I think your post is reasonable.
There are always diferent ways to approach any problem and Tom and I do it differently. If you look at my website you will see from the Summa paper that I started with "what" I wanted to achieve in a loudspeaker design and worked from there. I am very comfortable with my ability to predict that this would lead to a good loudspeaker system for a small room.
One can always get to a good design by cut and try, but thats a slow road and where you start determines how you will end. Usually people get to "good enough" and stop - I mean we all have to stop cutting and trying somewhere, or at least professionals do (hobbyests don't).
So Yes, I really do sit down with a hypothetical design on paper and work from there. I am well down the road before I actually build anything and, in general, I am looking to get the design right the first time, with no changes or tweaks.
And I have to chuckle about "Is it really that complicated?"
because I would look at the situation and say "Can it really be that simple?"
Its all perspective.
The ability to go from a theoretical "what I want to achieve" to a concrete set of design objectives like number of drivers, their phases and amplitudes and sizes is invaluable. But thats the way *I* do things.
There are always diferent ways to approach any problem and Tom and I do it differently. If you look at my website you will see from the Summa paper that I started with "what" I wanted to achieve in a loudspeaker design and worked from there. I am very comfortable with my ability to predict that this would lead to a good loudspeaker system for a small room.
One can always get to a good design by cut and try, but thats a slow road and where you start determines how you will end. Usually people get to "good enough" and stop - I mean we all have to stop cutting and trying somewhere, or at least professionals do (hobbyests don't).
So Yes, I really do sit down with a hypothetical design on paper and work from there. I am well down the road before I actually build anything and, in general, I am looking to get the design right the first time, with no changes or tweaks.
And I have to chuckle about "Is it really that complicated?"
because I would look at the situation and say "Can it really be that simple?"
Its all perspective.
The ability to go from a theoretical "what I want to achieve" to a concrete set of design objectives like number of drivers, their phases and amplitudes and sizes is invaluable. But thats the way *I* do things.
I used to think it was very complicated to design a good loudspeaker but its not really. There's some theory hurdles to be got over early on but the basic knowledge required to make a good loudspeaker is there for anyone to take and run with. Sure you can make it as complicated as you like during the design stages but opinion will always be divided on sound.
So no I don't think you need to be an expert, just enthusiastic with a willingness to do research.
So no I don't think you need to be an expert, just enthusiastic with a willingness to do research.
Hi Earl, musgofasa, all
Well I am not surprised there is a direct formula to do what I wanted and like is often the case, I didn’t know what it was.
I used a more mechanical approach, I used Mathcad to model the speaker height and front to back distances to the first and last rows and then had it plot out attenuation for each degree of coverage, starting at the rear seats.
Here, in the GH-60 the sources are at the same drive level and amplitude shading is done via the division of the entire length and assigning and area vs angle related to the proportion of each section, no electrical shading.
If one divided the 50 degree vertical coverage of a straight walled horn into say 32 sections, and then adjusted the level of each section by making the gap at the small end larger or smaller, you have it. A plus seems to be that while a line array is largely governed by its length, that also means what it does changes a lot with frequency or distance from the source.
In an outdoor side by side comparison / measurement of a large ribbon based line array, hung in the J shape the aimware suggested, the GH-60, an amplitude shaded point source, produced much less change in response vs distance and less change in SPL with distance and much greater front to back spl ratio.
You observation about a J source is on too, if one looks in time so far as what the vanes do, one can see that J shape if one looks at the woofers which joins in near the front edge of the vanes.
Musgofasa, you ask how does it happen?
Well, everyone has a different approach.
For me, it is often pondering one measurement that makes one think of something related later. Thinking about the refection that the wall mounted drivers produce just above there usable range, lead to the Tapped horn by asking what happens if is substitute a source at the closed end and one of opposite polarity such as one already has in the back side of the cone?.
I modeled this configuration and after a good bit of fiddling, found a geometry and driver parameters, which seemed to have “gain”. From this point, it was a matter of building a few Tapped Horns and comparing the measurements to the computer model and tweaking it based on what you learned each time. The latest generation of Tapped horns work better than the early ones, that learning process continues. I am very happy when I can model one and build it and have it be very similar.
I guess I would say my approach is to take your current understanding of how things work, when you have completed a design to the point where all the bug’s you can think of are worked out and you can picture it and all its parts working in your head, then you build one.
From here, you measure as many aspects as are relevant, when you see things that you didn’t expect, you pursue them until you understand “why”. You build another prototype with your solution.
This loop happens until what you measure is acceptable and agrees pretty much with your mental and computer model or you give up.
With the GH-60, the idea for amplitude shading sprouted from thinking about a ceiling speaker.
I had the idea of making a broad band asymmetric shaded amplitude projector, I had no idea for sure that it would work as imagined but I was pretty sure it would be very different than the Thumb rule for conventional CD horns.
Sometimes, when you have a weird idea, one your unsure of you have to build a test to see if your idea works / is worth pursuing.
As a proof of principal I though ok a ceiling speaker is one place this would be useful if it works.
I laid out a paraline lens horizontally, that was two feet across and would produce a 10 degree curved wavefront down the hall..
I divided up the aperture so that theoretically the energy projected horizontally was +20dB greater than the energy in the +45 -45 degree cell pointed straight down.
I allowed the Shading lens portion to project 6 inches below the ceiling.
What was clear in the prototype was that given its six inch projection down into the room (the size of the horn), it had far more directivity than one would expect and when hung in the air the effect was spooky.
For some distance, when you were in its horizontal beam, the sound was essentially the same loudness, for about 80 feet.
From that prototype, I was tasked to design real speaker, a system which could be arrayed into a 360 degree coverage if needed and could maintain the pattern control much lower frequency than the two foot by 10 degree paraline source permitted.
The GH-60 is meant to be hung plumb with adjacent boxes. The mouth size and geometry with it’s neighbor result in an array able system which typically has no audible seam when passing from the coverage of one box to another.
A 60 degree wide pie slice of coverage that has about 50 degrees of controlled vertical angle (+10 degrees of edge feathering at the top to suppress lobes from the loudest part)
Here, I used a paraline element vertically to combine the output of two compassion drivers into one wavefront but instead of the line source length setting its directivity, that is modified / governed by the shading lens and horn vertical boundaries, angle and large mouth size.
