Go listens to all the speakers within your budget , make a conclusion base on what you think sound best overall. After listening to audios for more than 30 years, starting with my obcession with a fantastic full range JVC ceramic speaker to now using modify KEF and a Yamaha surround sound AV rec. Yes I listened to many top high ends and honestly is underwhelmed mostly by the supposely incredibles speaker /amp review tales...
But if you want something to impress the diehard audiophiles go ahead its your ultimate satisfaction that matters.
Lastly I find ribbons impressive even though it don,t measures up to certain expectation . I have a set of low cost planar ribbons in my car now and i find the subjective sound incredible compare with a few domes. After swaping around and calibrating numerous times I kept the planar ribbon i the car. But I did not go further to replace the KEF as to me it sounds very good as it is after heavily modifying the crossover to eliminate the metallic midhigh sound and high pass it above 250hz to reduce substantial intermodulation distortion due to cone excursion. A lot of audiophile will tell you KEF don,t sound good but its a real good speaker let down by crossover and working it beyond its design space.
But if you want something to impress the diehard audiophiles go ahead its your ultimate satisfaction that matters.
Lastly I find ribbons impressive even though it don,t measures up to certain expectation . I have a set of low cost planar ribbons in my car now and i find the subjective sound incredible compare with a few domes. After swaping around and calibrating numerous times I kept the planar ribbon i the car. But I did not go further to replace the KEF as to me it sounds very good as it is after heavily modifying the crossover to eliminate the metallic midhigh sound and high pass it above 250hz to reduce substantial intermodulation distortion due to cone excursion. A lot of audiophile will tell you KEF don,t sound good but its a real good speaker let down by crossover and working it beyond its design space.
Interesting about the JVC ceramics, I tried to get a small pair, as I have the JVC wood cones, and was curious about the ceramics.
4" rockwool paneling on all surfaces would do it lower (in freq.) than what controlled directivity could achieve in-room (..most normal-sized rooms).
Of course controlled/limited directivity is much higher in pressure loss an octave + above the waveguide's "mouth" freq..
I personally think both are counter productive (except in cases where the loudspeaker needs to be very close to the wall).
I really like most ribbons, but it's NOT because of the vertical directivity in the top-octave reducing the pressure from vertical reflections.
Instead it's because of the lack of any real suspension and the lower mass for a given Sd.
Most dome's sound "broken" by comparison: lacking the transient clarity without "artifice" and particularly the decay of a good ribbon tweeter. It can be a much more natural/realistic sound with a good ribbon tweeter, though I've heard several designs with ribbons that still "get it wrong".
Last edited:
No, it only works for nearfield. Controlling directivity without screwing up the frequency response outside the controlled beam is not just difficult, it is neigh well impossible.
And this screwed up FR gets reflected and impacts the perceived timbre irreparably for any other listening use than near field.
I built a two way system with the Aurum Cantus G2 ribbons and I'm extremely pleased with the result, so much so that I've recently bought another pair of G2 for a new larger 3 way design - although this time in rectangular face plate form as the round face plate version I bought ~15 years ago is no longer available and the round G2Si (which looks very similar) doesn't have the power handling or sensitivity I need as it is a scaled down version of the G2 designed for small bookshelf speakers.
Like most smaller ribbons (these are about 50mm tall visible ribbon area) the key to success is don't cross them over too low and get the crossover right.
High order filters are needed and midrange drivers that can cross relatively high, cleanly. (at least 3Khz) If you don't like high order filters or higher crossover frequencies, don't use a ribbon, at least not a small one like this, as you will either damage it or be disapointed.
I'm using the AC G2 with a 3Khz 4th order Linkwitz-Reily passive crossover, and I would not go any lower in frequency than this - that's about their comfortable limit. I've also used them earlier in the design phase with a 3rd order 4Khz crossover. I would not use less than 3rd order.
If your midrange driver can perform well at higher frequencies my gut feeling is that 4Khz 4th order is optimal for this model, especially if you want to build a larger speaker that is operating at a higher SPL level, (raising the crossover frequency even from 3Khz to 4Khz reduces the power dissipation on the tweeter greatly) and that is the target for my 3 way design which is slowly coming together.
Ribbon tweeters can be a little delicate - they don't take kindly to metal filings getting into the exposed magnetic gap (keep the front sealed with cellotape while working on/building the speakers or transporting them!) and are easily over driven if the crossover is too low or too shallow.
