What are some really good values in speakers?

[MrBoat]

The fact that they can now be designed by computers, and the fact that most horns are now designed for the purpose of pattern control vs just loading. If you dig around and look at some of the research and designs introduced by Klipsch and Harman, you’ll find both of them have gone to great length to perfect them, with an emphasis on uniform controlled dispersion. Not surprisingly, both jbl and Klipsch horn designs are use similar principles in geometry. None of the modern designs used by both manufacturer are straight up horn curves (ie pure tractrix, pure CD, or pure exponential) but have been heavily modified.

Have a listen to a Klipsch Reference premier or jbl m2 monitor, or visit a cinema utilizing either modern Klipsch or jbl cinema speakers, and you’ll see what I mean.

I'm definitely no speaker expert and likely know the least about speaker design than anyone. I don't even want to know. It would probably ruin the whole sport for me.

I get to do quite a bit of back-to-back listening of different designs. I DIY enough different ones and I'm not even really caring about any real superiority tradeoffs to that effect. I still enjoy almost all of them. I'm not out to disprove any type either. I just like distortion free listening and that's really the extent of my criticisms if there are to be any.

Still I wonder, why would the likes of Harman and Klipsch pursue this technology if there is no merit to it? They've certainly covered a broad, if not all the current design types. It's from my recent experiences, one of those newer tricks that seems to do exactly as advertised and noticeably so. It's certainly stupid proofed my setup to a great degree. It's one of the most immersive experiences I have managed from my otherwise paltry, 2 channel setups.

 

[yepimonfire]

I'm definitely no speaker expert and likely know the least about speaker design than anyone. I don't even want to know. It would probably ruin the whole sport for me.

I get to do quite a bit of back-to-back listening of different designs. I DIY enough different ones and I'm not even really caring about any real superiority tradeoffs to that effect. I still enjoy almost all of them. I'm not out to disprove any type either. I just like distortion free listening and that's really the extent of my criticisms if there are to be any.

Still I wonder, why would the likes of Harman and Klipsch pursue this technology if there is no merit to it? They've certainly covered a broad, if not all the current design types. It's from my recent experiences, one of those newer tricks that seems to do exactly as advertised and noticeably so. It's certainly stupid proofed my setup to a great degree. It's one of the most immersive experiences I have managed from my otherwise paltry, 2 channel setups.

Klipsch originally designed horn speakers for extreme efficiency and ultra low distortion, especially modulation distortion, which is rarely talked about and often overlooked. Everyone talks about harmonic distortion, but most harmonic distortion generated by speakers is low order, and therefore not as audible or even disagreeable. IM distortion on the other hand is easily audible even at low levels, and is extremely fatiguing. Almost all 2 way speakers suffer from at least some IM distortion. If you want to get an idea of how bad it sounds, use REWs dual tone generator, pick the custom box, and play a 4:1 tone at the f3 of the speaker and about 100hz below the xover point. The larger the difference, the worse it becomes. A driver reproducing 50hz and 2500hz will likely have IM distortion levels as high as 10%, even at lower levels such as 80dB. Horns don’t suffer this problem, since excursion is minimal in horn loaded drivers. Even if the bass driver is direct radiating and the tweeter is a horn, a lower xover point afforded by a horn minimizes the distance between the lowest and highest frequencies the woofer has to reproduce.

Eventually, a focus toward constant directivity in their pro cinema series trickled down to their consumer lines, ironically their lower end icon series at first, and now into the reference series.

Distortion can be from more than just linear distortion and harmonic distortion, and the room reflections can cause their own distortion of the frequency response.

Floyd Toole has shown that listeners prefer speakers that have a uniform off axis response, if the dispersion is controlled and directivity constant, the sound bouncing off the walls will have the same timbre as the direct sound, and people sitting off axis will similarly receive the same timbre as those on axis. The only room problems you have to deal with now is room modes and excessive reverb/image shift, which can be controlled with modest absorption.


Sent from my iPhone using Tapatalk

 

[killdozzer]

A driver reproducing 50hz and 2500hz will likely have IM distortion levels as high as 10%, even at lower levels such as 80dB.

