John, most of these "devices" and tweaks address issues caused by 2nd order effects of materials and imperfections in the electrical network of the circuit, IME. Some ideas merely change the spectrum of the distortion artifacts, so they are not solutions - but others definitely resolve issues.
Since it's very hard to get absolutely "perfect" materials, that have zero "side effects", one has to work towards an understanding of where there is an audible, negative impact, and then work out a long term solution. Best is to completely bypass the nuisance imperfections, discard them - that's why I hardwire everything I can, for example - and for the rest use standard, cheap, techniques and materials to produce a proper solution.
I have not bought an expensive, "fancy" doodah, for 30 years to get the sound I'm after - there will be smarter ways, I'm sure of that, but at least assembling a good kitbag of ways of getting there is an excellent first step ...
Yes, oil money you know. The AD8229 is in gold ceramic only package for down hole instruments at 200C. The AD8429 is exactly the same die. The designer is a good friend for more than 30 yr. we are trying to do a crazy low noise FET amp and sneak it under the radar of the current establishment.
Last edited:
I await any of your findings. In the future (at least for the next year) if you need something just ask.
jcx won't thank for this, 
, but this has been pointed to again, Human hearing beats the Fourier uncertainty principle.
What I take from this is the ability of human hearing to become highly attuned to important characteristics in the sound, we can learn to discriminate to a far higher degree than simple models of hearing would predict - the examples of individuals in the article indicate that when hearing something in the sound is important then, well, that person just learns to be able to hear it - necessity is the mother of all ...
From the article:
, but this has been pointed to again, Human hearing beats the Fourier uncertainty principle.What I take from this is the ability of human hearing to become highly attuned to important characteristics in the sound, we can learn to discriminate to a far higher degree than simple models of hearing would predict - the examples of individuals in the article indicate that when hearing something in the sound is important then, well, that person just learns to be able to hear it - necessity is the mother of all ...
From the article:
This can happen when the mid to high freqs of a single speaker start becoming more directional. A 2-3 way speaker system often maintains the wider dispersion throughout the freq range better and generates a lot more nearby wall/room 'sound' - reflections etc. A multi-way speaker system can be made more directional -- similar to a single cone but more constant; Less than 90 degree total dispersion will give similar results as you describe.
THx-RNMarsh
THx-RNMarsh
Yes, there is a continuum of behaviour in what's going on, it's achievable using a huge variety of techniques, I'm sure of that - shutting down room reflections to a major degree reduces the "load" on the brain, trying to decode what's going on - it's getting the "message" of the recording with blinkers on, so to speak.
Which makes it all the more amazing when a system does it without the aid of that directionality - the sounds are bouncing off everything in sight, yet the soundscape of the recording still just sits there, in its own space, with full integrity; the first time you experience it, you think to yourself, this is just plain nuts!!!
Which makes it all the more amazing when a system does it without the aid of that directionality - the sounds are bouncing off everything in sight, yet the soundscape of the recording still just sits there, in its own space, with full integrity; the first time you experience it, you think to yourself, this is just plain nuts!!!
Hi fas42
Yeah, this is a neat area of loudspeaker development I think as it’s linked directly to how well a stereo image can be created (if in the recording).
I think you see what I see except I would say room reflections etc are and additional factor to how the speaker radiates. The loudspeaker part of it works outside where there are no room effects.
I like your analogy about being drowned out by the information in the recording, I think that is a pretty good description of one’s 3d hearing process as It seeks information and throws out noise without our conscious knowledge until the voice or whatever is at the edge of audibility (such as listening to a conversation in a crowd).. .
The first time I heard this, I had cut the annoying protection spark gaps on my boss’s esl-63’s and had them at home to fiddle with. With a soft voice, the sound “sounded like” it was coming from behind much more so that the flat homebuilt ESL’s I made. The 63’s were made from rings of electrostatic element, driven at he center first and progressively outwards using a passive delay line. It is a Flat radiator surface that radiates a simple spherical patch all the way up pretty high, something like as if the sound had come from a single point in time and space and passed through an aperture limiting it’s angle. I thought this was curious and unlike he flat ESL’s I had made but then, I was mostly fiddling with servomotor drive woofers at work and went on.