You do get a little bit of visual warning though - the ribbon will start to bow slightly when it heats and expands, if you see this you're overdriving it. If they overheat too much the ribbons will stretch and not return to their original corrugated form, becoming a little bit slack. They will still work but their response will change slightly.
I did manage to damage one ribbon very early on when I was doing quick testing with an unsuitable crossover however after replacing the ribbon foil they have been in use for 15 years now without incident, so they have stood the test of time. They still sound great, better than ever in fact as a couple of years ago I did a ground up redesign of the crossovers using Vituixcad.
A common theme I hear among ribbon sceptics is that they can't/don't sound good because they have higher THD than similar domes, and suggest that somehow this higher THD is actually pleasing to the ear and account for preferences towards ribbons among some people. Rubbish.
While it's true they do have slightly higher THD than similar performing domes, this is mostly at the top end of the treble so most of the harmonics generated fall outside the audible hearing range anyway. (What are the second and third harmonics of 10Khz ?) So I don't buy that argument at all.
Non-linear distortion has a very strong threshold effect - if distortion is below a certain threshold (or is masked) it is inaudible. A driver with 0.2% of low order harmonic distortion won't sound worse than one with 0.1% distortion as both are too low to be audible.
Comparing two tweeters and criticising one because it has slightly higher non-linear distortion when both are a long way below the threshold of audibility in music is folly, especially if the lower distortion driver is deficient in some other metric, such as directivity control or flatness of frequency response. If you optimise the wrong parameter at the expense of other more important parameters you won't get the best result.
As a result I don't take much stock in the opinions of people who obsess over minute level distortion measurements of tweeters. (I won't name names but they know who they are 🙂 )
For many years I was also a believer in very low levels of non-linear distortion being very important but work from people like Earl Geddes on the audibility of non-linear distortion made me start to question that belief and I now believe that linear distortions (especially amplitude and time) are much more important. That's not to say non-linear distortion doesn't matter at all, just that obsessing over getting super low values is unnecessary especially when you have to sacrifice elsewhere.
I only have experience with waveguide loaded ribbon tweeters, so when I talk about ribbon tweeters in this post some of their characteristics come from the ribbon itself and would apply to all ribbons, and some are a result of the directivity pattern imparted by the waveguide which would not necessarily apply to a large open ribbon.
There is just something about the sound of a ribbon tweeter that is hard to quantify or explain. It's one of those "you have to hear it to understand" things. I hesitate to use subjective descriptions but I would say "naturalness", "transparency" and freedom from audible high Q resonances that lead to harshness are the main points. Waterfall plots on ribbon tweeters are usually exceptionally clean.
With a good ribbon tweeter you can get crispness without harshness. They can be at the same time detailed yet laid back and easy to listen to without listener fatigue. Sibilance is not usually an issue with a ribbon tweeter (unless you mess up the crossover) as sibilance is typically a result of high Q cone breakup in the 6-8Khz region which is common on soft dome tweeters.
Some ribbon tweeters (including the G2) do have a slight emphasis above 10Khz due to a cavity resonance behind the ribbon as well as the directivity effects of the waveguide - if left uncorrected (and I suspect many designs don't correct for this) this can lead to the treble sounding a little toppy and thin however this is easily corrected in the crossover by shelving down the high treble slightly (about 0.5dB) for a more natural tonal balance.
If it's well crossed over a ribbon tweeter will disappear and not draw attention to itself.
The directivity afforded by the narrow tall ribbon and waveguide has some interesting properties. In the vertical plane they have the same directivity you'd expect from a 50mm diameter cone driver - starting to fall off a bit in the top octave once you go about 20 degrees off the vertical axis. As you go further off axis the treble rolls off more in a smooth progressive way without significant additional lobing.
Because they're relatively directional in the vertical plane this dramatically reduces high frequency treble bounce from the floor and ceiling when you're listening from a distance. I find this helps keep the treble well focused and avoids the floor and ceiling from stretching the treble image into a vertical mush.
This higher than usual vertical directivity also reduces the overall power response significantly at high frequencies, which means you can design for a flat on axis response without the speaker sounding too bright in a room, whereas a speaker with a lower treble DI (dome tweeter) typically has to have the on axis response sloped down a bit in the treble to avoid sounding over bright in room, but this sacrifices the neutrality of the direct path sound.