How much of IM distortion can be resolved with 2way + sub/subs? Your example with 50Hz and 2500Hz obviously should, but I'm guessing this is not the only IM distortion occurring.

Do 2.1 and 2.2 bookshelves systems suffer from not having a regular woofer? Is it genuinely missed when listening to stereo music? Also, if you're going for a 2.1 or 2.2 system with 2way speakers, should you play the mains full range and have IM or opt for some sort of XO that takes some load of the midd drivers?

I think my bookshelves are tuned to 52Hz, which also draws the question of upper harmonics and IM. I tried playing low freq. from a test CD through my speakers and although they should in no way reproduce 25 or 30Hz, you do hear something. I guess these must be the upper harmonics. Will you get IM distortion with them as well and will they be as high as you stated?

 

[MrBoat]

Klipsch originally designed horn speakers for extreme efficiency and ultra low distortion, especially modulation distortion, which is rarely talked about and often overlooked. Everyone talks about harmonic distortion, but most harmonic distortion generated by speakers is low order, and therefore not as audible or even disagreeable. IM distortion on the other hand is easily audible even at low levels, and is extremely fatiguing. Almost all 2 way speakers suffer from at least some IM distortion. If you want to get an idea of how bad it sounds, use REWs dual tone generator, pick the custom box, and play a 4:1 tone at the f3 of the speaker and about 100hz below the xover point. The larger the difference, the worse it becomes. A driver reproducing 50hz and 2500hz will likely have IM distortion levels as high as 10%, even at lower levels such as 80dB. Horns don’t suffer this problem, since excursion is minimal in horn loaded drivers. Even if the bass driver is direct radiating and the tweeter is a horn, a lower xover point afforded by a horn minimizes the distance between the lowest and highest frequencies the woofer has to reproduce.

Eventually, a focus toward constant directivity in their pro cinema series trickled down to their consumer lines, ironically their lower end icon series at first, and now into the reference series.

Distortion can be from more than just linear distortion and harmonic distortion, and the room reflections can cause their own distortion of the frequency response.

Floyd Toole has shown that listeners prefer speakers that have a uniform off axis response, if the dispersion is controlled and directivity constant, the sound bouncing off the walls will have the same timbre as the direct sound, and people sitting off axis will similarly receive the same timbre as those on axis. The only room problems you have to deal with now is room modes and excessive reverb/image shift, which can be controlled with modest absorption.


Sent from my iPhone using Tapatalk

All I can say is, I am thankful I was subjected to portable record players, AM transistor radios, mono car radios, and that awful stereo equipment my parents bought that my older brothers blew the speakers on playing Hendrix too loud when they weren't home when I was younger.

I'm also thankful that an imposing WAF, or even perhaps Social Appreciation Factor (SAF) doesn't have me constantly trying to cheat the notion of there being "no replacement for displacement" with regard to that excursion issue that you mention. That some of these issues can sometimes be overcome with sheer brute force.

 

[shadyJ]

Almost all 2 way speakers suffer from at least some IM distortion. If you want to get an idea of how bad it sounds, use REWs dual tone generator, pick the custom box, and play a 4:1 tone at the f3 of the speaker and about 100hz below the xover point. The larger the difference, the worse it becomes. A driver reproducing 50hz and 2500hz will likely have IM distortion levels as high as 10%, even at lower levels such as 80dB. Horns don’t suffer this problem, since excursion is minimal in horn loaded drivers. Even if the bass driver is direct radiating and the tweeter is a horn, a lower xover point afforded by a horn minimizes the distance between the lowest and highest frequencies the woofer has to reproduce.

What is your source for that 10% figure?

If you want to know more about IM, look at some of the papers by Alex Voishvillo, and also Klippel. Voishvillo and his team have done a lot of research here, and have produced some very comprehensive papers. I think your view of intermodulation distortion is too simplistic. Horns do not make a speaker immune to intermodulation distortion, and non-horn speakers can still have excellent linearity and very low distortion. Bandwidth makes a difference here, and horn-loaded drivers can be given a greater bandwidth than non-horn loaded drivers.

 

[yepimonfire]

How much of IM distortion can be resolved with 2way + sub/subs? Your example with 50Hz and 2500Hz obviously should, but I'm guessing this is not the only IM distortion occurring.

