limono, no that is not the question at all. What would be done today in a cinema is another issue entirely. But since you bring it up, imo most "cinemas" today are dramatically smaller "multiplex" type rooms anyhow. Consider Radio City as the biggest of the big... but again we're talking the application being reasonably large listening spaces for "hi-fi", not theater, not PA/SR.
ra7, thanks for the link to the paper mache horns! 😀
At this point in time I am thinking that the best way to go is not a traditional tractrix... but they still could be good.
1audiohack, a great many folks, including Speaker Dave seem to argue with this premise. I think it is important...
I agree with limono that the issue of alignment of the horn with any additional HF driver WRT "time" is quite vexing and problematic, maybe the most difficult aspect of the entire concept. But, the question remains, is this a sufficiently big compromise to dismiss the rest of the benefits compared to another speaker, especially a more traditional one with the xover in the middle of the range?
_-_-bear
ra7, thanks for the link to the paper mache horns! 😀
At this point in time I am thinking that the best way to go is not a traditional tractrix... but they still could be good.
1audiohack, a great many folks, including Speaker Dave seem to argue with this premise. I think it is important...
I agree with limono that the issue of alignment of the horn with any additional HF driver WRT "time" is quite vexing and problematic, maybe the most difficult aspect of the entire concept. But, the question remains, is this a sufficiently big compromise to dismiss the rest of the benefits compared to another speaker, especially a more traditional one with the xover in the middle of the range?
_-_-bear
The Altec "Giant Voice" 290 drivers with the phenolic 'frams are supposed to cover about 300-5000Hz, tho I have not been able to get them that low. Maybe with a really big horn.
Thanks for the link to the compression drivers with the paper diaphragms, never seen anything like that!
Thanks for the link to the compression drivers with the paper diaphragms, never seen anything like that!
The problems with tractrix are driver problems. most of them are simply not designed to work with neutral horns. Back to WE horns ,I have Altec Model 15 with WE 32a bent horn and it's not a good horn really. I had WE tubes and they were beautifuly made, just pleasure to look and listen. What I find amusing is that Japanese bought most of neglected american heritage surplus for close to nothing and now sell it back charging 100 fold to american snobs (Shindo and alike)
FYI regarding horn/driver/diaphragm experience:
RE: The 4001 has too many anomalies in the highest frequencies - Bruce Edgar - High Efficiency Speaker Asylum
RE: The 4001 has too many anomalies in the highest frequencies - Bruce Edgar - High Efficiency Speaker Asylum
There is a picture from the Bell Laboratory Record (Feb 1930, p 256) showing a pair of 15A horns in towers, one in each corner of a large room, toed-in like a proper stereo setup. It might be dual mono, but why not place the horns side by side, and splay them outwards, as they usually did in the theatres?
(OT: Bell labs experimented with 2-way TV at that time 🙂 )
Regards,
Bjørn
Attachments
By coincidence, just last night, I started work on a 300-3kHz solution that's a bit out of the ordinary--a Karlson tube: 2" ID x 24" long with a 15"-long expo slot. Potentially suitable drivers in my inventory right now are the Community M200, JBL 2485, and Dynaudio D54af. It will be the centerpiece of a 3-way K-tube system.
Karlson tubes offer an interesting mix of attributes--high efficiency, wide dispersion and small form factor. I'm excited to see how it sounds and measures.
Whoa! pretty live (very reverberant) room there!
Those stands look to the eye like the same sort of stands used in the actual behind the screen set up for theaters, that included a bass horn hung in there, with the mouth on the bottom, driver on top, 90 degree turn at the bottom...
_-_-bear
Those stands look to the eye like the same sort of stands used in the actual behind the screen set up for theaters, that included a bass horn hung in there, with the mouth on the bottom, driver on top, 90 degree turn at the bottom...
_-_-bear
Imo, Karlson tubes do not work particularly well... interesting idea of course, but they work out to be merely not very good horns with a lot of comb filter effect... might be interesting if some variant of the exponential slot could be teamed with a compression driver and also provide proper loading.
Good luck with the thing... might be something to start a new thread about?
_-_-bear
Thanks! Yes, I'll start a new thread when I have something to show.
Do you have experience with K-tubes? This is my first foray.