This way, the way it behaves is nearly constant once above the pattern control loss Frequency instead of a variable. This is (I think) why it measured / sounded so much better than the ribbon line source, it is a CD point source that projects more strongly to the back row than the front.
Cordraconis
Bass directivity another holy grail or at least marketing thing in pro-sound.
You have two approaches, one to use acoustic size to produce a favored radiation direction. Two is to use an anti-phase source to produce a cardioid like radiation pattern.
In cardioid approach, one is canceling out some of the energy the front source is producing, it is some what like governing ones speed with the brake pedal.
This is your only choice IF size is not your option.
The simplest directivity is when you have two subs positioned about 1/2 wavelength apart.
Here, one produces a significant on axis lobe forward (and back, well radially actually).
More than two sources can be used in this configuration in which case one has what is called a broad side array in antenna world.
Another useful geometry is the “End fire” antenna array.
Here, one lines up the sources in line a half wavelength apart and drives every other one with an reverse phase.
This produces a pattern similar to the broadside array but 90 degrees to the line of sources.
The problem with these geometry’s is that they work best at one frequency and only at all over a narrow band.
An end fire array can be converted to a more broad band device by placing the woofers about 1 /2 wl apart at there highest frequency of use and then driving the rear source first, then the next one and so on ,with a delayed signal, delayed an amount approximating the speed of sound and physical distance between boxes.
Lastly, placing a source on an acoustically large simple round or square baffle, causes forward directivity. It becomes a 180 degree horn in effect if it is large enough.
In general, one might say that approaches, which use physical size to produce forward directivity relative to the unaided source, make the source louder on axis by confining what had been rear radiation, forward.
Cancellation based systems reduce the sound going nearly everywhere but on axis relative to that source without the cardioid cancellation effect.
A less obvious advantage from one perspective, if one is canceling out some of the original source, with an additional source, to get less total sound than the original source, these systems take more parts, Watt’s and $ to make a given SPL and really make mfr’s happy.
Time delays and large array number and size seem the best of all so far as pattern.
So far as having two point source radiators within one horn, I would put it like this.
If your sources are so far apart that they can produce a radiation pattern, they are well too far apart to combine coherently into one source within that horn like I described.
Only in a small zone between say one quarter and one half wavelength apart can one get some “free lunch”.
A larger acoustic spacing and your producing HOM’s for sure.
Hope that helps.
Best,
Tom Danley
Well I am not surprised there is a direct formula to do what I wanted and like is often the case, I didn’t know what it was.
I used a more mechanical approach, I used Mathcad to model the speaker height and front to back distances to the first and last rows and then had it plot out attenuation for each degree of coverage, starting at the rear seats.
Here, in the GH-60 the sources are at the same drive level and amplitude shading is done via the division of the entire length and assigning and area vs angle related to the proportion of each section, no electrical shading.
If one divided the 50 degree vertical coverage of a straight walled horn into say 32 sections, and then adjusted the level of each section by making the gap at the small end larger or smaller, you have it. A plus seems to be that while a line array is largely governed by its length, that also means what it does changes a lot with frequency or distance from the source.
In an outdoor side by side comparison / measurement of a large ribbon based line array, hung in the J shape the aimware suggested, the GH-60, an amplitude shaded point source, produced much less change in response vs distance and less change in SPL with distance and much greater front to back spl ratio.
You observation about a J source is on too, if one looks in time so far as what the vanes do, one can see that J shape if one looks at the woofers which joins in near the front edge of the vanes.
Musgofasa, you ask how does it happen?
Well, everyone has a different approach.
For me, it is often pondering one measurement that makes one think of something related later. Thinking about the refection that the wall mounted drivers produce just above there usable range, lead to the Tapped horn by asking what happens if is substitute a source at the closed end and one of opposite polarity such as one already has in the back side of the cone?.
I modeled this configuration and after a good bit of fiddling, found a geometry and driver parameters, which seemed to have “gain”. From this point, it was a matter of building a few Tapped Horns and comparing the measurements to the computer model and tweaking it based on what you learned each time. The latest generation of Tapped horns work better than the early ones, that learning process continues. I am very happy when I can model one and build it and have it be very similar.
I guess I would say my approach is to take your current understanding of how things work, when you have completed a design to the point where all the bug’s you can think of are worked out and you can picture it and all its parts working in your head, then you build one.
From here, you measure as many aspects as are relevant, when you see things that you didn’t expect, you pursue them until you understand “why”. You build another prototype with your solution.
This loop happens until what you measure is acceptable and agrees pretty much with your mental and computer model or you give up.
With the GH-60, the idea for amplitude shading sprouted from thinking about a ceiling speaker.
I had the idea of making a broad band asymmetric shaded amplitude projector, I had no idea for sure that it would work as imagined but I was pretty sure it would be very different than the Thumb rule for conventional CD horns.
Sometimes, when you have a weird idea, one your unsure of you have to build a test to see if your idea works / is worth pursuing.
As a proof of principal I though ok a ceiling speaker is one place this would be useful if it works.
I laid out a paraline lens horizontally, that was two feet across and would produce a 10 degree curved wavefront down the hall..
I divided up the aperture so that theoretically the energy projected horizontally was +20dB greater than the energy in the +45 -45 degree cell pointed straight down.
I allowed the Shading lens portion to project 6 inches below the ceiling.
What was clear in the prototype was that given its six inch projection down into the room (the size of the horn), it had far more directivity than one would expect and when hung in the air the effect was spooky.
For some distance, when you were in its horizontal beam, the sound was essentially the same loudness, for about 80 feet.
From that prototype, I was tasked to design real speaker, a system which could be arrayed into a 360 degree coverage if needed and could maintain the pattern control much lower frequency than the two foot by 10 degree paraline source permitted.
The GH-60 is meant to be hung plumb with adjacent boxes. The mouth size and geometry with it’s neighbor result in an array able system which typically has no audible seam when passing from the coverage of one box to another.