Do 2.1 and 2.2 bookshelves systems suffer from not having a regular woofer? Is it genuinely missed when listening to stereo music? Also, if you're going for a 2.1 or 2.2 system with 2way speakers, should you play the mains full range and have IM or opt for some sort of XO that takes some load of the midd drivers?

I think my bookshelves are tuned to 52Hz, which also draws the question of upper harmonics and IM. I tried playing low freq. from a test CD through my speakers and although they should in no way reproduce 25 or 30Hz, you do hear something. I guess these must be the upper harmonics. Will you get IM distortion with them as well and will they be as high as you stated?

What is your source for that 10% figure?

If you want to know more about IM, look at some of the papers by Alex Voishvillo, and also Klippel. Voishvillo and his team have done a lot of research here, and have produced some very comprehensive papers. I think your view of intermodulation distortion is too simplistic. Horns do not make a speaker immune to intermodulation distortion, and non-horn speakers can still have excellent linearity and very low distortion. Bandwidth makes a difference here, and horn-loaded drivers can be given a greater bandwidth than non-horn loaded drivers.

I was in no way implying a horn is the only solution, it’s just a very good solution in two way speakers.

The “ideal” loudspeaker would be an active 4 way design, with each driver covering a maximum of ~3 octaves, dispersion of each driver can be matched by using drivers that are smaller than the wavelengths of the xo frequency. A 4 way speaker could theoretically avoid IM distortion completely, and ensure each driver is crossed over well before it starts beaming. For 99% of us, this is impractical and very expensive. Who wants to use 8 amps just for a stereo pair? Outside of that, a three way speaker is likely to show even less IM and beaming problems (at least in the midrange), and adding a sub would reduce distortion even further.

A horn can extend the bandwidth of a tweeter at the lower end, often by an entire octave, since excursion is well controlled due to impedance matching, IM distortion is substantially less. Using rews dual tone generator, playing an 1800hz tone atop a 20khz tone at a 4:1 ratio at 85dB @1m (didn’t want to blow the tweeter) I measured .883% IM distortion. At 50hz and 1500hz for the woofer, again at a 4:1 ratio, but this time at 95dB @1m, I measured 1.3% IM distortion. A pair of Polk Audio speakers I have appear to have a crossover point (based on close mic measurement of the individual drivers) of 3khz. Playing a 50hz tone and a 2800hz tone, IM distortion reached 6% @95dB. Testing the tweeter, a 3.2khz tone and a 20khz tone at 85dB produced 4% IM distortion, which was clearly audible as a nasty warbling sound. The extra octave the woofer has to cover is the problem here, since running a 50hz and 1500hz tone showed 1.8% IMD (I suspect the doped plastic woofer starts flexing a bit more at similar spl, since it’s not as stiff as the aluminum/ceramic.)

As far as dispersion goes (especially at the very top of the spectrum), you’re just not going to get the kind of uniform controlled response a waveguide/horn offers with direct radiating tweeters. Sure, you can achieve an off axis response that doesn’t have issues like lobing, but you can’t get around the laws a physics, a 1” dome tweeter is going to have a very wide dispersion at 2500hz, and a progressively narrowing response above ~8khz. A speaker like this will sound different in every room, and different with each variation of placement, because the angle of dispersion progressively gets smaller and smaller as the frequency goes up. A speaker like this might measure flat in an anechoic chamber, but as soon as it starts spewing off axis sound at varying intensity at the walls, that response will change.

A waveguide/horn can achieve pattern control up to all but the very highest frequencies. The sound reflecting off the walls will be similar in timbre, and (generally for better) be lower in intensity. There are many non waveguide speakers that are designed for extremely wide dispersion (without any form of pattern control), and these have their place. Some people like the way those sound, and there isn’t anything inherently wrong with them, so long as you can get good placement dialed in and your room jives well with them acoustically, but if you want to minimize the impact of reflections on sound accuracy and timbre, provide a uniform response to all seats, on and off axis, and still maintain tight imaging (which is somewhat diffuse with omnidirectional speakers), waveguides/horns are the way to go.



I’ve actually read a bunch of Klippel’s stuff, I’ll have to look at Voishvillo.