Regarding comb filtering, just to put it in perspective, the K-tube radiates along a continuous edge, not like an array of discrete sources, so it is smoother off-axis. (The vast majority of horns create far lumpier off-axis effects due to their mouth diffraction.) Unlike most horns, the expanding slot of the K-tube makes the acoustic length of the tube indistinct. The advantage of this in comparison to most horns can be seen in the clean impulse/rapid decay.
Anyway, it's all theory until I hear/measure for myself. 🙂
Do you have experience with K-tubes? This is my first foray.
Regarding comb filtering, just to put it in perspective, the K-tube radiates along a continuous edge, not like an array of discrete sources, so it is smoother off-axis. (The vast majority of horns create far lumpier off-axis effects due to their mouth diffraction.) Unlike most horns, the expanding slot of the K-tube makes the acoustic length of the tube indistinct. The advantage of this in comparison to most horns can be seen in the clean impulse/rapid decay.
Anyway, it's all theory until I hear/measure for myself. 🙂
Last edited:
Very interesting picture and it looks like and early stereo test. This is close to the era where Bell labs did the early Stokowski stereo tests. Nice web site on that here:
Stokowski, Harvey Fletcher, and the Bell Labs Experimental Recordings
My recollection was that they primarily used a later speaker system with a square bass bin woofer and shorter HF horn, but maybe they preceeded it with this? (Just speculating)
I don't see any advantage to twin systems and a mono source and I am not aware of any cinemas using such a setup. Obviously the interference between a pair of speakers (and the same signal) would be considerable. Later, when higher output was required, a number of systems used throat adapters that allowed multiple compression drivers on one horn. Certainly Altec offered a number of systems configured that way.
I believe stereo optical soundtracks weren't common until the 50's. Disney Fantasound was more of a steered mono experiment.
Regards,
David
Nice link, thanks.
Yes, the Fletcher system, with 594 on the multicells, and a huge bass compression driver on a huge bass horn. The drivers were patented in 1933, but I'm sure they experimented with stereo before that. I'll see if I can find the article the picture above is taken from, it may reveal something.
I have a picture of a 3-horn installation somewhere. It was not uncommon to use 2, 3 and even 4 15A horns to cover the audience. Take a look at this page at moviemice, about halfway down the page. There are several drawings of possible configurations of 15A and 16A horns, along with measurements of the improved coverage from splaying two 22A horns. It's kind of a discrete multicell system. The horns are close enough to act as one source at LF, and beam enough to not interfere too much at HF. Still not an ideal solution, and probably some interference in the midrange, but better than letting 1/3rd the audience lack HF...
From the JAES paper "Fantasia - Innovations in sound" (Jan/Feb 1991):
"The basic scheme of the Fantasound playback system involves an optical film sound track, which is a control track. The control track synchronizes to a second projector with four optical sound tracks as follows: 1) control track; 2) screen left; 3) screen right; and 4) screen center. The house (surround) right, left, and rear channels are derived from the the screen right and left channels. The control track has various amplitude and frequency tones that drive Voltage Gain amplifiers(VGAs). "
I'm not sure when stereo optical sound tracks came, but it may have been born out of the Fantasound project.
Regards,
Bjørn
Yes, the Fletcher system, with 594 on the multicells, and a huge bass compression driver on a huge bass horn. The drivers were patented in 1933, but I'm sure they experimented with stereo before that. I'll see if I can find the article the picture above is taken from, it may reveal something.
I have a picture of a 3-horn installation somewhere. It was not uncommon to use 2, 3 and even 4 15A horns to cover the audience. Take a look at this page at moviemice, about halfway down the page. There are several drawings of possible configurations of 15A and 16A horns, along with measurements of the improved coverage from splaying two 22A horns. It's kind of a discrete multicell system. The horns are close enough to act as one source at LF, and beam enough to not interfere too much at HF. Still not an ideal solution, and probably some interference in the midrange, but better than letting 1/3rd the audience lack HF...
From the JAES paper "Fantasia - Innovations in sound" (Jan/Feb 1991):
"The basic scheme of the Fantasound playback system involves an optical film sound track, which is a control track. The control track synchronizes to a second projector with four optical sound tracks as follows: 1) control track; 2) screen left; 3) screen right; and 4) screen center. The house (surround) right, left, and rear channels are derived from the the screen right and left channels. The control track has various amplitude and frequency tones that drive Voltage Gain amplifiers(VGAs). "
I'm not sure when stereo optical sound tracks came, but it may have been born out of the Fantasound project.