A 60 degree wide pie slice of coverage that has about 50 degrees of controlled vertical angle (+10 degrees of edge feathering at the top to suppress lobes from the loudest part)
Here, I used a paraline element vertically to combine the output of two compassion drivers into one wavefront but instead of the line source length setting its directivity, that is modified / governed by the shading lens and horn vertical boundaries, angle and large mouth size.
This way, the way it behaves is nearly constant once above the pattern control loss Frequency instead of a variable. This is (I think) why it measured / sounded so much better than the ribbon line source, it is a CD point source that projects more strongly to the back row than the front.
Cordraconis
Bass directivity another holy grail or at least marketing thing in pro-sound.
You have two approaches, one to use acoustic size to produce a favored radiation direction. Two is to use an anti-phase source to produce a cardioid like radiation pattern.
In cardioid approach, one is canceling out some of the energy the front source is producing, it is some what like governing ones speed with the brake pedal.
This is your only choice IF size is not your option.
The simplest directivity is when you have two subs positioned about 1/2 wavelength apart.
Here, one produces a significant on axis lobe forward (and back, well radially actually).
More than two sources can be used in this configuration in which case one has what is called a broad side array in antenna world.
Another useful geometry is the “End fire” antenna array.
Here, one lines up the sources in line a half wavelength apart and drives every other one with an reverse phase.
This produces a pattern similar to the broadside array but 90 degrees to the line of sources.
The problem with these geometry’s is that they work best at one frequency and only at all over a narrow band.
An end fire array can be converted to a more broad band device by placing the woofers about 1 /2 wl apart at there highest frequency of use and then driving the rear source first, then the next one and so on ,with a delayed signal, delayed an amount approximating the speed of sound and physical distance between boxes.
Lastly, placing a source on an acoustically large simple round or square baffle, causes forward directivity. It becomes a 180 degree horn in effect if it is large enough.
In general, one might say that approaches, which use physical size to produce forward directivity relative to the unaided source, make the source louder on axis by confining what had been rear radiation, forward.
Cancellation based systems reduce the sound going nearly everywhere but on axis relative to that source without the cardioid cancellation effect.
A less obvious advantage from one perspective, if one is canceling out some of the original source, with an additional source, to get less total sound than the original source, these systems take more parts, Watt’s and $ to make a given SPL and really make mfr’s happy.
Time delays and large array number and size seem the best of all so far as pattern.
So far as having two point source radiators within one horn, I would put it like this.
If your sources are so far apart that they can produce a radiation pattern, they are well too far apart to combine coherently into one source within that horn like I described.
Only in a small zone between say one quarter and one half wavelength apart can one get some “free lunch”.
A larger acoustic spacing and your producing HOM’s for sure.
Hope that helps.
Best,
Tom Danley
Excuse me, Dr Geddes, when you said this:
"Easy for you, difficult for me."
It's been over 30 years since I touched Fourier Transforms, so has this been simplified to a rules or better yet a user friendly program?
Or does this require an intimate knowledge of calculus and your training and experience?
I feel the need to paraphrase Señor Wences:
"Easy for you, difficult for me."
It's been over 30 years since I touched Fourier Transforms, so has this been simplified to a rules or better yet a user friendly program?
Or does this require an intimate knowledge of calculus and your training and experience?
Tom Danley said:
Greets!
I wonder how it compares to Altec's single driver Vari Intense horns, he asks rhetorically?
http://www.lansingheritage.org/images/altec/catalogs/1993-pro/1993-24.JPG
http://www.lansingheritage.org/images/altec/catalogs/1993-pro/1993-25.JPG
Regardless, cool design TD!
GM
HK26147 said:
Well yes it would require some knowledge of calculus, what engineering problem doesn't? It also involves an understanding of "k-space" which are used a lot in optics, its the way lenses are designed. But in the end all of the usual things that we understand about time-frequency transforms hold for velocity-k-space transforms. For example the LF resolution limitation of finite sampling over a finite time equates to the inability of a finite length source to have a directional pattern. Aliasing occurs and is a fundamental limitation steming from the finite number of sources. I have just always wondered why people don't know this and use it in line array design.
There is also an equivalent relationship for a spherical source to the polar pattern or a square source, etc. With these types of transforms one can design any desired directivity and work backwards to the array that achieves it. The theory then also clearly spells out the limitations of what can be done. Not doing things this way is simply the slow road to the end result.
gedlee said:
My guess is that this level of knowledge is uncommon. For instance, I worked for a wireless phone carrier for years, and was constantly surprised at the limited knowledge of the people working on these products. You would think that the "best and the brightest" would be assigned to the challenging problems. But frequently this wasn't the case; there were plenty of "average people" who just happened to be in the right place at the right time and landed the gig.
Of course there were one or two people who were simply brilliant, but 98% of the people doing the most technical work weren't much different that people in other lines of work.
I'd venture to guess you'd find the same thing in loudspeaker companies; there are probably hundreds of professionals who have less knowledge of the subject than many hobbyists.
I chuckle as I think the simple answer to my question is "yes" and "no". In all reality, this is exactly the answer I expected only with far more detail. I really appreciate the insight from both of you gentlemen.
It proves that, over the years, my approach has been down the right track. I have always considered this to be as much art as science. Neither more than the other. For each room, space, or venue the only thing set in concrete at the beggining is the size of the location. My objective has always been to try and find a known "good" that was as close to that size as I could get. If there is no reference, then measurements must be taken and the work begun. As you both noted, you eliminate as many variables and improvements as you can by looking at what is already known and then see where you can take it from there.
I have a pretty intimate knowledge of the math involved in the calculations. there is no doubt I could put them to work to solve similar problems. The thing I would not be able to do is to build models and test them and then model the changes I intend. That would have to be left to people like you guys who have the equipment and the knowledge to use it.
What is fun for the hobbyist is our ability to keep cutting and trying till we get what we want in our own space. We don't have the problem of making something adaptable to multiple uses most of the time. That is why there is a commercial market, a professional market, and a DIY market. I am just now, after too many years, realizing this.
Thanks for the very friendly responses guys. It is this type of communication that keeps me coming back to music and speaker design year after year. Perhaps I will find an avenue to get myself back into the business one day. For now I can enjoy the learning process and try to keep myself up to date. Very interesting things happening even after so many years.