The higher vertical directivity is also great for minimising diffraction - if the tweeter is near the top of the cabinet the frequency at which diffraction from that top edge is a problem isn't actually a problem because the driver hardly radiates at those frequencies at 90 degrees from perpendicular. The result is that the top edge of the cabinet has little if any effect on the treble, unlike a dome tweeter where it has a major effect due to diffraction.
In the horizontal plane a waveguide loaded ribbon acts as a constant directivity driver, in the case of the G2 it's a relatively wide 120 degrees or so from about 3Khz up. As you go off the horizontal axis the response hardly changes at all out to a fairly wide angle then most of the treble starts dropping together - what you'd expect from CD. It's not perfect CD due to the small waveguide but it's a decent approximation.
This means that the sidewall reflections are attenuated (depending on how much you angle the speakers) but relatively spectrally flat. The off axis response in the horizontal plane is actually better than a dome tweeter at the top end since the ribbon itself is <10mm wide. It changes very little until you get out to +/- 60 degrees or so.
While diffraction from the sides of the cabinet is still an issue, the signal radiated along the cabinet face to the sides is still dramatically reduced from a plain dome tweeter, thus any diffraction anomalies are greatly reduced even before you start applying absorption and cabinet rounding.
In fact in the very early stages of testing I simply had the G2 sitting on top of an existing cabinet with no baffle around it at all - despite this the deviation in the frequency response was only about 1dB from the baffled response due to the directivity. In the same situation a free standing dome tweeter would be suffering from horrific diffraction response anomalies with an unmounted round face plate.
So all together you can design for flat on axis response, (without excessive power response in the room) have greatly attenuated ceiling and floor treble reflection (avoiding vertical stretching and smearing of the image) and have a nice spectrally balanced but attenuated sidewall reflection for good ambience.
Couple this with an extremely flat frequency response free of typical cone resonances and it's no wonder they can sound so good.
In some ways they're more challenging to use than a dome tweeter (crossover design - steeper filters etc, finding a suitable midrange driver that can go high enough to partner well with it) but in other ways they're much easier. (Greatly reduced diffraction artefacts due to their inherent directivity and a much flatter frequency response that doesn't need much tailoring at all)
Also the impedance peak (and acoustic peak) at resonance is typically very flat - so flat that I don't even have any resonance compensator/notch in my crossover, it just isn't needed.
Like most smaller ribbons (these are about 50mm tall visible ribbon area) the key to success is don't cross them over too low and get the crossover right.
High order filters are needed and midrange drivers that can cross relatively high, cleanly. (at least 3Khz) If you don't like high order filters or higher crossover frequencies, don't use a ribbon, at least not a small one like this, as you will either damage it or be disapointed.
I'm using the AC G2 with a 3Khz 4th order Linkwitz-Reily passive crossover, and I would not go any lower in frequency than this - that's about their comfortable limit. I've also used them earlier in the design phase with a 3rd order 4Khz crossover. I would not use less than 3rd order.
If your midrange driver can perform well at higher frequencies my gut feeling is that 4Khz 4th order is optimal for this model, especially if you want to build a larger speaker that is operating at a higher SPL level, (raising the crossover frequency even from 3Khz to 4Khz reduces the power dissipation on the tweeter greatly) and that is the target for my 3 way design which is slowly coming together.
Ribbon tweeters can be a little delicate - they don't take kindly to metal filings getting into the exposed magnetic gap (keep the front sealed with cellotape while working on/building the speakers or transporting them!) and are easily over driven if the crossover is too low or too shallow.
You do get a little bit of visual warning though - the ribbon will start to bow slightly when it heats and expands, if you see this you're overdriving it. If they overheat too much the ribbons will stretch and not return to their original corrugated form, becoming a little bit slack. They will still work but their response will change slightly.
I did manage to damage one ribbon very early on when I was doing quick testing with an unsuitable crossover however after replacing the ribbon foil they have been in use for 15 years now without incident, so they have stood the test of time. They still sound great, better than ever in fact as a couple of years ago I did a ground up redesign of the crossovers using Vituixcad.
A common theme I hear among ribbon sceptics is that they can't/don't sound good because they have higher THD than similar domes, and suggest that somehow this higher THD is actually pleasing to the ear and account for preferences towards ribbons among some people. Rubbish.