About 15 years later, staring over with a new company and I was working on the crossover for our first 3 way Synergy horn, an SH-50. As the TDS measurements showed I was getting closer to having one source in time it also took on that weird effect, it got harder and harder to hear how far away it was with that soft voice test. These are a conical horn that begins inside the driver at about 3/8 inch in diameter and expands to a 28 inch mouth. Unlike the esl 63, these do radiate a spherical segment but start with high frequencies, then down the horn mid-range is added and farther the lf’s. There are reasons why I won’t go into here but what comes out of the mouth appears to have come from one single full range horn driver at the apex and can reproduce a square wave over a broad band with a passive xover. The drivers all add coherently so you can walk up, put your head inside the horn mouth while playing music and you cannot hear anything but the sound coming from somewhere in front of you, no individual drivers. I saw you referenced a post from a guy that rented some live sound versions and because they radiate as if the sound came from an inch or two from the rear, you really can put them flat against the wall (or more often another cabinet) and not have any interference, with two cabinets, no audible seam. That makes ALL the near wall reflections go away and the near field a lot larger.
I heard a pretty good version of this kind of simple single point radiation from a cone the first time at a small hifi gathering once. The fellow had two small fostex full range drivers on two large flat baffles. A 3 inch radiator will radiate as a simple source up to 1200-1500Hz and then it’s behavior starts to change and up high it’s in controlled breakup but up to that is a pretty broad range where it radiates a hemisphere as if it were a small simple acoustic source.
Multiple sources can add together into one new source, it easy with subwoofers because the rule is, they sources can’t be any farther apart than ¼ wavelength at the highest frequency of interest. Once two sources are ½ wavelength or more apart, they radiate independently and produce an interference pattern, seen as lobes an nulls in the polar patterns at crossover.
Anyway, fun stuff.
Best,
Tom
Hey i mentioned to Doug Jones (friend of Dick Heyser) there was a thread about Dick maybe he will pop in in this.
One of his things is stereo hearing as well. His LEDR stereo image recordings might be fun to play with.
Online LEDR Sound Test | Listening Environment Diagnostic Recording Test
> we are trying to do a crazy low noise FET amp and sneak it under the radar of the current establishment......
I think you did not mean AD8429 ?
150nA input bias current does not sound like FET input ??
I have built a 12000x low noise amplifier with a Felix Levinson front end followed by an AD8429 with 100x gain.
(No where near as clever as your circuit using the VO1263 for bias.)
The original PCB was designed for SSM2019 before I found out about the AD8429.
A small conversion board did the trick ......
Nice chip, the AD8429.
Patrick
.
I think you did not mean AD8429 ?
150nA input bias current does not sound like FET input ??
I have built a 12000x low noise amplifier with a Felix Levinson front end followed by an AD8429 with 100x gain.
(No where near as clever as your circuit using the VO1263 for bias.)
The original PCB was designed for SSM2019 before I found out about the AD8429.
A small conversion board did the trick ......
Nice chip, the AD8429.
Patrick
.
Attachments
Last edited:
Tom you are right that the ESL 63's radiate in a mostly spherical pattern, because they radiate front and back (there is some side cancellation of course). I would assume that your horns do not do that, but rather radiate in a hemispherical pattern. I also found the 63's fascinating to listen to, and they do tend to disappear, but they do NOT make the room disappear, because they do a pretty good job of illuminating the whole room.
I have spent some time with some mostly-omnidirectional speakers (that is, effectively omni through most of the audio band) and it is also a fascinating experience. There they are spraying sound literally in all directions, early reflections, late reflections, whatever, so by conventional thinking they ought to be easy to identify spatially, and be lousy at producing a stereo image. But that's not what happens. Rather, sounds have a very natural feel, because they sound like a sound produced in that room. Not only is the stereo image very good, but it is maintained very well as one moves around the room. The speakers are much harder to locate than conventional box speakers.