Take care,
Robert
It proves that, over the years, my approach has been down the right track. I have always considered this to be as much art as science. Neither more than the other. For each room, space, or venue the only thing set in concrete at the beggining is the size of the location. My objective has always been to try and find a known "good" that was as close to that size as I could get. If there is no reference, then measurements must be taken and the work begun. As you both noted, you eliminate as many variables and improvements as you can by looking at what is already known and then see where you can take it from there.
I have a pretty intimate knowledge of the math involved in the calculations. there is no doubt I could put them to work to solve similar problems. The thing I would not be able to do is to build models and test them and then model the changes I intend. That would have to be left to people like you guys who have the equipment and the knowledge to use it.
What is fun for the hobbyist is our ability to keep cutting and trying till we get what we want in our own space. We don't have the problem of making something adaptable to multiple uses most of the time. That is why there is a commercial market, a professional market, and a DIY market. I am just now, after too many years, realizing this.
Thanks for the very friendly responses guys. It is this type of communication that keeps me coming back to music and speaker design year after year. Perhaps I will find an avenue to get myself back into the business one day. For now I can enjoy the learning process and try to keep myself up to date. Very interesting things happening even after so many years.
Take care,
Robert
Tom Danley said:
I know this thread is bit old but I just joined so I missed this one!
The idea of a shaded ceiling speakers sound really interesting to me. I was just trying imagine what materials would be good for those dividing walls?
On the one hand it needs to be quite thin(I think)
On the other hand it needs to be rather rigid
Any other details you might be willing to share without compromising company secrets would be appreciated
Tom, once again apologies for the late post but I have also just joined this forum.
You mention the GH60 measurements taken to 150 feet but what happens after this sort of distance? I have always been amused (bemused?) to see line arrays being used outdoors with all the hyperbola about their tremendous 'throw' (yes, that word again) but then seeing delay towers every 40 or 50 metres to reach the back of the arena or field. Are the GH60s capable of extremely long throw say 100-150 metres or do you see the need for a larger version or a 'top box' with narrow(er) Horizontal and/or vertical dispersion to reach this distance? Are there other physical difficulties in designing a speaker to throw several hundred metres - other than the amount of power needed or overcoming HF absorption and the like?
Regards,
Jazomir
Hey Jazomir,
Although line arrays or Finite Line sources can throw long distances there's a couple of reasons not too.
1. It seems that Line sources seem more prone to wind effects, using a shorter throw distance can help obviate this.
2. If you consider, as pointed out earlier, that directivity is a function of line source length, then we have an array thats pushing HF in a narrow controlled pattern but the LF is basically acting in an omni fashion, Unless some other means of directivity control is employed (horn loaded). Given the conventional position of the system alongside the performers you can see this might be an issue with open microphones.
So it doesn't look like delay stacks are going away soon.
Although line arrays or Finite Line sources can throw long distances there's a couple of reasons not too.
1. It seems that Line sources seem more prone to wind effects, using a shorter throw distance can help obviate this.
2. If you consider, as pointed out earlier, that directivity is a function of line source length, then we have an array thats pushing HF in a narrow controlled pattern but the LF is basically acting in an omni fashion, Unless some other means of directivity control is employed (horn loaded). Given the conventional position of the system alongside the performers you can see this might be an issue with open microphones.
So it doesn't look like delay stacks are going away soon.
Jazomir said:
Yes, a couple. Those distances are only going to happen outdoors. A light breeze could ruin your sound at that distance. Also, you would have to mount the speakers insanely high to keep from blowing out the front row
Maybe not impossible to solve, but who would when delay towers would keep things in check more practically
Martin,
I would have thought that any wind-induced movement would have been a result of the low energy of the high frequencies, which would affect line & point source arrays similarly rather like the eddies seen around the Gulf Stream where the friction or temperature gradient between cold & warm waters induces breakaway streams to form. I have a sneaking feeling that this is something that the folks at Funktion One utilise when overseeing outdoor set-ups of their rigs - they are aimed 'high', possibly to induce a buffer zone to reduce interference at high frequencies in a similar way that feathering the flare of a CD horn does. However, from
This the point that Tom Danley was trying to make when he talked about the downsides of the 'line array' - after all, even with line arrays, you usually have the subs ground stacked either centrally or in a 'stereo' arrangement either side of the stage (unless you are flying the subs, when you will need a very long array to achieve a good balance of directivity & efficiency). Alternatively, utilising the directivity offered by precisely stacked horns (tapped or otherwise) or utilising cardioid arrays (accepting the power wastage these present) the proximity of the subs to the performers ought not to be a problem. Having a precise radiation pattern at higher frequencies (as in the synergy horn) would also seem to obviate this problem.
Regards,
Jazomir
Huh, a sleeping thread has awoken.
Jazomir you ask about “throw” and about making a deep projection box and issues with that.
Ferrit mentions wind effects etc and actually all of these things tie together in a way.
Throw is a misnomer perhaps, normally one thinks of long throw as being a narrow coverage angle horn.
In reality, it is long throw because it's narrow angle allows it to project more energy into a smaller angle, thus appears to have a higher sensitivity and maximum output..
A line source is one which one dimension is very large acoustically “infinitely large” which is the asymptote on the graph of change vs size.
When it is infinitely long acoustically, it radiates a cylindrical wavefront with an impulsive input, viewed from the side, a flat plane wave.
Radiating into one plane instead of two, H and V, the sound pressure fall off is half that of a point source or –3 dB per doubling of distance from the source..
In real life, line arrays are not infinitely long and so their acoustic size is a constant function of frequency. This variable acoustic size means there is a narrow pattern in one plane up high and the system is a point source at the bottom end.
Also, the real line sources are made of separate sources, often too far apart to combine coherently so there is self interference in the vertical plane especially..
These things cause several unexpected things.
In front of the line array set up out doors where there are no room effects, you measure the response, then you move back and measure again and now it’s different, move further back and its more different, you are seeing the spectral balance change because the system has differing directivities as a function of frequency as a result of length.