While it's true they do have slightly higher THD than similar performing domes, this is mostly at the top end of the treble so most of the harmonics generated fall outside the audible hearing range anyway. (What are the second and third harmonics of 10Khz ?) So I don't buy that argument at all.
Non-linear distortion has a very strong threshold effect - if distortion is below a certain threshold (or is masked) it is inaudible. A driver with 0.2% of low order harmonic distortion won't sound worse than one with 0.1% distortion as both are too low to be audible.
Comparing two tweeters and criticising one because it has slightly higher non-linear distortion when both are a long way below the threshold of audibility in music is folly, especially if the lower distortion driver is deficient in some other metric, such as directivity control or flatness of frequency response. If you optimise the wrong parameter at the expense of other more important parameters you won't get the best result.
As a result I don't take much stock in the opinions of people who obsess over minute level distortion measurements of tweeters. (I won't name names but they know who they are 🙂 )
For many years I was also a believer in very low levels of non-linear distortion being very important but work from people like Earl Geddes on the audibility of non-linear distortion made me start to question that belief and I now believe that linear distortions (especially amplitude and time) are much more important. That's not to say non-linear distortion doesn't matter at all, just that obsessing over getting super low values is unnecessary especially when you have to sacrifice elsewhere.
I only have experience with waveguide loaded ribbon tweeters, so when I talk about ribbon tweeters in this post some of their characteristics come from the ribbon itself and would apply to all ribbons, and some are a result of the directivity pattern imparted by the waveguide which would not necessarily apply to a large open ribbon.
There is just something about the sound of a ribbon tweeter that is hard to quantify or explain. It's one of those "you have to hear it to understand" things. I hesitate to use subjective descriptions but I would say "naturalness", "transparency" and freedom from audible high Q resonances that lead to harshness are the main points. Waterfall plots on ribbon tweeters are usually exceptionally clean.
With a good ribbon tweeter you can get crispness without harshness. They can be at the same time detailed yet laid back and easy to listen to without listener fatigue. Sibilance is not usually an issue with a ribbon tweeter (unless you mess up the crossover) as sibilance is typically a result of high Q cone breakup in the 6-8Khz region which is common on soft dome tweeters.
Some ribbon tweeters (including the G2) do have a slight emphasis above 10Khz due to a cavity resonance behind the ribbon as well as the directivity effects of the waveguide - if left uncorrected (and I suspect many designs don't correct for this) this can lead to the treble sounding a little toppy and thin however this is easily corrected in the crossover by shelving down the high treble slightly (about 0.5dB) for a more natural tonal balance.
If it's well crossed over a ribbon tweeter will disappear and not draw attention to itself.
The directivity afforded by the narrow tall ribbon and waveguide has some interesting properties. In the vertical plane they have the same directivity you'd expect from a 50mm diameter cone driver - starting to fall off a bit in the top octave once you go about 20 degrees off the vertical axis. As you go further off axis the treble rolls off more in a smooth progressive way without significant additional lobing.
Because they're relatively directional in the vertical plane this dramatically reduces high frequency treble bounce from the floor and ceiling when you're listening from a distance. I find this helps keep the treble well focused and avoids the floor and ceiling from stretching the treble image into a vertical mush.
This higher than usual vertical directivity also reduces the overall power response significantly at high frequencies, which means you can design for a flat on axis response without the speaker sounding too bright in a room, whereas a speaker with a lower treble DI (dome tweeter) typically has to have the on axis response sloped down a bit in the treble to avoid sounding over bright in room, but this sacrifices the neutrality of the direct path sound.
The higher vertical directivity is also great for minimising diffraction - if the tweeter is near the top of the cabinet the frequency at which diffraction from that top edge is a problem isn't actually a problem because the driver hardly radiates at those frequencies at 90 degrees from perpendicular. The result is that the top edge of the cabinet has little if any effect on the treble, unlike a dome tweeter where it has a major effect due to diffraction.
In the horizontal plane a waveguide loaded ribbon acts as a constant directivity driver, in the case of the G2 it's a relatively wide 120 degrees or so from about 3Khz up. As you go off the horizontal axis the response hardly changes at all out to a fairly wide angle then most of the treble starts dropping together - what you'd expect from CD. It's not perfect CD due to the small waveguide but it's a decent approximation.
