basic sound production, finagling the ambiance is the big challenge. Much thought has gone into that by speaker makers (not much at DIYaudio forum which really needs an architectural acoustic sub-forum). In a way, round speakers is an anti-theoretical approach that actually leaves ambiance entirely to how you decorate your listening room and locate your speakers.
Backing a large ESL on to a wall just might be a recipe for terrible ambiance. So partial AMT might make good sense. Walker's scheme also "lifts" the sound off the speakers.
Big ESL speakers are maybe 85% of the way there already - as far as tossing sound indiscriminately around the room in directions not directly towards the listener.
Ben
Last edited:
here are a couple of post that will give you an idea as to what I was referring to.
http://www.diyaudio.com/forums/planars-exotics/153220-electrostatic-amt-4.html#post1979874
http://www.diyaudio.com/forums/planars-exotics/153220-electrostatic-amt-4.html#post1984943
There was one post in particular that showed a drawing of a simple stacked ESL construction, But I will have to do some searching to find it.
jer 🙂
http://www.diyaudio.com/forums/planars-exotics/153220-electrostatic-amt-4.html#post1979874
http://www.diyaudio.com/forums/planars-exotics/153220-electrostatic-amt-4.html#post1984943
There was one post in particular that showed a drawing of a simple stacked ESL construction, But I will have to do some searching to find it.
jer 🙂
Here is another post showing what I am talking about but not the one I am looking for.
http://www.diyaudio.com/forums/planars-exotics/112577-idea-wide-dispersion-esl.html#post1361711
jer 🙂
http://www.diyaudio.com/forums/planars-exotics/112577-idea-wide-dispersion-esl.html#post1361711
jer 🙂
Geez, I hate it when my Alzheimer's acts up (kidding, I hope). Spherical wave-front? Pleated driver to produce a wide-dispersion line-source? Deadly beaming (which shouldn't be an issue for a multi-panel big speaker)? Can you explain?
Thanks.
Ben
Thanks.
Ben
With a finite rather than infinite segmented wall, there will be left and right boundaries to the listening area(shown in the figure) you would need to stay inside of for the image to stay centered in front of you. The listening area boundaries narrow as delay is increased and the angle of the virtual sources are spread further apart. Also, moving further away from the wall of ESLs narrows the left and right boundaries of the listening area. Would be pretty cool to have a DSP setup where you could dial in the source angles from you listening chair.
What intrigues me about this idea is that(theoretically) you could have 4 people sitting side by side on a couch listening to music each experiencing the same image spread centered directly in front of them. Obviously not a good thing for home theater where you want the sound source centered on the screen, but for sharing a music listening experience it would be nice to have everybody listening in their own personal "sweet spot"
That's a very good question...you're the one with all those panels 😉
R-C transmission lines progressively roll off high frequencies the further you get down the line.
So probably not a good fit for this application requiring full-bandwidth "steered" wave-fronts.
Last edited: