They may not think about it, but they do notice it. That is the remarkable thing about very good acoustic playback, its tone. Everyone notices it. It's colored, to be sure, but the colorations seem natural and pleasant - lifelike. From what I've read, Edison and company taught the singers to imitate the colorations of the playback to better fool the audience. My guess would be that some singer heard himself on the record and was goofing around imitating the sound. Presto! Life imitates art and whole series of demos was born.
Rare to read at this forum an ego-maniac post that so baldly shouts,"My system is FABULOUS but yours is crap."
Gosh, I wonder how time-alignment (which is just dandy on my system north of my horn sub, thank you) affects sound "down the hall"? In any case, many readers of this forum would most likely include time-alignment in any trustworthy list of the 500 most important audio design factors.
Maybe I should re-post weltersys comments to the ESL forum...might give them quite a stir. Actually, might give anybody (a) with a couple of subs, (b) with a couple of listener seats, or (c) a TH or other horn quite a stir.
B.
Last edited:
I think you are on to something.
But the distinction usually made is between some kind of engineering perfection versus some kind of personal enjoyment perfection. The two, as you point out, don't have to be quite the same thing. Indeed, if producers didn't crank up the solo parts in orchestra recordings (which obviously violates engineering objectivity)*, we'd think the sound was very deficient. Likewise for the often noted comment that bass needs to be boosted over "flat" to make it sound right at home.
Almost needless to mention, on pop recordings producers do all kinds of tricks to enhance enjoyment without fussing about objectively capturing the sound of the performers.
Ben
*in a concert setting, your "ears" point to, say, a woodwind soloist inside the orchestra (OK, I mean your attention and vision) so you attend/hear the soloist stand out like the score intended. Not possible when you are the passive recipient of sound from a recording.... so the producer cranks up the soloists to better capture the intention of the score.
Last edited:
Ben,
As Earl Geddes wrote in post #2:
"I have a recording of a classical pianist playing a piano with no room acoustics. When played back in my room, the piano IS in the room with you. It can be done."
A large source of sound like a piano can be emulated by a pair of loudspeakers with distance between them in the same room, but your goals specified in post #1:
"*When stepping down the hall or into another room, have EVER thought Rene Fleming was singing in your music room? Or even a simple flute or guitar? "
are asking a compact point source to be reproduced by a pair of loudspeakers "down the hall" and still sound like a compact point source.
Your pair of ESL speakers and tweeters (or Earl's, or anyone's speakers) can only possibly time align at one point in space, unlike Edison's single speaker reproducing a voice, and trained vocalists copying the reproduced sound.
The sound of two separated speakers, even if they were "perfect" won't sound like a single source other than at the point where they are in time alignment (presumably at the listening position) because they are not a single source.
Even using a single ESL the dispersion and radiation pattern is not similar to a human voice, a flute or guitar, so other than in the stereo triangle the illusion of reality tends to break down.
I'd expect your goal of stepping down the hall and hearing a single point source sound like a single point source would be better achieved by using just one of your ESL speakers.
Give it a try, DIY
.Art
Art - I had this same thought. We accept the completely unconvincing nature of film without question, but expect the untenable from the audio. Just as film in and of itself IS the art-form two channel playback is exactly the same way. That it can never be completely convincing recreating a large venue in a small room, neither can film make me actually believe that the building is coming down around me. To expect such a situation is simply absurd. Your brain knows that you are in a small room and it is not going to allow you to believe otherwise. But that in no way limits the scope of the medium as "it is what it is".
This was uncalled for. Art is one of the more reasonable people around here. And that's NOT what he was saying. Thus far I agree with him far more than I agree with you.
The discussion is narrowing to "Believers" and "Non-believers." I'll resist suggesting that bears any resemblance to the debate about seeing the Loch Ness Monster*.
I think it is fair to say that thinking the Chicago Symphony Orchestra led by the long-dead Fritz Reiner is in your room, whether you are in the room or down the hall, is not claimed by anyone. On the other hand, there's no argument about whether you can tell when the drummer is scrubbing his or her feet on the floor.