Sent from my iPhone using Tapatalk

 

[yepimonfire]

How much of IM distortion can be resolved with 2way + sub/subs? Your example with 50Hz and 2500Hz obviously should, but I'm guessing this is not the only IM distortion occurring.

Do 2.1 and 2.2 bookshelves systems suffer from not having a regular woofer? Is it genuinely missed when listening to stereo music? Also, if you're going for a 2.1 or 2.2 system with 2way speakers, should you play the mains full range and have IM or opt for some sort of XO that takes some load of the midd drivers?

I think my bookshelves are tuned to 52Hz, which also draws the question of upper harmonics and IM. I tried playing low freq. from a test CD through my speakers and although they should in no way reproduce 25 or 30Hz, you do hear something. I guess these must be the upper harmonics. Will you get IM distortion with them as well and will they be as high as you stated?

A sub definitely lowers IMD, especially in ported speakers since excursion is ridiculous below the tuning frequency. IM distortion increases with excursion, so if your speaker is tuned to 52hz, the cone movement will be minimal at this frequency, and IM distortion minimal as well. You really want to aim for no more than 3-4 octaves per driver, so for example, from 88hz-1420hz is 4 full octaves. If a speaker were crossed over at 3khz, that’s 5 1/4 octaves. With a ported speaker, I’d say you could ignore that octave, since cone movement is at a minimum, so if a speaker has a port tuning of say, 60hz, from about 50hz-70hz the cone excursion will be significantly reduced, so you’d likely be able to run it from 60hz-1420hz and not have any more distortion than if it were crossed over at 80hz.

As for getting sound at 25&30hz, I’ve noticed some speakers, even ported speakers, tend to reproduce undistored bass slightly below the tuning, I can get about 40hz out of my mains, about 6dB down, which are -3dB at 47hz.

Could also be the room. Room gain doesn’t apply as much to ported speakers, but I’ve seen some rooms impart an extra half an octave. A friend bought a pair of KB-15s from me, which measure -3dB at 60hz. His room has a carpet covered concrete floor, and the walls surrounding his room are brick on the other side outdoors. I measured his room for him for eq purposes and he had a natural -3dB response of 38hz, measured again several times to confirm it. I have a pair of speakers attached to my ceiling for atmos. They have a -3dB spec of 90hz, my ceiling is my room is 20’x12’x8’, giving me calculated room modes of 28hz along the length, 47hz along the width, and 70hz along the height. Since the speakers are about 7” from the ceiling, the boost at 70hz extends the response down to 70hz -3dB.

If the manufacturer is being honest, the -3dB spec should be anechoic. Depending on the -3dB point of your speakers and the dimensions/material of your room, you could get an extension. For example, if you have a pair of tower speakers that have a LF response down to 30hz, and you put them in a 25’x20’, a foot from the short wall, it’s very possible you could get an in room response down to 25hz, since the room modes would be 28hz and 25hz.


Sent from my iPhone using Tapatalk

 

[Johnny2Bad]

" ...
As for getting sound at 25&30hz, I’ve noticed some speakers, even ported speakers, tend to reproduce undistored bass slightly below the tuning, I can get about 40hz out of my mains, about 6dB down, which are -3dB at 47hz.
..."

With ported speakers, the port resonance frequency is the point of minimal cone excursion (due to mechanical impedance) so distortion should be low at that point.

You can screw that up, but in that case the speaker is not properly designed, so we shouldn't find it in the well-regarded loudspeakers someone into HiFi would be expected to choose.

 

[killdozzer]

@yepimonfire thanks for your answer. I'm actually thinking about pairing LS50 with sealed box sub SVS SB Ultra13 (mainly music). That's why I asked. Here, where I live, you don't really get the easy trial period, so I'm trying to get as much info as possible, without having them in my home for a test drive (don't worry though, If I decide to go for them, I'll find a way to hear them).

White Paper actually says the port is tuned to 55 and not 52.


But I'm thinking about alleviating them of some bass.

 

[yepimonfire]

@yepimonfire thanks for your answer. I'm actually thinking about pairing LS50 with sealed box sub SVS SB Ultra13 (mainly music). That's why I asked. Here, where I live, you don't really get the easy trial period, so I'm trying to get as much info as possible, without having them in my home for a test drive (don't worry though, If I decide to go for them, I'll find a way to hear them).

