I took a quick look because I thought it would be funny if you were right, but looking at the 12a it seems the crossover is around 700hz.
I really doubt a 1" CD would be happy at 700hz. I'm pretty sure the abbey would be crossed over in the range of 1.2khz which makes it "unlivable" to Dr. Geddes
You might want to look more closely at Earls polar response app, and what each line means. 😉
Red and Black do not represent the individual responses of the woofer and tweeter, which I suspect is what you're thinking. (Yes its a bit confusing that they are not labelled, I fell for the same thing myself)
From memory Red is power response, Black is directivity index on the chosen axis, and the colour line is the frequency response on the chosen axis. All three measurements are with both drivers active - there are no individual driver measurements in that data that would allow you to determine the crossover frequency.
The Summa is about 800 Hz, the Abbey about 900 Hz, the Nathan about 1100 Hz and the Harper about 1500 Hz. No 1" tweeter is going to get anywhere near those points at the efficiencies and SPL capabilities that I obtain. Not even close.
Ah, I guess I'm not looking at the charts correctly then, I was thinking that they were driver axis response where we'd be able to guess the crossover point by looking at their roll off. I guess I'll have to get that polar response app and take a closer look.
It would certainly be helpful if the graph was labelled... 😉 Although it's possible to deduce what each line is when playing with the "live" version of the app, its not at all clear from just looking at the static screenshot. I bet the two of us are not the only people to be fooled by Earls graph before. 😀
Why not the Bohlender Graebener Neo3-PDRW Planar Tweeter for most speakers that should be enough performance or for the ultimate we should use Raal ribbon tweeters then??
I love my Neo 3 PDRs but I'm not under the impression that they're the best. Certainly worth their $65 price tag though, at least imo.
Thanks for all this info - much appreciated.
This very thorough 1962 Electronics World article is where I learned most about how those lenses operated: "The Acoustical Lens"
Btw during my childhood, one of my parent's friends owned a Sansui/Garrard SuperZero/JBL Century L100 quad system & he kept all the catalogs for JBL speakers, including one for separate drivers. I read them every time we visited and those funky lenses always fascinated me).
Look up Bart Locanthi and his early paper on these. Not someone I would accuse of being short on science.
David S.
I'm not suggesting there is no science behind diffraction slots in general, obviously they're used for focusing light and electromagnetic radiation, however in both those applications the relative bandwidth of operation is an extremely narrow percentage of the operating frequency compared to the several octaves a tweeter is expected to cover.
What I am suggesting is that the practical implementations of an "acoustic lens" for tweeters back in the day was rather wishful thinking. Show me a cone tweeter of the era with severe beaming without such a lens, then show me that placing the lens in front of it improves the dispersion at high frequencies whilst simultaneously providing a smooth frequency response, clean diffraction free impulse response and a smoothly monotonically falling off axis polar response.
There is no doubt that some increased dispersal of high frequencies will take place, but the drawback is that performance of the lens will vary greatly with frequency due to the fixed size slots, will introduce spurious diffraction and reflection that will cause narrowband ripples in the frequency response and make it impossible to get a clean impulse response.
If you look at the evolution of constant directivity wave-guides, diffraction slots were once considered a necessity of the design, but many years later it was realised that the diffraction from the sharp edge was a serious and under appreciated problem for sound quality and designs evolved to avoid it.
With modern speakers (at least some of them) going to great pains to eliminate sources of diffraction and reflection from the cabinet and mounting, any kind of diffractive grating in front of a tweeter is a big backwards step.
If you're talking about 50 years ago when high quality wide dispersion tweeters were thin on the ground, it might have been a useful band-aid to increase high frequency dispersion - after all the same drivers probably already had a rather ragged treble response to begin with, and you can bet no attention was paid to diffraction in the rest of the speaker and cabinet design.
But with today's modern tweeters which can be both very smooth and wide in dispersion, a diffraction based "acoustic lens" is nowhere near state of the art, so I can see no reason at all to use one.
I did my MS thesis on slant plate acoustic lenses. This was also an AES paper. There is some good theory behind them, but implimentation was always a problem. The plates would vibrate, etc. things the theory did not address. The real measured responsese were never as good as one hoped.
The use of foam in a waveguide was a direct result of an attempt to use the foam as a "lense". It really is a good material in this regard. But I fould that the "lense" effect was less pronounced than the "sonic" effect so I opted to optimize the sonic effect.
The use of foam in a waveguide was a direct result of an attempt to use the foam as a "lense". It really is a good material in this regard. But I fould that the "lense" effect was less pronounced than the "sonic" effect so I opted to optimize the sonic effect.
As long as the horn or waveguide makes it worth going there, it can.
As we do 😉 ..it's under 'Comments' on the default plot in the app.
re seebert "a full set of (as an example) pre or power amplifier specs will not give anyone reading them a good indication of how the unit will perform in their own real world."