Every room, every environment, has a sonic signature. A room with "bad" acoustics can be quite jarring when one first enters it, but after hearing some familiar sounds (eg voices) for a few minutes one adjusts to the room and ceases to "hear" it (barring some really pathological environments, of course). If one then hears sounds that don't sound like they were produced in that room, they sound "wrong". That is the problem with limited directivity in loudspeakers (or directivity that varies widely with frequency). Since the speaker fails to illuminate the room we are robbed of those nasty "early reflections" that contribute to a room's sound. It's like listening through headphones: really good headphones can produce an amazing sound, but it never sounds "natural" like listening to sounds produced in your environment. The same with highly directional speakers where one needs to sit in the "sweet spot" to try to get only near-field sound and blot out the sonic signature of one's environment. Good omnis and dipoles do a much better job of letting the room contribute, so the net sound is effortless and sounds "right" for the environment where one is listening. This may be less "transporting" but in my experience is very natural sounding, at least with most program material.
I have spent some time with some mostly-omnidirectional speakers (that is, effectively omni through most of the audio band) and it is also a fascinating experience. There they are spraying sound literally in all directions, early reflections, late reflections, whatever, so by conventional thinking they ought to be easy to identify spatially, and be lousy at producing a stereo image. But that's not what happens. Rather, sounds have a very natural feel, because they sound like a sound produced in that room. Not only is the stereo image very good, but it is maintained very well as one moves around the room. The speakers are much harder to locate than conventional box speakers.
Every room, every environment, has a sonic signature. A room with "bad" acoustics can be quite jarring when one first enters it, but after hearing some familiar sounds (eg voices) for a few minutes one adjusts to the room and ceases to "hear" it (barring some really pathological environments, of course). If one then hears sounds that don't sound like they were produced in that room, they sound "wrong". That is the problem with limited directivity in loudspeakers (or directivity that varies widely with frequency). Since the speaker fails to illuminate the room we are robbed of those nasty "early reflections" that contribute to a room's sound. It's like listening through headphones: really good headphones can produce an amazing sound, but it never sounds "natural" like listening to sounds produced in your environment. The same with highly directional speakers where one needs to sit in the "sweet spot" to try to get only near-field sound and blot out the sonic signature of one's environment. Good omnis and dipoles do a much better job of letting the room contribute, so the net sound is effortless and sounds "right" for the environment where one is listening. This may be less "transporting" but in my experience is very natural sounding, at least with most program material.
Indeed, fun stuff
. The surprising thing, this happened first for me with very conventional speakers, B&W DM10s, about as straightforward as one can get with simple 2-ways. When the system was in "ordinary" mode the the drivers were quite obvious, it was easy to distinguish the treble emerging from where it was supposed to, and distinct from the mid/bass output - when it achieved "turbo" mode, , the two drivers coalesced into one - if you placed your ear midway between the two, at and near to the front of the cabinet, it was impossible to pick that the treble was from higher in the vertical plane, the sound was just coming from somewhere behind the front baffle. Big problem was, it slipped out of turbo, back to ordinary extremely fast - it then became the exercise to understand what had to be done to sustain the "elevated" operation ...I've done the trick many times with a variety of ordinary speakers, so it wasn't anything special about the transducers - the key was achieving exceptionally clean sound from the system overall. With conventional speakers, which require the electronics to drive them fairly hard, in terms of current draw from the mains, this is not trivial to do - a lot of care is needed, in each circumstance, to get the necessary low levels of distortion.
Sorry Frank, this is a technical nonsense, again.
BTW, even floor-standing 3way speakers "disappear", in case they are properly placed in the room, the room has reasonable reverberation time with frequency and room reflections are considered in speakers placement, and if the speakers have been designed properly.