White Paper actually says the port is tuned to 55 and not 52.
View attachment 22791

But I'm thinking about alleviating them of some bass.

Coaxial drivers in theory have high levels of IM distortion, however, with the amount of engineering and research kef has done, they may have found a solution to it. Never heard them and never seen measurements so I can’t comment.


Sent from my iPhone using Tapatalk

 

[shkumar4963]

@yepimonfire thanks for a very informative and detailed discussion.

I have LS50s. They have a very small woofer. While they call it 5.5 inch, if you actually measure it, the cone area is more like 4.5 inch (with the center cut out for a tweeter) . Based on your discussion, the small size should help in managing directivity at the cross over frequency (1.5k??). And waveguide loading allows tweeter to perform well at that frequency as well.

But how does that small woofer produces sound till about 60-70 Hz. The port is tuned to 52 Hz. What is the physics behind it. Why other manufacturers need 6 inch or larger woofers while kef ls50 can get away with 4.5 inch?

Also what could be a possible reason for relative high levels of distortion in these speakers.

http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:nrc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153

Sent from my SAMSUNG-SM-G930A using Tapatalk

 

[everettT]

What is your source for that 10% figure?

If you want to know more about IM, look at some of the papers by Alex Voishvillo, and also Klippel. Voishvillo and his team have done a lot of research here, and have produced some very comprehensive papers. I think your view of intermodulation distortion is too simplistic. Horns do not make a speaker immune to intermodulation distortion, and non-horn speakers can still have excellent linearity and very low distortion. Bandwidth makes a difference here, and horn-loaded drivers can be given a greater bandwidth than non-horn loaded drivers.

Horns baby, Horns, that's the only solution

https://workforcenow.adp.com/jobs/apply/posting.html?client=VOXX&ccId=19000101_000001&type=MP&lang=en_US

 

[Steve81]

I have LS50s. They have a very small woofer. While they call it 5.5 inch, if you actually measure it, the cone area is more like 4.5 inch (with the center cut out for a tweeter). Based on your discussion, the small size should help in managing directivity at the cross over frequency (1.5k??). And waveguide loading allows tweeter to perform well at that frequency as well.

The LS50s are crossed over at 2.2kHz. This is facilitated by using a concentric layout to manage any potential directivity mismatch, vs crossing over low enough that the mid/woof isn't beaming.

But how does that small woofer produces sound till about 60-70 Hz. The port is tuned to 52 Hz. What is the physics behind it. Why other manufacturers need 6 inch or larger woofers while kef ls50 can get away with 4.5 inch?

Also what could be a possible reason for relative high levels of distortion in these speakers.

The low end response is mostly just a matter of engineering of the driver (T/S parameters) combined with the enclosure. However, the relatively limited volume displacement of the driver means that output is necessarily limited, hence the high distortion figures you found. The LS50s are really designed as a near field monitor, not a speaker that's supposed to fill a large living room with reference level sound. That much should be apparent in how it's advertised:


The SoundStage measurements plot distortion at 90dB, 2 meters distance in an anechoic chamber. That's pretty loud for how the speaker is intended to be used, hence the high distortion on the low end.

 

[killdozzer]

But how does that small woofer produces sound till about 60-70 Hz. The port is tuned to 52 Hz. What is the physics behind it. Why other manufacturers need 6 inch or larger woofers while kef ls50 can get away with 4.5 inch?

Low sensitivity, heavy membrane, strong motor, picture perfect CFD port.

However, not all manufacturers strive for this type of speaker. Technics tried to follow these footsteps.

 

[shkumar4963]

@Steve81. Yes. I had heard that the cross over freq. Is 2.2 k. But look at the woofer response below. At 2.2 K it's response is at -15dB.

Sent from my SAMSUNG-SM-G930A using Tapatalk

 

[shkumar4963]

@yepimonfire thanks for a very informative and detailed discussion.

I have LS50s. They have a very small woofer. While they call it 5.5 inch, if you actually measure it, the cone area is more like 4.5 inch (with the center cut out for a tweeter) . Based on your discussion, the small size should help in managing directivity at the cross over frequency (1.5k??). And waveguide loading allows tweeter to perform well at that frequency as well.