That is correct when the specifications are either incomplete or do not represent the way the equipment will be used in the real world. For example, when it comes to power ampilifers (which don't actually amplify power but voltage and have a final stage capable of delivering a lot of current) the standard method inherited from the 1920s and 1930s uses a resistor as a load and measures frequency response at 1 watt output. Real world speaker loads are neither purely resistive nor are they passive. Their reactive components and their back emf can adversely affect some amplifiers and not others. The concept of slewing rate and transient intermodulation distortion stems from the fact that at higher power ouput levels the bandwidth is reduced and FR can be much poorer. If FR were specified at full output and all levels below, TIM and slew rate would be unnecessarily redundant.
The way in which loudspeakers radiate sound into solid space and how that sound interacts with a real room is rarely if ever more than hinted at with lip service. No data is generally published for FR at all solid angles in all planes even at one SPL level let alone at multiple SPLs. There are no commercial loudspeakers which are designed with engineered provisions to control how speakers interact with real world rooms by making provisions to compensate for variables of different rooms so that the end result will be similar performance no matter what room the speaker is installed in. The use of tweeter and midrange controls and equalizers is entirely unsatisfactory because they affect both the direct and reflected sound fields simultaneously, that is they can't be altered independently of one another.
The current design philosophy seems to be to defeat the acoustics of the listening room one way or another through restricting high frequency dispersion, placing speakers far from reflective surfaces, and using sound absorbing material. Short of creating and anechoic chamber, the world's worst place to listen to music, the battle against listening room acoustics is one the speaker designer cannot win.
Last edited:
A multidirectional array of tweeters has the following advantages over a single tweeter
1) uniform dispersion over a very wide angle is possible by aiming them in different directions. This is much closer to the way musical instruments propagate sound. Just walk around a piano at a piano bar or a street musician (buskar) and notice how little the tonality changes. This can never be true of single tweeter designs. Multiple tweeters also create multiple comb filters where the the number of peaks and dips are so many and change in what frequencies they occur at so much with slight changes in orientation that for all practical intents and purposes they don't exist. When you are on the direct field of one driver, you're at an angle to all the others. Therefore the total power output versus frequency comes much closer to the FR of the shortest path length for any direction between you and the array. The term "on axis" loses its meaning.
2) An array of tweeters cuts the power requirement into small fractions of what a single tweeter must handle. Efficiency of 90 db for each driver is typical. Dynamic compression is sharply reduced or eliminated. Tweeters rated at 20 to 30 watts can each operate at only a fraction of a watt producing almost no temperature rise in their voice coils.
3,4) Tweeters should be crossed over at a high frequency and their passband in the audible range restricted to no more than 2 to 2 1/2 octaves. That means a crossover no lower than 3500 to 5 khz. A tweeter with a lower crossover must handle too much power at frequencies it is not optimized for. It is nearly impossible to get a 2 way design to cover the entire audible spectrum of 10 octaves. Invariably there is a tradeoff somewhere. Usually it's in the deep bass but often there is an abrupt change in radiating angle at the crossover frequency. Adding a subwoofer turns it into a 3 way system, usually meaning the end user must experiment to get the best possible match up. The 4 way design may seem complex but it allows each driver to cover a range for which it is best suited. Loudspeaker drivers are resonant devices with a useful range of about 2 to 2 1/2 octaves, rarely much more. The only exceptions are headphones but they can only prodice a very small amount of soundpower.
When well executed, the 4 way is probably the system with the fewest crossover points necessary for high fidelity. Use of multiple drivers in each range allows greater design flexibility in angular radiation characteristics and increased power handling.
An inexpensive array of multiple poly dome tweeters easily outperforms the most expensive single tweeter in all significant aspects except one. If you want a tweeter to beam high frequencies at you with few early reflections so that you can listen in a sweet spot and hear some special imaging that only such an arrangement can produce, a multidirectional array of tweeters is not the answer.
I don't think our speakers are intended to be stand-ins for instruments, but instead working together to form an image. When listening to music, if I look directly at my speakers they appear to be completely silent. Even when the stereo image places a performer around the location of one of the speakers, it still doesn't sound like it is coming from the speaker. I have only been able to achieve this since paying closer attention to diffraction and reflections, both of which I have improved with narrowed directivity.
Tonality is one of the four basic elements of music. Imaging is not. For me, if a sound reproducing system cannot reproduce the beautiful tonality of acoustic musical instruments closely to the way they sound in the real world, that system is not high fidelity and whatever other atributes it has are not worth a whole lot.It is the tonality of the instruments which is a crucial element in the appreciation and enjoyment of music. Recording systems invariably come in a poor second at best.
- Status
- Not open for further replies.