The Believer high-ground is to argue you should record with no room ambiance and then play back in the same room (with its ambiance being necessarily present on reproduction). Moreover, despite the fact that a single speaker can't emit sound like a singer or instrument (electric instruments excepted), two speakers can, it is claimed.
For sure, I believe you can make great music at home. I just am a Non-believer that I would be fooled using present gear into thinking Rene Fleming is in my music room (even with eyes closed and a clothespin on my nose).
Ben
*all those owning hotels around Loch Ness have seen the monster many times, sadly only when alone.
I think it is fair to say that thinking the Chicago Symphony Orchestra led by the long-dead Fritz Reiner is in your room, whether you are in the room or down the hall, is not claimed by anyone. On the other hand, there's no argument about whether you can tell when the drummer is scrubbing his or her feet on the floor.
The Believer high-ground is to argue you should record with no room ambiance and then play back in the same room (with its ambiance being necessarily present on reproduction). Moreover, despite the fact that a single speaker can't emit sound like a singer or instrument (electric instruments excepted), two speakers can, it is claimed.
For sure, I believe you can make great music at home. I just am a Non-believer that I would be fooled using present gear into thinking Rene Fleming is in my music room (even with eyes closed and a clothespin on my nose).
Ben
*all those owning hotels around Loch Ness have seen the monster many times, sadly only when alone.
All I wanted to do in starting this thread is to offer the observation that there are no design or material obstacles to making a fine coffee brewer. But today, speaker reproduction gear is a mess of compromises*.
For example why can't you buy a woofer with a free-air resonance a safe octave or two below the range demanded of that driver (like 7 Hz)? And also capable of playing just fine an octave or two above the crossover point? And not having a screwy impedance curve? And having reasonably uniform dispersion?
Ben
* I buy amps decades old because it has been decades since amps became 95% capable of doing what they need to do.
For example why can't you buy a woofer with a free-air resonance a safe octave or two below the range demanded of that driver (like 7 Hz)? And also capable of playing just fine an octave or two above the crossover point? And not having a screwy impedance curve? And having reasonably uniform dispersion?
Ben
* I buy amps decades old because it has been decades since amps became 95% capable of doing what they need to do.
Last edited:
Physics.
I don't know a lot about you or your system but I do know you don't simulate anything before you build and I don't think you measure anything after it's built or once it's placed into the system in the room.
Add to that the fact that all the products in your sig are extremely old and don't reflect current parts or design capabilities. In fact I think every product in your sig is an antique (although I can't be sure without exact model numbers.)
For example, if you are using esl mids and esl tweeters, what's the center to center distance between them and the crossover point? For a 3200 hz crossover they shouldn't be any further apart than 4 inches c-t-c maximum or you are going to have lobing and comb filtering problems. You can't fix that with eq or even dsp.
And then there's the klipschorn. I don't know much about that horn either but I doubt it's a good horn by today's standards despite Klipsh's proud statement "The only speaker in the world to be in continuous production for 60 years, and remain relatively unchanged since its inception." Here's a pic of the Khorn's frequency response from here - North Reading Engineering KV1 KV2 Crossovers - Klipsch Upgrades and Restorations
If you had a massive room mode right around 70 hz that might be bearable but otherwise ...
An externally hosted image should be here but it was not working when we last tested it.
Now the antique Stevens woofer probably has xmax of around 2 mm or so and on any reasonably sized OB it's probably grossly distorting at moderate spl. And you probably run the OB and the Khorn at the same time in the same passband and they are probably not phase cohesive.
Anyway, the point is that if you are running a bunch of antiques and ignoring best design practices it's not really fair to say that loudspeaker technology is primitive. Loudspeakers can sound spectacularly realistic if you use the best technology, set them up properly, and use a recording that was produced with this goal in mind.
Last edited:
(edited quote) I think that the above is a very valid criticism. My guess would be that there isn't much of a profit incentive in solving the above mentioned faults of speaker drivers. If as much research were poured into the design of loudspeaker drivers as say has been done with computer technology, I wouldn't doubt that contemporary speaker drivers would be much easier to work with.