Regards,
Bjørn
Last edited:
The tight clustering of the horns as shown in your second reference is common to this day. The intention is to get the sources as close as possible to prevent interference between their outputs. Also, by angling each unit outwards some, then the directivity broadens as the lobes of each unit add. The 16A, as two horns joined at the mouth, will guarantee coherent output from the multiple drivers. I see there are also several examples of twin compression drivers with a split manifold, to increase power handling.
What was unusual in your first photograph was that it wasn't a tight cluster, as is more typically used, but a pair of well seperated and toed in speakers. This is much like we would now use for a stereo.
Clearly this would give a great deal of interference and a "phasey" sound as you move across the midpoint between the units, but maybe that effect was not well understood at the time (and yes, this effect is audible even in the reverberent field, hence the big horn clusters you see in basketball arenas).
Thanks for the links,
David
What was unusual in your first photograph was that it wasn't a tight cluster, as is more typically used, but a pair of well seperated and toed in speakers. This is much like we would now use for a stereo.
Clearly this would give a great deal of interference and a "phasey" sound as you move across the midpoint between the units, but maybe that effect was not well understood at the time (and yes, this effect is audible even in the reverberent field, hence the big horn clusters you see in basketball arenas).
Thanks for the links,
David
Hi Bear,
I somewhat agree about the 300-3000 band.
I've always tried to do 300-5000 in one band.
Of course the WE driver does this.
Thats probably why it sounds so good.
No xover no time problem 300-5000.
That's my opinion.
For those interested in how the WE 15a launched waves at various frequencies, John Sheerin did these simulations awhile back. (click on the links marked WE15a.)
Last edited:
Hi bear
It will be a while before much of our stuff is retired and finds it’s way into homes (unless I can get folks at work more interested in the home) so I wouldn’t say any of our stuff is competitive.
Also you have identified a fatal flaw as you say.
To make a horn large enough to have a nice response curve down low, it is much too large to couple to another horn above as a single source. Unless one places sources less than about ¼ wavelength apart (like with subwoofers), you produce an interference pattern of lobes and nulls and do not get coherent addition into one source.
The crossover is also a deviation in time of what had been a single event in time. For a high pass and low pass filter like a crossover, the high frequency component emerges first and the low, last.
If the object was to reproduce the input signal as closely as possible, spreading it out in time and or frequency dependant distribution are both loudspeakers artifacts which are audible as a character or identity and ideally avoided yet unavoidable without a single broad band point source driver.
Our Synergy horns accomplish much of this and act much more like a single impossibly wide band driver than any conventional approach I have seen. They do employ crossovers but they are not ‘named types” and they accommodate the front to back spacing of the drivers and raw responses and are driven by that.
Also while multi way, all the drivers couple at dimensions less than ¼ wavelength and so they have no interference pattern, the crossovers are invisible in the phase and polar measurements as if there was one driver connected to a large horn. The mid and low drivers couple to the horn through an acoustic low pass filter so that the harmonic distortion is very low, sound above the operating range is blocked.
Some, like an SH-50, can reproduce a square wave across a broad band over a decade wide, spanning both crossovers and not just in one spot.
The limit here has been the power of one hf driver, Y throats produce an interference pattern up high and a well known 4 driver combiner is only about 4 dB louder up high than one driver. The problem is, one has a 5/8 inch wavelength at 20KHz and so to add two sources, they too must be a quarter wavelength or less apart or interfere less than a quarter wavelength.
It took some months to find the strategy, but I found a way to combine more than one hf driver into a synergy horn without producing an interference pattern two years ago and that allowed the JH series cabinets we supply to sports stadiums replace the familiar large long arrays with such a huge sonic improvement that the brand name doesn’t matter there.
The cool part is that if driver add coherently and aren’t partially canceling out, it takes fewer of them too so this way often cheaper as well.
Anyway, if any of you are in the commercial sound industry and are going to the Infocom trade show in Lost Wages, stop by and hear a different way to emulate a single impossibly wide band driver with high degree of constant directivty..