Line arrays made of acoustically separate sources also typically have very ugly repose because of fine self interference between sources and so nice color bars are used in prediction programs vs high res measurements.
These acoustic size / distance issues can be reduced by curving the array like a J or banana and or tapering the response and amplitude, then one has created system with two origins or focal points, this is a point source with astigmatic distortion as in optics.
Also though, to the degree one moves toward a point source, one is moving away from the reduced SPL fall off.
Wind effects are a problem for both a point source and line array system.
The degree of problem subjectively is related to how homogeneous the pattern is as it is moved around by the wind. A speaker that sounds exactly the same up and down, left and right would have MUCH less problem in wind than one where everywhere was different as the pattern blew around.
A “point source” I would describe first as a sound source small enough that it has no directivity of it’s own and so radiates spherically, in X and in Y.
One can assume that a source that was a quarter wavelength or less across does this.
If you place a point source at the apex of a plain conical horn, one finds it radiates a spherical segment at a fixed angle (constant directivity) and down to a frequency governed by the mouth dimension and horn wall angle.
Because the radiation is still spherical, the pressure falls off at –6 dB per doubling of distance as before with a full point source.
What one finds too with a real point source is that unlike the line array, the spectral balance does not change as you move away, just the loudness.
Making a full range point source is no small task but can be done haha.
The Genisis is a different approach to a point source, which can do what you can’t normally do by having a specific amplitude shaping vs angle.
An example of that is a variation used at turner field in Atlanta.
Picture 2 and 3 show a goofy speaker along the edge of the 2nd deck.
http://www.danleysoundlabs.com/danleyport.asp?ID=61
These are a shaded amplitude Genesis horn which faces down and in picture 3 you can see the coverage. This coverage starts directly below the speaker and ends at row 1 at the dirt warning track.
The people directly below are 20 feet away, the first row is 110 feet for reference the horn mouth is 40 inches in the front / back dimension..
As a result of the shaded amplitude, the contractor measured the first row (farthest) was about 1dB louder than the level directly below the speaker and by the time you get to the grass, it’s –16dB down already.
A whole pile of our speakers were used here, maybe someone here has heard the system this season and can comment but I haven’t actually heard it yet.
I can tell you that in the not distant future, there will be a Genesis type shaded amplitude speakers for large scale concert / NFL stadium use.
At least over some area, one can eliminate the fall off with distance and make a significant span a constant amplitude and spectrum.
For installation and commercial sound, they are already proving effective but like other things, there is a learning curve when using new inventions dang it.
I wish I had Earl’s math ability, I am stuck doing it the hard way bit by bit..
Hope that helps fill in some blanks.
Best,
Tom Danley
Jazomir you ask about “throw” and about making a deep projection box and issues with that.
Ferrit mentions wind effects etc and actually all of these things tie together in a way.
Throw is a misnomer perhaps, normally one thinks of long throw as being a narrow coverage angle horn.
In reality, it is long throw because it's narrow angle allows it to project more energy into a smaller angle, thus appears to have a higher sensitivity and maximum output..
A line source is one which one dimension is very large acoustically “infinitely large” which is the asymptote on the graph of change vs size.
When it is infinitely long acoustically, it radiates a cylindrical wavefront with an impulsive input, viewed from the side, a flat plane wave.
Radiating into one plane instead of two, H and V, the sound pressure fall off is half that of a point source or –3 dB per doubling of distance from the source..
In real life, line arrays are not infinitely long and so their acoustic size is a constant function of frequency. This variable acoustic size means there is a narrow pattern in one plane up high and the system is a point source at the bottom end.
Also, the real line sources are made of separate sources, often too far apart to combine coherently so there is self interference in the vertical plane especially..
These things cause several unexpected things.
In front of the line array set up out doors where there are no room effects, you measure the response, then you move back and measure again and now it’s different, move further back and its more different, you are seeing the spectral balance change because the system has differing directivities as a function of frequency as a result of length.
Line arrays made of acoustically separate sources also typically have very ugly repose because of fine self interference between sources and so nice color bars are used in prediction programs vs high res measurements.
These acoustic size / distance issues can be reduced by curving the array like a J or banana and or tapering the response and amplitude, then one has created system with two origins or focal points, this is a point source with astigmatic distortion as in optics.
Also though, to the degree one moves toward a point source, one is moving away from the reduced SPL fall off.
Wind effects are a problem for both a point source and line array system.
The degree of problem subjectively is related to how homogeneous the pattern is as it is moved around by the wind. A speaker that sounds exactly the same up and down, left and right would have MUCH less problem in wind than one where everywhere was different as the pattern blew around.
A “point source” I would describe first as a sound source small enough that it has no directivity of it’s own and so radiates spherically, in X and in Y.
One can assume that a source that was a quarter wavelength or less across does this.
If you place a point source at the apex of a plain conical horn, one finds it radiates a spherical segment at a fixed angle (constant directivity) and down to a frequency governed by the mouth dimension and horn wall angle.
Because the radiation is still spherical, the pressure falls off at –6 dB per doubling of distance as before with a full point source.
What one finds too with a real point source is that unlike the line array, the spectral balance does not change as you move away, just the loudness.
Making a full range point source is no small task but can be done haha.
The Genisis is a different approach to a point source, which can do what you can’t normally do by having a specific amplitude shaping vs angle.
An example of that is a variation used at turner field in Atlanta.
Picture 2 and 3 show a goofy speaker along the edge of the 2nd deck.
http://www.danleysoundlabs.com/danleyport.asp?ID=61
These are a shaded amplitude Genesis horn which faces down and in picture 3 you can see the coverage. This coverage starts directly below the speaker and ends at row 1 at the dirt warning track.
The people directly below are 20 feet away, the first row is 110 feet for reference the horn mouth is 40 inches in the front / back dimension..
As a result of the shaded amplitude, the contractor measured the first row (farthest) was about 1dB louder than the level directly below the speaker and by the time you get to the grass, it’s –16dB down already.
A whole pile of our speakers were used here, maybe someone here has heard the system this season and can comment but I haven’t actually heard it yet.
I can tell you that in the not distant future, there will be a Genesis type shaded amplitude speakers for large scale concert / NFL stadium use.