I don't know how attacking my little old HiFi became of such interest in this thread. Please change the subject.
You feel my 8-octaves of ESL speakers (12kv bias, sealed in HS6 gas, formerly direct-driven by my home-brew 2400v B+ amp) are antique as compared to state of the art Rice-Kellogg drivers (General Electric patent, 1924) most others on this forum choose? What can I say?
I wish I could talk about time-alignment but when your ESL speakers are a meter square and formed on a section of a sphere and work as dipoles, a bit hard to get your arms (ooops, I mean model) around it. But as best as I can measure it, I'd say my antique tweeter ESLs (the array is in a section of a cylinder pattern) are misaligned with the centre of the ESL mids by approximately zero.
BTW, the greatly admired driver from Stephens has a free-air resonance of 20 Hz, substantial Xmax, and an edge-wound voice coil. Not bad for an antique, eh. As for the Klipsch bass, eBay might provide you with some perspective on how people feel about it.
Well, gotta go wind up my Victrola and put on a lacquer.
Pano, I tried as hard as I could to address just a guy's withering criticism of my "antique" system and failure at pre-modeling stuff. But it isn't possible to defend yourself without at the same time indicating you feel the critic was mistaken.
Ben
Funny thing, not many people have their principal speaker gear listed in their signature. Why so shy?
You feel my 8-octaves of ESL speakers (12kv bias, sealed in HS6 gas, formerly direct-driven by my home-brew 2400v B+ amp) are antique as compared to state of the art Rice-Kellogg drivers (General Electric patent, 1924) most others on this forum choose? What can I say?
I wish I could talk about time-alignment but when your ESL speakers are a meter square and formed on a section of a sphere and work as dipoles, a bit hard to get your arms (ooops, I mean model) around it. But as best as I can measure it, I'd say my antique tweeter ESLs (the array is in a section of a cylinder pattern) are misaligned with the centre of the ESL mids by approximately zero.
BTW, the greatly admired driver from Stephens has a free-air resonance of 20 Hz, substantial Xmax, and an edge-wound voice coil. Not bad for an antique, eh. As for the Klipsch bass, eBay might provide you with some perspective on how people feel about it.
Well, gotta go wind up my Victrola and put on a lacquer.
Pano, I tried as hard as I could to address just a guy's withering criticism of my "antique" system and failure at pre-modeling stuff. But it isn't possible to defend yourself without at the same time indicating you feel the critic was mistaken.
Ben
Funny thing, not many people have their principal speaker gear listed in their signature. Why so shy?
Last edited:
Well, that's like saying cars are outdated because the Model T was made a long time ago. Things have changed since 1924. Just because they were pioneered then doesn't mean the state of the art hasn't changed. What's the xmax of your Stevens woofer? Things have changed a lot since the 60's even.
You don't have to talk about time alignment. I'm talking about center to center distance which is something completely different (a whole different axis). If your esl panel is a meter square your center to center distance between the mid and the tweeter has to be at least 50 cm (unless the tweeter is physically in front of the mid). Probably closer to 60 cm (and it could be even more, depending how big the tweeter is and how far it is from the mid panel). If your mid and tweeter are separated by 2 feet center to center with a 3200 hz crossover you have some terrible comb filtering and lobing issues. That's physics, you don't need a simulation to know that. Are you familiar with this formula quoted below?
Ideally, any two drivers playing the same frequency at the crossover would be within 1/4 wavelength but that's REALLY restrictive so most people use the formula above, which allows for only about 4 inches c-t-c for a 3200 hz xo.
Also, what do you think the dispersion is like on your meter wide mid panel? What's the dispersion like on your tweeter? Do you think they blend well? Do you think there's a reason that esl panels are famous for having a small "head in a vice" sweet spot?
I know how fiercely you protest the validity of simulations but a simulation can tell you ALL of this stuff. It will show you how ugly comb filtering and lobing is when your mid and tweeter are so far apart. It can show you how dispersion will be affected by cone size. It will show where your lobes are pointed and how wide they are.