Frank, while I would never deny the room effects, nevertheless my experience is that spekaers which work well in say my room by doing the disappearing act, will tend to do the same in any other room as well (using the same system). In one room, it will do as well as in mine, in another room not so well, true enough, but will still sound better than many a system.
My point is, I feel rooms are used far too much as an excuse for mediocre sound, caused by a poorly matched system, which will sound like it anywhere.
My point is, I feel rooms are used far too much as an excuse for mediocre sound, caused by a poorly matched system, which will sound like it anywhere.
Nonsense? Not IME - when I say distortion I am not talking specifically of straightforward THD or IMD results, but every mechanism that contributes to the resulting sound not being sufficiently 'correct'. And a clear trend is that systems which require greater draws of current, quite often in pulses with high harmonic content are the ones that have greater problems - a nice counter example was in a store where a very average AV receiver hooked up to nicely balanced, high sensitivity, benign impedance Klipsch speakers, was able to deliver PA level sound quite cleanly - way better than the majority of setups at the more recent audio show.
The higher the current, the higher the frequency of the harmonics in the current pulses, the greater the opportunity for interference effects to manifest, degrading the sound - it's been a good rule of thumb for me over the years. Everything can be solved, but are the manufacturers committed to doing so ?
With disappearing speakers, I have already said there are numerous ways - your approach is a perfectly valid one ... I'm pointing out a method which turns out to have multiple benefits: the room considerations don't matter; the "trick" works everywhere, including outside the room; it works at any volume level; and all recordings show their best side, even the worst of the worst.
Dejan, people who are sensitive to room interactions will still hear obvious impact, they will be the same as those who go to a live, non-amplified concert and react strongly to less than brilliant acoustics - I'm not one of those, I'm too busy tuning into the impact and texture of the musical content. I know people can be very different in their expectations - what I'm after is that the illusion fools you into not realising that it is reproduction, that's the goal ...
Tom, in your experience, does time aligning the driver array contribute mcuh, little or not at all?
In my case, the front baffle is at an angle of 7 degrees towards the back, and at a 3 metre distance, this puts the centers of the magnets in the same plane. My typical listening distance is about 3 metres, so I am usually in the ideal position to listen to them as far as time alignment is concerned.
I ask becaue my backup speakers, AR94 (fully refreshed, all new PP capacitors inside, new rubber suspension, pure silver wiring inside, but still the world's first 2.5 way speaker from 1986, twin 8" bass drivers of different kind, lower bass driver cut off by a single inductor at around 150 Hz) also deliver a sound way above what I'd expect from them, although not nearly as linear as my speakers (you never heard of them, local manufacture, only 7 pairs ever made, all drivers from the late Son Audax) and arranged in the traditional single line, no staggering, nothing. And not too gentle a load, AR speakers were never easy loads to drive.
You may remember how much hullabaloo was mae about time alignment in the late 70ies and early 80ies, initially by B&W DM6 ("The Fat Man"), and later taken up by Technics and KEF. At the time a friend of mine and I who were developing my speakers didn't have the time and money to make another sample but with a classic bafle, which would be the ideal way to check this out.
In my case, the front baffle is at an angle of 7 degrees towards the back, and at a 3 metre distance, this puts the centers of the magnets in the same plane. My typical listening distance is about 3 metres, so I am usually in the ideal position to listen to them as far as time alignment is concerned.
I ask becaue my backup speakers, AR94 (fully refreshed, all new PP capacitors inside, new rubber suspension, pure silver wiring inside, but still the world's first 2.5 way speaker from 1986, twin 8" bass drivers of different kind, lower bass driver cut off by a single inductor at around 150 Hz) also deliver a sound way above what I'd expect from them, although not nearly as linear as my speakers (you never heard of them, local manufacture, only 7 pairs ever made, all drivers from the late Son Audax) and arranged in the traditional single line, no staggering, nothing. And not too gentle a load, AR speakers were never easy loads to drive.
You may remember how much hullabaloo was mae about time alignment in the late 70ies and early 80ies, initially by B&W DM6 ("The Fat Man"), and later taken up by Technics and KEF. At the time a friend of mine and I who were developing my speakers didn't have the time and money to make another sample but with a classic bafle, which would be the ideal way to check this out.