Best,
Tom Danley
It will be a while before much of our stuff is retired and finds it’s way into homes (unless I can get folks at work more interested in the home) so I wouldn’t say any of our stuff is competitive.
Also you have identified a fatal flaw as you say.
To make a horn large enough to have a nice response curve down low, it is much too large to couple to another horn above as a single source. Unless one places sources less than about ¼ wavelength apart (like with subwoofers), you produce an interference pattern of lobes and nulls and do not get coherent addition into one source.
The crossover is also a deviation in time of what had been a single event in time. For a high pass and low pass filter like a crossover, the high frequency component emerges first and the low, last.
If the object was to reproduce the input signal as closely as possible, spreading it out in time and or frequency dependant distribution are both loudspeakers artifacts which are audible as a character or identity and ideally avoided yet unavoidable without a single broad band point source driver.
Our Synergy horns accomplish much of this and act much more like a single impossibly wide band driver than any conventional approach I have seen. They do employ crossovers but they are not ‘named types” and they accommodate the front to back spacing of the drivers and raw responses and are driven by that.
Also while multi way, all the drivers couple at dimensions less than ¼ wavelength and so they have no interference pattern, the crossovers are invisible in the phase and polar measurements as if there was one driver connected to a large horn. The mid and low drivers couple to the horn through an acoustic low pass filter so that the harmonic distortion is very low, sound above the operating range is blocked.
Some, like an SH-50, can reproduce a square wave across a broad band over a decade wide, spanning both crossovers and not just in one spot.
The limit here has been the power of one hf driver, Y throats produce an interference pattern up high and a well known 4 driver combiner is only about 4 dB louder up high than one driver. The problem is, one has a 5/8 inch wavelength at 20KHz and so to add two sources, they too must be a quarter wavelength or less apart or interfere less than a quarter wavelength.
It took some months to find the strategy, but I found a way to combine more than one hf driver into a synergy horn without producing an interference pattern two years ago and that allowed the JH series cabinets we supply to sports stadiums replace the familiar large long arrays with such a huge sonic improvement that the brand name doesn’t matter there.
The cool part is that if driver add coherently and aren’t partially canceling out, it takes fewer of them too so this way often cheaper as well.
Anyway, if any of you are in the commercial sound industry and are going to the Infocom trade show in Lost Wages, stop by and hear a different way to emulate a single impossibly wide band driver with high degree of constant directivty..
Best,
Tom Danley
Tom,
From a business standpoint, I expect there's lots of reasons not to get into consumer audio. How about licensing the tech out? I'm sure someone would be interested in building on your good name.
From a business standpoint, I expect there's lots of reasons not to get into consumer audio. How about licensing the tech out? I'm sure someone would be interested in building on your good name.
The Community Light & Sound M4 driver is an example of a compression driver designed specifically to cover 250-2500 Hz, it works well 300 to 3K also. It can play at a far greater SPL level with less distortion than the W.E. drivers.
The EV DH1 on a proper horn is usable from as low as 300 Hz (at home listening levels) and sounds good to 16 kHz.
A more recent development is the BMS coaxial 4592 compression driver, with a frequency range of 300-22K Hz, having low distortion down to 600 Hz at extremely loud levels, and more than adequate level for home use (on a proper horn) down to 300 Hz.
That said, there are literally hundreds of driver options that can be combined in a single horn as DSL (Synergy series and various Paraline horns with Biblical names) and Renkus Heinz (Co-Entrant) do commercially that can give seamless reproduction over a far greater range than just 300 to 3K.
The requirement of using a single driver over a wide range makes good sense when separate horns are employed, but is no longer a necessity.
Listening, or looking at the polar, phase, distortion or frequency response charts, it is impossible to determine where the crossover points are in a well executed multiple driver horn.
Crossovers in the "speech range" no longer must be a "fatal flaw".
Although most of the commercial offerings are aimed at sound reinforcement in large venues (as was most of W.E. equipment), it is possible to make excellent sounding (arguably better sounding, certainly better in any measurable sense than W.E.) multiple driver shared horn units for very low relative cost, as the demands on drivers are much less critical when they cover a decade or less of range.