At least over some area, one can eliminate the fall off with distance and make a significant span a constant amplitude and spectrum.
For installation and commercial sound, they are already proving effective but like other things, there is a learning curve when using new inventions dang it.
I wish I had Earl’s math ability, I am stuck doing it the hard way bit by bit..
Hope that helps fill in some blanks.
Best,
Tom Danley
Tom, thanks for the recap on line array theory and in particular the explanation for the audible differences caused by the effect of wind on line arrays and point sources - that was most informative.
Also, other than the high frequency-related artefacts shown by line arrays and highlighted in your comparison with the GH60, are there any major problems that present themselves in extreme long throw cabinet design that current technology (your own included) has yet to overcome?
Regards,
Jazomir
Edge feathering - is this the same as the edge feathering in the outer flare of a CD horn which I believe serves the same purpose?
Also, other than the high frequency-related artefacts shown by line arrays and highlighted in your comparison with the GH60, are there any major problems that present themselves in extreme long throw cabinet design that current technology (your own included) has yet to overcome?
Regards,
Jazomir
Tom Danley said:
Hi Tom
Thanks, but I don't think math ability does much in the end. It appears that people skills are what I need. Doesn't seem to matter how good the stuff is you gotta kiss _ss if you want sell your product
On the wind thing, wind affects sound in highly nonintuitive ways. Thats because the velocity has almost no effect, its too slow, and the major effect is density changes. There are reported events where sound has traveled much farther upsteam than downsteam. This was the result of a density gradient effect.
Hi Earl, Jazomir
“It appears that people skills are what I need. Doesn't seem to matter how good the stuff is you gotta kiss _ss if you want sell your product ”
Well I can’t say I know how to sell speakers but I can say that one’s ability to argue a point is unrelated to the validity of that idea .
On line and maybe because audio is a hobby or personal pursuit, you occasionally encounter the emotional baggage that can tag along in a technical argument.
This makes the engineering discussion not about X Yor Z but about “being right” even if technically wrong. I ran into this and you have too, with at least one of the same people and with some it’s hard to stop.
There will always be some fraction who will not see or cannot see a technical argument regardless of how thoroughly it’s explained.
From everything I have heard, your speakers sound very good THAT is what your sales angle has to be, not convincing individuals who have not heard them and don’t want to concede your point not technically but because of pride.
Earl I have read a bunch of your papers starting with the bandpass systems a long time ago, I took a deep breath when you stood up to ask a question at AES when I was giving a talk on the Servodrives at AES ages ago.
I would consider you among the very smartest people have known in audio, possibly the sharpest in math of all.
That isn’t going to sell speaker for you but it’s certainly not going to hurt either in fact you do have a very nice resume which should be part of it.
I can tell you a couple things about the business I am in now that knock on wood is doing well. These are different than times past for me and may or may not be important for you..
Our business is mostly going into installations, work which is designed then specified and out on bid and then installed. We sell to the contractor who got the job to install.
If you have something, which actually does a better job doing something, then a side by side demo is often enough to get your widget specified on a trial job.
If the end customer is happy, then the contractor is happy then your happy because he is likely to use the product again.
If it makes sense in your case, demo the people who can design in your speakers to home theater installations or like dealers that can sell them for you.
Beyond a given size, it takes people selling for you.
I do not try to sell anything, I don’t know what our stuff costs more than approximately for a few things .
That way by honestly not knowing the selling prices off the top of my head, I don’t get very involved in the least fun part of speakers.
I had always sort of decided what to design based on my best guess at adapting the technology at hand to the imagined need.
I was pretty surprised how often customers used the speakers I designed to do things I never thought of.
I still do that occasionally, entirely follow a whim, but now what has worked better is to have the system designers and installers tell me what they need and then design around that.
In your case with the home market, you already have a nice looking design and like they say, sell what you have first, have to find the right or better ”how” maybe.
Fwiw, the cabinet shop we use QMS is getting into short run molded parts.
Our first is a 60 by 60 horn with an 18 inch mouth and so far so good.
These are tooled with an SLA or milled part which is then used to make the mold which then clamps together in the needed places. This is a cold cast urethane similar to what we are using on the sh-100 and 100B cabinets (Protoshapes in Michigan I think).
Anyway, the SLA part isn’t exactly cheap but the process is way cheaper than tooling and molding the old way.
If you want their number and contact let me know, they are good folks.
Anyway, hope things are well with you and yours, I imagine you’ve had a cool wet spring too.
Jazomir asked about edge feathering and the flare near the end of a conical horn.
Don Keele described a pattern loss rule of thumb for the point where the horn is no longer able to confine the radiation to an angle defined by the horn walls.
Also he described that as you approach the pattern loss frequency , that right before you loose pattern control, the pattern momentarily becomes narrower, then wider.
He described how making the outer third of the conical horn a larger angle, that this waste banding was eliminated or reduced.
This same “pattern controlling dimension is why say an exponential horn’s radiation angle decreases and the frequency increases, the controlling point moves back towards the driver where the horn is smaller and narrower.
In any case, making the mouth bigger or angle larger lowers the pattern loss frequency AND making a simple conical horn significantly asymmetric, causes “pattern flip” because the narrow angle looses pattern control first because it’s dimension is too small.
An old EV T-35 tweeter worked better mounted “up and down” because pattern flip was so strong more energy was radiated 90 degrees to the obvious than the obvious.
The pattern loss thumb rule is pretty accurate for a simple conical horn and like I mentioned, if one makes the horn angle larger or makes the mouth larger across, then the pattern control frequency goes down.
I lack the descriptor for the feathering in the shaded horn here.
Picture that the last 15 degrees of a conical horn’s pattern were made of say 5 cells, each 3dB lower in level than the towards the center.
In other words, before one has reached the angle defined by the outer wall, one has tapered the amplitude in each cell by say 15 dB by the time you reach the outer horn wall..
Doing this tapering at the edges appears to allow more directivity with a nicer polar pattern than not.. On the other hand, I have not used feathering on all of them, the ones in the ball park were only shaded to account for the distance and angle to the seats front to back. The feathering seems most useful when at the edge of a very narrow pattern but ones concept of operation must be dictated by the measured results so this is evolving.