I suspect it's because the gear itself isn't as important as what you do with it. To be honest though, there's not a lot you can do with the gear in your sig. A 1 meter square mid panel is a bad idea (dispersion, comb filtering, lobing), a 60's woofer on OB is a bad idea (ultra low xmax driver in the most excursion hungry alignment possible), a Khorn seems like a bad idea (what little I know of them is summed up in the response graph I just posted). As far as I can tell there's no way to make a quality system from these parts. I'm sure if you measured and eq'ed it heavily and sat right in the sweet spot lobe and didn't move even an inch it could sound quite good. But without the heavy eq and outside the sweet spot it's not hard to tell why it doesn't sound realistic.
Last edited:
To address your edits:
Your reference point IS the subject. You can't say loudspeaker technology is primitive if all your gear is antique and explicitly defies everything we know about creating good sound.
What does "substantial Xmax" mean in mm? I could simulate your Stephens woofer OB and show you how it fails even in comparison to the inexpensive Alpha 15 on a relatively small baffle. 20 hz fs is not what you want in an OB woofer, it's way too low to be useful unless you have fairly complicated dsp capability. And the qts is probably way too low as well. But I'd be very interested to see the specs if you would care to post them, or at least a model number of the woofer.
I'm not really interested in opinions on ebay from people that don't measure frequency response. I can find a bunch of good reviews on ebay for $10 clock radios too. You have never heard a tapped horn but you consistently give them terrible reviews. 3rd party reviews are worthless unless they contain full measurements and state the measurement conditions.
Your reference point IS the subject. You can't say loudspeaker technology is primitive if all your gear is antique and explicitly defies everything we know about creating good sound.
What does "substantial Xmax" mean in mm? I could simulate your Stephens woofer OB and show you how it fails even in comparison to the inexpensive Alpha 15 on a relatively small baffle. 20 hz fs is not what you want in an OB woofer, it's way too low to be useful unless you have fairly complicated dsp capability. And the qts is probably way too low as well. But I'd be very interested to see the specs if you would care to post them, or at least a model number of the woofer.
I'm not really interested in opinions on ebay from people that don't measure frequency response. I can find a bunch of good reviews on ebay for $10 clock radios too. You have never heard a tapped horn but you consistently give them terrible reviews. 3rd party reviews are worthless unless they contain full measurements and state the measurement conditions.
Last edited:
For the purposes of this conversation there's generally no difference between a cone on a moving coil loudspeaker and a diaphragm on an esl (or even horn mouths) when talking about directivity, comb filtering and lobing. The rules are the same. The rules don't change based on size and shape of the source, but size and shape must be properly accounted for. (There's a rule of thumb for cone size vs directivity too, but it assumes you want constant flat directivity, which isn't always the best choice.)
Maybe Jim Griffin's line array paper could help here. http://www.diy-audio.narod.ru/litr/nflawp.pdf
I'm not sure if that paper would answer your questions sufficiently or not.
The Edge software is also good for this type of thing. Good crossover simulators can also show effects of directivity (of a rigid piston), lobing and comb filtering too.
Maybe Jim Griffin's line array paper could help here. http://www.diy-audio.narod.ru/litr/nflawp.pdf
I'm not sure if that paper would answer your questions sufficiently or not.
The Edge software is also good for this type of thing. Good crossover simulators can also show effects of directivity (of a rigid piston), lobing and comb filtering too.
Same reason most people keep their pants zipped up. As mentioned "the gear itself isn't as important as what you do with it".
Sorry to drag it back to the original post, but the problem is extremely difficult. For example, designing an RF antenna that covers three decades is non-trivial. What kind of antenna would cover everything from 20 MHz to 20 GHz, with response flat to +/- 3 dB?
The transduction problem is compounded by not being able to do it in one step, with the exception of ionic speakers, which have their own set of problems, like ozone emission (which poisons the listeners!). I've heard the astonishing Plasmatronics, which ionized helium as the acoustical emitter, but the practical drawback was replacing the canister of helium every month (neither cheap nor convenient).