Dejan, we in fact have a set of those Technics, the DIY gainclones drive them - and at one stage many years ago I had one of the Technics as the left channel, and an extremely standard, flat front, shoebox Brit 2-way on the right ... what the ... ???!!
Worked very nicely, soundscape had good integrity, no weirdness - no-one who heard the combo commented on anything untoward, they didn't realise what the setup was ... conclusion, time alignment was not a major issue,
.Exactly, Kindhornman. A poor speaker is a poor speaker, period, while a good speaker may sound better in some rooma than in others, but it will always be a good speaker.
As a personal scale, I'd say a poor speaker will depend on the room to range from poor to awful, a good speaker will range from good to very good, possibly even excellent.
My own sounds good wherever we took it to while developing it, but it did sound better in rooms which tended to be square(ish) than in clearly recangular rooms, in which it was more picky regarding positioning (the reflex port is on the back side, so that's understandable). It really comes on song when biamped by either vintage Marantz (2*85, 2*76W, bass/misd and treble) but beomes truly great when bass driven by H/K PA2400 (2*170W) and mid/treble driven by H/K Citation 24 /2*100W). The clarity and definition, especially of the soundstage width, breadth and depth really make one's heart sing. Connection to amps via van den Hul Hybrid 352 cable, 5.5 mm.sq. 2*256 strands; this size was dictated by cable length, the room outlay is such that I need 2*6m (app. 13.5 ft), and I see no sense in using extremely capable amps (peaks of 400/768 Watts 4/2 Ohms for the bass) connected with hair width, telephone grade cabling. I settled on these after extensive testing of many vables, including solid core (which provided all the "oomph!" but the range covered by the 10" woofer, up to 800 Hz, was not as coherent as with the ones I have, detail was lost.
As a personal scale, I'd say a poor speaker will depend on the room to range from poor to awful, a good speaker will range from good to very good, possibly even excellent.
My own sounds good wherever we took it to while developing it, but it did sound better in rooms which tended to be square(ish) than in clearly recangular rooms, in which it was more picky regarding positioning (the reflex port is on the back side, so that's understandable). It really comes on song when biamped by either vintage Marantz (2*85, 2*76W, bass/misd and treble) but beomes truly great when bass driven by H/K PA2400 (2*170W) and mid/treble driven by H/K Citation 24 /2*100W). The clarity and definition, especially of the soundstage width, breadth and depth really make one's heart sing. Connection to amps via van den Hul Hybrid 352 cable, 5.5 mm.sq. 2*256 strands; this size was dictated by cable length, the room outlay is such that I need 2*6m (app. 13.5 ft), and I see no sense in using extremely capable amps (peaks of 400/768 Watts 4/2 Ohms for the bass) connected with hair width, telephone grade cabling. I settled on these after extensive testing of many vables, including solid core (which provided all the "oomph!" but the range covered by the 10" woofer, up to 800 Hz, was not as coherent as with the ones I have, detail was lost.
Dejan, we in fact have a set of those Technics, the DIY gainclones drive them - and at one stage many years ago I had one of the Technics as the left channel, and an extremely standard, flat front, shoebox Brit 2-way on the right ... what the ... ???!!
Worked very nicely, soundscape had good integrity, no weirdness - no-one who heard the combo commented on anything untoward, they didn't realise what the setup was ... conclusion, time alignment was not a major issue,
In our case, Frank, the deal was this simple: why the hell not when we can anyway? Nothing to lose, perhaps something to gain? True, manufacturing is a little bit more expensive, but that's negligible in view of the total price, like +1% or some such. And it does make for a good sales point, just like mentioning the detail that our front baffle is MDF 68 mm (2.7 inches) thick and is sculptured for less refraction a la Avalon, and that the seemingly simple three way weighs in at 28.3 kilos (app. 62 lb).
- Status
- Not open for further replies.