So far as “long distance sound”, with our point sources do that now, week before last in the town the company is in, Mike set up a system at for a town event which involved canned music and then music with fireworks.
He used one TH-812 subwoofer and one SH-96 with an SH-46 on top set in a stack.
It was powered by one amplifier, one channel for bass, one for the two tops with a simple electronic crossover.
At 520 feet, at the edge of the park they read 90dBa SPL slow weighted with the heavier program with what they said was wonderful sound.
The trick for doing this indoors for an audience is making the loudness and spectrum near constant for the seats and projecting as little sound as possible where the people aren’t which only drives the room’s reverberant field. Obviously, vastly harder..
The non loudspeaker problems with really long distances are “steering” and absorption.
Hf absorption is more of a minor issue at normal distances “far away” it’s why thunder is just a low bass rumble. I suppose if you talking really far away, getting enough acoustic power is not insignificant as the more sources you have at a frequency, the more interference your likely have (one reason why large concert sound systems sound little like the input signal)
Like Earl mentioned weird effects can be density or thermally caused.
For common example imagine a sound system set up at an open field or parking lot.
After dark, the ground is warm, the air above it is cool and remember the speed of sound goes up with temperature.
A band playing might find that their sound simply disappears a few hundreds of feet away.
What happens is the bottom part of the radiating wave is traveling with a slightly higher velocity and gradually bends upwards and away.
At Redstone arsenal they found they needed to have large horns on towers to test the air before testing a Saturn five engine.
This “need” was brought about because occasionally when they tested, they would break windows in towns about 20 miles away.
In this case, the upper air was warmer which gradually bent that nice fat bass (sounds like thunder that doesn’t stop) back down to earth and better yet, they found there can even be focusing.
Anyway, a toot on the big bass horns and readings from the microphones in the towns with problems, told them when it was ok to test.
I have run on, time to stop haha,
Best,
Tom
“It appears that people skills are what I need. Doesn't seem to matter how good the stuff is you gotta kiss _ss if you want sell your product ”
Well I can’t say I know how to sell speakers but I can say that one’s ability to argue a point is unrelated to the validity of that idea .
On line and maybe because audio is a hobby or personal pursuit, you occasionally encounter the emotional baggage that can tag along in a technical argument.
This makes the engineering discussion not about X Yor Z but about “being right” even if technically wrong. I ran into this and you have too, with at least one of the same people and with some it’s hard to stop.
There will always be some fraction who will not see or cannot see a technical argument regardless of how thoroughly it’s explained.
From everything I have heard, your speakers sound very good THAT is what your sales angle has to be, not convincing individuals who have not heard them and don’t want to concede your point not technically but because of pride.
Earl I have read a bunch of your papers starting with the bandpass systems a long time ago, I took a deep breath when you stood up to ask a question at AES when I was giving a talk on the Servodrives at AES ages ago.
I would consider you among the very smartest people have known in audio, possibly the sharpest in math of all.
That isn’t going to sell speaker for you but it’s certainly not going to hurt either in fact you do have a very nice resume which should be part of it.
I can tell you a couple things about the business I am in now that knock on wood is doing well. These are different than times past for me and may or may not be important for you..
Our business is mostly going into installations, work which is designed then specified and out on bid and then installed. We sell to the contractor who got the job to install.
If you have something, which actually does a better job doing something, then a side by side demo is often enough to get your widget specified on a trial job.
If the end customer is happy, then the contractor is happy then your happy because he is likely to use the product again.
If it makes sense in your case, demo the people who can design in your speakers to home theater installations or like dealers that can sell them for you.
Beyond a given size, it takes people selling for you.
I do not try to sell anything, I don’t know what our stuff costs more than approximately for a few things .
That way by honestly not knowing the selling prices off the top of my head, I don’t get very involved in the least fun part of speakers.
I had always sort of decided what to design based on my best guess at adapting the technology at hand to the imagined need.
I was pretty surprised how often customers used the speakers I designed to do things I never thought of.
I still do that occasionally, entirely follow a whim, but now what has worked better is to have the system designers and installers tell me what they need and then design around that.
In your case with the home market, you already have a nice looking design and like they say, sell what you have first, have to find the right or better ”how” maybe.
Fwiw, the cabinet shop we use QMS is getting into short run molded parts.
Our first is a 60 by 60 horn with an 18 inch mouth and so far so good.
These are tooled with an SLA or milled part which is then used to make the mold which then clamps together in the needed places. This is a cold cast urethane similar to what we are using on the sh-100 and 100B cabinets (Protoshapes in Michigan I think).
Anyway, the SLA part isn’t exactly cheap but the process is way cheaper than tooling and molding the old way.
If you want their number and contact let me know, they are good folks.
Anyway, hope things are well with you and yours, I imagine you’ve had a cool wet spring too.
Jazomir asked about edge feathering and the flare near the end of a conical horn.
Don Keele described a pattern loss rule of thumb for the point where the horn is no longer able to confine the radiation to an angle defined by the horn walls.
Also he described that as you approach the pattern loss frequency , that right before you loose pattern control, the pattern momentarily becomes narrower, then wider.
He described how making the outer third of the conical horn a larger angle, that this waste banding was eliminated or reduced.
This same “pattern controlling dimension is why say an exponential horn’s radiation angle decreases and the frequency increases, the controlling point moves back towards the driver where the horn is smaller and narrower.
In any case, making the mouth bigger or angle larger lowers the pattern loss frequency AND making a simple conical horn significantly asymmetric, causes “pattern flip” because the narrow angle looses pattern control first because it’s dimension is too small.
An old EV T-35 tweeter worked better mounted “up and down” because pattern flip was so strong more energy was radiated 90 degrees to the obvious than the obvious.
The pattern loss thumb rule is pretty accurate for a simple conical horn and like I mentioned, if one makes the horn angle larger or makes the mouth larger across, then the pattern control frequency goes down.
I lack the descriptor for the feathering in the shaded horn here.
Picture that the last 15 degrees of a conical horn’s pattern were made of say 5 cells, each 3dB lower in level than the towards the center.