But how does that small woofer produces sound till about 60-70 Hz. The port is tuned to 52 Hz. What is the physics behind it. Why other manufacturers need 6 inch or larger woofers while kef ls50 can get away with 4.5 inch?

Also what could be a possible reason for relative high levels of distortion in these speakers.

http://www.soundstagenetwork.com/index.php?option=com_content&view=article&id=941:nrc-measurements-kef-ls50-loudspeakers&catid=77:loudspeaker-measurements&Itemid=153

Sent from my SAMSUNG-SM-G930A using Tapatalk

Actually the woofer dia in less than 4 inch (cone size excluding surround)

Sent from my SAMSUNG-SM-G930A using Tapatalk

 

[Steve81]

@Steve81. Yes. I had heard that the cross over freq. Is 2.2 k. But look at the woofer response below. At 2.2 K it's response is at -15dB.

Given that the graph in question is pulled from a section on the low frequency alignment of the LS50s, it's not necessarily representative of what's happening on the top end of the woofer's bandwidth. Looking at the measurements of the LS50's raw drive units, a 1.5kHz XO would be a tall order for the tweeter, with no upside to speak of. 2.2kHz is what appears in all the literature I've seen on the LS50s (brochure, manual, product information sheet, website, reviews, etc.), and makes a heck of a lot more sense design-wise.

 

[DubPlate]

I was mostly referring to two way designs. Your run of the mill speakers with 5.25” or 6.5” woofers and a 1” dome tweeter with a 2500hz xover.

If you follow the rule that a driver should have a ka=2 at the crossover frequency, an effective piston diameter of 5” sets this point at 1706hz, a 6” diameter 1312hz. Most soft dome tweeters would be destroyed by such a low crossover point, ignoring the massive distortion from them that low as well. With a two way design, the only way to manage such a low xover point is a waveguide or horn, which controls the excursion of the driver within it operating range. A titanium dome tweeter attached to a horn/waveguide in, for example, the RP-150m speakers I own, are crossed over at 1500hz, perfectly matching the directivity of the drivers. Klipsch uses the exact same tweeter for all of their reference speakers, but each one is crossed over differently depending on the woofer diameter and horn size. An rb 10 with a 4” woofer has an xover of 2200hz, while the 150m using a 5” driver has an xover of 1500hz, matching the directivity of both woofers.

You are correct that a horn by itself cannot reduce beaming, but horn/waveguide mounted tweeters are usually fit with a phase plug in order to direct the highest frequencies into the horn/waveguide for dispersion control.

I 100% agree with what you have said, that good crossover design and matching of drivers can provide perfect off axis response. As previously stated, I was only referring to typical direct radiating two way designs. 3 way or 4 way speakers can be better matched with good crossover design for excellent off axis dispersion. With a two way direct radiating design, a “proper crossover point” matching the drivers directivity doesn’t exist in designs utilizing 5”+ woofers. The woofer will start to beam as the wavelength of the frequency approaches the diameter of the woofer, and even if some magic tweeter could somehow survive a low xover point, a 1” tweeter will start beaming at about 8500hz. 9/10 polar plots I’ve seen of common two way designs display problems at the 2500 xover frequency vertically off axis, and a rapid drop off off axis beginning at around 8khz because of this.

The solution is (like you mentioned) a multi way design with good driver/crossover matching based on directivity or a tweeter mounted to a waveguide/horn that allows a lower crossover frequency, using a phase plug to direct the highest frequencies into the waveguide for controlled dispersion.

So I agree, you don’t necessarily need a waveguide or horn to get wide off axis dispersion, but it is certainly one way of achieving it, and really the only way to achieve it in two way designs.

While I don’t agree with everything he says (especially his aversion to line arrays in live sound) I think this guy makes a good case as to why he insists on 4 way designs (or 3 way with a sub) http://education.lenardaudio.com/en/05_speakers.html




Sent from my iPhone using Tapatalk

Would having two or more woofers mitigate beaming in a two way design, and perhaps allow for a higher crossover?