Digital signal processing and inexpensive test gear has made crossover design for seamless integration of multiple drivers relatively easy, it is now possible for a knowledgeable and competent DIYer to surpass the drivers and designs of the olden (golden?) days of audio without spending a fortune.
Art Welter
Thanks Bjørn, very cool photo. Sure does look like stereo.
Somewhere here I have a book that shows a photo of a man doing video projection at Radio City Music Hall in the 1930's. Looks like a typical 6x9 foot rear projection screen with a speaker on each side of the screen. Stereo? Probably not.
Will see if I can find and will scan and post.
Tom,
I agree with you about almost everything you have said. And, of course, the proof would be in the listening... no doubt the issue of the time delay between the tweeter and a wide range horn is problematic. But if you go with what Art Welter suggests then that might be manageable via DSP mean, assuming you believe that DSP is transparent (I do not think it is there quite yet, but would embrace it if it were).
I have a concern about your designs, been busy, but I will PM you to keep that discussion private.
The report of the ability to reproduce a square wave over a decade bandwidth is incredible!
I am unaware of any other speakers that can do that, or even come close.
Art Welter, I understand your claim, but what examples other than Tom Danley's design can you cite are multiple driver horns?
And in any case, the issue still remains of the harmonic spectra of the multiple drivers is still going to vary between the drivers. At best what one could say is that one has reduced the attendant distortions and so the audibility of differences is diminished to the point where it is no longer of concern. Of course, this is the same argument that designers of "0.001%" distortion solid state amps made for many years. Turns out that the mere reduction of distortion is insufficient to claim inaudibility.
So, the ideas are very interesting.
The next question is how can they be applied to home audio - and maybe that is another thread, not quite on topic in this one?
Before I forget, I was not impressed by what I saw on paper by the BMS coax drivers... great idea, but still not quite as smooth and clean on paper as one might wish for. At high power levels, maybe they represent a viable option, but we are talking at low power levels for home use.
I am presently covering ~250Hz. up past 10kHz, very very flat with only gentle HF EQ (1st order) to correct for the usual... the drivers you mentioned I do not think will do that well if at all. Again, home SPL levels, not PA/SR. Dunno about the EV DH1 going that high, isn't that a 2" or larger exit? What would be a proper horn in your opinion? The Community M4 is a major beast, one that I always have been interested in trying and working with... but with that size, I'd have hoped it went lower than it is spec'd for! 😀
_-_-bear
I agree with you about almost everything you have said. And, of course, the proof would be in the listening... no doubt the issue of the time delay between the tweeter and a wide range horn is problematic. But if you go with what Art Welter suggests then that might be manageable via DSP mean, assuming you believe that DSP is transparent (I do not think it is there quite yet, but would embrace it if it were).
I have a concern about your designs, been busy, but I will PM you to keep that discussion private.
The report of the ability to reproduce a square wave over a decade bandwidth is incredible!
I am unaware of any other speakers that can do that, or even come close.
Art Welter, I understand your claim, but what examples other than Tom Danley's design can you cite are multiple driver horns?
And in any case, the issue still remains of the harmonic spectra of the multiple drivers is still going to vary between the drivers. At best what one could say is that one has reduced the attendant distortions and so the audibility of differences is diminished to the point where it is no longer of concern. Of course, this is the same argument that designers of "0.001%" distortion solid state amps made for many years. Turns out that the mere reduction of distortion is insufficient to claim inaudibility.
So, the ideas are very interesting.
The next question is how can they be applied to home audio - and maybe that is another thread, not quite on topic in this one?
Before I forget, I was not impressed by what I saw on paper by the BMS coax drivers... great idea, but still not quite as smooth and clean on paper as one might wish for. At high power levels, maybe they represent a viable option, but we are talking at low power levels for home use.
I am presently covering ~250Hz. up past 10kHz, very very flat with only gentle HF EQ (1st order) to correct for the usual... the drivers you mentioned I do not think will do that well if at all. Again, home SPL levels, not PA/SR. Dunno about the EV DH1 going that high, isn't that a 2" or larger exit? What would be a proper horn in your opinion? The Community M4 is a major beast, one that I always have been interested in trying and working with... but with that size, I'd have hoped it went lower than it is spec'd for! 😀
_-_-bear