The classical Rice-Kellogg direct radiator converts electrical current to magnetism to mechanical force to acoustical radiation; three conversion steps in all. There are losses and problems at every stage, with acoustical emitter (the diaphragm) the most difficult challenge. The problems in the diaphragm are inherent in the materials we have available, trading rigidity against self-resonance, as well as extremely poor acoustical coupling to the air, resulting in conversion efficiency typically below 1%.
What about electrostats? Not as many conversion steps, which is good, but raising the stator voltage to increase efficiency results in difficult problems with arc-over, as well as serious problems with limited excursion (compared to conventional direct-radiators). Electrical efficiency is lower than direct-radiators, and the electrical load is a nearly pure capacitance, which is especially brutal for feedback amplifiers. (To explain: a reactive load reflects the power back to the final device of the power amplifier, which is either an array of heatsinked transistors or the plates of an array of vacuum tubes. In addition to heating effects, a reactive load can drive a feedback amplifier into an unstable region of operation.)
Magnetic-planars? Even lower efficiency, thanks to very low BL factor (weak magnetic coupling). The flat impedance curve is actually the result of the low BL factor. Instead of the amplifier supplying the damping, it comes from the air-load (which perhaps is a good thing, depending on the amplifier).
Horns? The efficiency is pretty good, from 10% to 50%, but bass horns are inherently large. If you want flat response down to 30 Hz, it's going to be the size of a small car. You can decrease the size by folding, but now you get peaks and dips in the response and degraded time response, with many internal reflections. Horns also have a somewhat narrower bandwidth than direct-radiators, with a decade of useful response. The primary challenge is reducing diffraction and internal reflections; this is where Dr. Gedde's Oblate Spheroid and LeCleac'h horns and waveguides have a measurable advantage over other horn profiles.
The transduction problem is compounded by not being able to do it in one step, with the exception of ionic speakers, which have their own set of problems, like ozone emission (which poisons the listeners!). I've heard the astonishing Plasmatronics, which ionized helium as the acoustical emitter, but the practical drawback was replacing the canister of helium every month (neither cheap nor convenient).
The classical Rice-Kellogg direct radiator converts electrical current to magnetism to mechanical force to acoustical radiation; three conversion steps in all. There are losses and problems at every stage, with acoustical emitter (the diaphragm) the most difficult challenge. The problems in the diaphragm are inherent in the materials we have available, trading rigidity against self-resonance, as well as extremely poor acoustical coupling to the air, resulting in conversion efficiency typically below 1%.
What about electrostats? Not as many conversion steps, which is good, but raising the stator voltage to increase efficiency results in difficult problems with arc-over, as well as serious problems with limited excursion (compared to conventional direct-radiators). Electrical efficiency is lower than direct-radiators, and the electrical load is a nearly pure capacitance, which is especially brutal for feedback amplifiers. (To explain: a reactive load reflects the power back to the final device of the power amplifier, which is either an array of heatsinked transistors or the plates of an array of vacuum tubes. In addition to heating effects, a reactive load can drive a feedback amplifier into an unstable region of operation.)
Magnetic-planars? Even lower efficiency, thanks to very low BL factor (weak magnetic coupling). The flat impedance curve is actually the result of the low BL factor. Instead of the amplifier supplying the damping, it comes from the air-load (which perhaps is a good thing, depending on the amplifier).
Horns? The efficiency is pretty good, from 10% to 50%, but bass horns are inherently large. If you want flat response down to 30 Hz, it's going to be the size of a small car. You can decrease the size by folding, but now you get peaks and dips in the response and degraded time response, with many internal reflections. Horns also have a somewhat narrower bandwidth than direct-radiators, with a decade of useful response. The primary challenge is reducing diffraction and internal reflections; this is where Dr. Gedde's Oblate Spheroid and LeCleac'h horns and waveguides have a measurable advantage over other horn profiles.
Last edited:
Damn! All of those wasted agonizing hours making certain that sort of "info' was inaudible.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.