 

[yepimonfire]

Would having two or more woofers mitigate beaming in a two way design, and perhaps allow for a higher crossover?

Not exactly, in a three way design, a large woofer for the bass and low midrange, and a smaller woofer for the midrange/lower treble, would do so, at least for the midrange. The problem one runs into is that something like a 4” midrange driver will still begin to beam at about 2500hz. A 1” dome tweeter will still beam at 8khz. If you were to design a 4 way speaker with a 12” bass driver, an 5” upperbass/lower mid driver, a 2” dome midrange, and a 1/2” tweeter for the harmonics, using xover points of 120hz, 1700hz, and 4khz, directivity would be matched fairly well. A 1/2” tweeter wouldn’t start becoming directional until about 17khz. The issue of varying directivity is still a problem though. Each driver will be very wide at the low end of their operating range, and get slightly more narrow as frequency increases.

This is why I personally think waveguides/horns are a good solution instead. You can achieve matched directivity easily with a 2 way or three way design, since the horn allows a much lower xover than if the tweeter were direct radiating, and the waveguide will control the directivity across most of the bandwidth. A 5.25” woofer can be run up to about 1800hz before beaming becomes an issue. A 1” titanium driver can be crossed over as low as 1000hz if the horn is large enough. A three way system with midrange and high frequency horns gives even more directivity control of the lower midrange, down to about 500hz.

The problems associated with timbre and imaging from uneven dispersion and room reflections of that uneven response occur mostly above 1khz according to Tooles findings. Considering this, a horn that controls directivity above 1khz, should solve these problems. It doesn’t matter if the dispersion is narrow or wide, as long as the directivity is uniform up to at least 10khz. Wider dispersion allows more reflections, narrower allows less, Geddes suggests narrow dispersion speakers (60 degrees comical) to greatly reduce room reflections. Toole suggests wider dispersion speakers. I personally think a middle ground is best. +-45 degrees seems just about perfect for most home audio setups. All seats fall into the pattern, and most room reflections fall on the outside edge of it, reducing their level compared to direct sound, but ensuring they’re of the same timbre as the direct.


Sent from my iPhone using Tapatalk

 

[everettT]

So explain how all 4” drivers beam at 2500hz.

Not exactly, in a three way design, a large woofer for the bass and low midrange, and a smaller woofer for the midrange/lower treble, would do so, at least for the midrange. The problem one runs into is that something like a 4” midrange driver will still begin to beam at about 2500hz. A 1” dome tweeter will still beam at 8khz. If you were to design a 4 way speaker with a 12” bass driver, an 5” upperbass/lower mid driver, a 2” dome midrange, and a 1/2” tweeter for the harmonics, using xover points of 120hz, 1700hz, and 4khz, directivity would be matched fairly well. A 1/2” tweeter wouldn’t start becoming directional until about 17khz. The issue of varying directivity is still a problem though. Each driver will be very wide at the low end of their operating range, and get slightly more narrow as frequency increases.

This is why I personally think waveguides/horns are a good solution instead. You can achieve matched directivity easily with a 2 way or three way design, since the horn allows a much lower xover than if the tweeter were direct radiating, and the waveguide will control the directivity across most of the bandwidth. A 5.25” woofer can be run up to about 1800hz before beaming becomes an issue. A 1” titanium driver can be crossed over as low as 1000hz if the horn is large enough. A three way system with midrange and high frequency horns gives even more directivity control of the lower midrange, down to about 500hz.

The problems associated with timbre and imaging from uneven dispersion and room reflections of that uneven response occur mostly above 1khz according to Tooles findings. Considering this, a horn that controls directivity above 1khz, should solve these problems. It doesn’t matter if the dispersion is narrow or wide, as long as the directivity is uniform up to at least 10khz. Wider dispersion allows more reflections, narrower allows less, Geddes suggests narrow dispersion speakers (60 degrees comical) to greatly reduce room reflections. Toole suggests wider dispersion speakers. I personally think a middle ground is best. +-45 degrees seems just about perfect for most home audio setups. All seats fall into the pattern, and most room reflections fall on the outside edge of it, reducing their level compared to direct sound, but ensuring they’re of the same timbre as the direct.


Sent from my iPhone using Tapatalk

1. explai