In other words, before one has reached the angle defined by the outer wall, one has tapered the amplitude in each cell by say 15 dB by the time you reach the outer horn wall..
Doing this tapering at the edges appears to allow more directivity with a nicer polar pattern than not.. On the other hand, I have not used feathering on all of them, the ones in the ball park were only shaded to account for the distance and angle to the seats front to back. The feathering seems most useful when at the edge of a very narrow pattern but ones concept of operation must be dictated by the measured results so this is evolving.
So far as “long distance sound”, with our point sources do that now, week before last in the town the company is in, Mike set up a system at for a town event which involved canned music and then music with fireworks.
He used one TH-812 subwoofer and one SH-96 with an SH-46 on top set in a stack.
It was powered by one amplifier, one channel for bass, one for the two tops with a simple electronic crossover.
At 520 feet, at the edge of the park they read 90dBa SPL slow weighted with the heavier program with what they said was wonderful sound.
The trick for doing this indoors for an audience is making the loudness and spectrum near constant for the seats and projecting as little sound as possible where the people aren’t which only drives the room’s reverberant field. Obviously, vastly harder..
The non loudspeaker problems with really long distances are “steering” and absorption.
Hf absorption is more of a minor issue at normal distances “far away” it’s why thunder is just a low bass rumble. I suppose if you talking really far away, getting enough acoustic power is not insignificant as the more sources you have at a frequency, the more interference your likely have (one reason why large concert sound systems sound little like the input signal)
Like Earl mentioned weird effects can be density or thermally caused.
For common example imagine a sound system set up at an open field or parking lot.
After dark, the ground is warm, the air above it is cool and remember the speed of sound goes up with temperature.
A band playing might find that their sound simply disappears a few hundreds of feet away.
What happens is the bottom part of the radiating wave is traveling with a slightly higher velocity and gradually bends upwards and away.
At Redstone arsenal they found they needed to have large horns on towers to test the air before testing a Saturn five engine.
This “need” was brought about because occasionally when they tested, they would break windows in towns about 20 miles away.
In this case, the upper air was warmer which gradually bent that nice fat bass (sounds like thunder that doesn’t stop) back down to earth and better yet, they found there can even be focusing.
Anyway, a toot on the big bass horns and readings from the microphones in the towns with problems, told them when it was ok to test.
I have run on, time to stop haha,
Best,
Tom
Hey Tom
Yea, we have run into the same "people"
I am well aware of the install market, but unfortunately HT is very driven by trade names. Without one of those you won't get into a contractors specs.
Go how I wish I had a sales staff!!
I've used SLA before, many times. Its idea for quick turn arround. Downside that I found, along with the cost, is they are very brittle. Its like working with soft glass. My polyurethane waveguides can be hit directly with a sledge hammer and won;t budge.
Given I've only been doing this a year, I now have five products and enough sales to fill the pipeline. Couldn't ask for anything more.
Yea, we have run into the same "people"
I am well aware of the install market, but unfortunately HT is very driven by trade names. Without one of those you won't get into a contractors specs.
Go how I wish I had a sales staff!!
I've used SLA before, many times. Its idea for quick turn arround. Downside that I found, along with the cost, is they are very brittle. Its like working with soft glass. My polyurethane waveguides can be hit directly with a sledge hammer and won;t budge.
Given I've only been doing this a year, I now have five products and enough sales to fill the pipeline. Couldn't ask for anything more.
Tom Danley said:
I agree with this 110%. I am in the software biz, and have met a handful of guys who've made seven figures in the industry. 30 year old millionaires aren't terribly uncommon in Bellevue, Sunnyvale, and Santa Monica.
It's basically an entirely different demographic than the DIY crowd. The DIY crowd is very knowledgeable, but demanding and frugal. The millionaire crowd is often looking for something durable and flashy, something that will "wow" their friends and is unique. I think the Summa and the Synergy Horns fits the bill pretty well, except for the cosmetics. (sorry guys, neither one is pretty
But my girlfriend likes the space age finish There's a lot of good alternatives to SLA. Even if you stick with an RP process you can get a much more durable piece. There are lots of cheap FDM sources available which can print in almost anything, but there's other good choices too. If I remember rightly there was a pretty big company in Georgia(which is close to you?) we almost used for a John Deere project(though smaller companies have advantages too). I can dig it up if you want. I would carefully consider putting and SLA process piece in any tour grade equipment
RP is great for devloping and testing. I've seen some DIYers do some rather impossible/artistic designs too. But as soon as you get to a production quantity of about.... 5 (and even 3)pieces I do much better with other methods.
There is a particular advantage though, that FDM may have for an "evil genius" like gedlee with the possibilty of samples with shapes not possible to machine(and with 0.0006in precision, in a durable material)
See link : http://www.objet.com/3D-Printer/Eden350_Eden350V/
This possibly could allow for CD Phase plugs whose designs are sonically superior but impractical to produce otherwise.
Speaking of which, has Mr. Danley considered using a phase plug(of sorts) between the walls of the horn and woofer(s) of a synergy type horn? Probably an asymmetric one considering the vents are edge mounted. Now that you are switching to molded maybe some more exotic shapes can be had.
PS "evil genius" is a compliment of the highest honor in my family, Please don't misunderstand
RP is great for devloping and testing. I've seen some DIYers do some rather impossible/artistic designs too. But as soon as you get to a production quantity of about.... 5 (and even 3)pieces I do much better with other methods.
There is a particular advantage though, that FDM may have for an "evil genius" like gedlee with the possibilty of samples with shapes not possible to machine(and with 0.0006in precision, in a durable material)
See link : http://www.objet.com/3D-Printer/Eden350_Eden350V/
This possibly could allow for CD Phase plugs whose designs are sonically superior but impractical to produce otherwise.
Speaking of which, has Mr. Danley considered using a phase plug(of sorts) between the walls of the horn and woofer(s) of a synergy type horn? Probably an asymmetric one considering the vents are edge mounted. Now that you are switching to molded maybe some more exotic shapes can be had.
PS "evil genius" is a compliment of the highest honor in my family, Please don't misunderstand
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.