This means that the sidewall reflections are attenuated (depending on how much you angle the speakers) but relatively spectrally flat. The off axis response in the horizontal plane is actually better than a dome tweeter at the top end since the ribbon itself is <10mm wide. It changes very little until you get out to +/- 60 degrees or so.
While diffraction from the sides of the cabinet is still an issue, the signal radiated along the cabinet face to the sides is still dramatically reduced from a plain dome tweeter, thus any diffraction anomalies are greatly reduced even before you start applying absorption and cabinet rounding.
In fact in the very early stages of testing I simply had the G2 sitting on top of an existing cabinet with no baffle around it at all - despite this the deviation in the frequency response was only about 1dB from the baffled response due to the directivity. In the same situation a free standing dome tweeter would be suffering from horrific diffraction response anomalies with an unmounted round face plate.
So all together you can design for flat on axis response, (without excessive power response in the room) have greatly attenuated ceiling and floor treble reflection (avoiding vertical stretching and smearing of the image) and have a nice spectrally balanced but attenuated sidewall reflection for good ambience.
Couple this with an extremely flat frequency response free of typical cone resonances and it's no wonder they can sound so good.
In some ways they're more challenging to use than a dome tweeter (crossover design - steeper filters etc, finding a suitable midrange driver that can go high enough to partner well with it) but in other ways they're much easier. (Greatly reduced diffraction artefacts due to their inherent directivity and a much flatter frequency response that doesn't need much tailoring at all)
Also the impedance peak (and acoustic peak) at resonance is typically very flat - so flat that I don't even have any resonance compensator/notch in my crossover, it just isn't needed.
Last edited:
When i got mine i got both round & rectangualr bezels, i prefer the latter, i might still have the circular bezels if you really prefer them.
dave
Not really much of use in this thread when it comes to evaluating ribbons.
I would offer:
1) the ribbon is only as good as the integration to the rest. If the ribbon "sounds" considerably "different" or "better" than what is below it, it can not be integrated
2) No architecture has the market cornered in sound quality. There are excellent solutions from domes, cones, horns, electrostats, AMTs, magneto planar and true ribbons.
3) Many solutions are philosophical in nature. Constant directivity is an example of this, as is wide dispersion. Unfortunately the directivity discussion is ALWAYS clouded by copy/paste bogusness.
I generally agree that limiting directivity is a virtue in domestic living spaces (your mileage may vary). Yes, we can compensate with room treatment to get reasonable response also off axis, BUT HF reflections in a small living room (small compared to the live concert space) give our brain indications to how small the room really is. I have had best results with a live end/dead end room with the (directional) speakers in the dead end. This setup gives the best chance to get the recorded ambience to the ear without additional confusion.
I have personally never been able to integrate plasma or ribbon tweeters myself. I do remember hearing stacked Quads with the Kelly/Decca ribbon sounding very compelling in a VERY large room.
One post drew my attention: beliefs are not acceptable - we only should consider the facts. This must be the most ignorant statement that I have ever read. If there were such persuasive "facts", all speakers would be built the same way. We have so much diversity because there are many solutions to playback enjoyment. Sure, there is science behind many of the solutions, but there is no one speaker that is considered by all to be the "reference". The demands are too diversified and even those very intimate with live sound do not agree.
My answer to the original question: No, ribbons are not the best. In certain situations they can be the "best" choice but that is a very loaded statement. Many speakers out there with ribbons have really miserable integration with no hope of ever working properly. In my world 99% of the speaker is the intentions of the designer not the architecture chosen.
I would offer:
1) the ribbon is only as good as the integration to the rest. If the ribbon "sounds" considerably "different" or "better" than what is below it, it can not be integrated
2) No architecture has the market cornered in sound quality. There are excellent solutions from domes, cones, horns, electrostats, AMTs, magneto planar and true ribbons.
3) Many solutions are philosophical in nature. Constant directivity is an example of this, as is wide dispersion. Unfortunately the directivity discussion is ALWAYS clouded by copy/paste bogusness.
I generally agree that limiting directivity is a virtue in domestic living spaces (your mileage may vary). Yes, we can compensate with room treatment to get reasonable response also off axis, BUT HF reflections in a small living room (small compared to the live concert space) give our brain indications to how small the room really is. I have had best results with a live end/dead end room with the (directional) speakers in the dead end. This setup gives the best chance to get the recorded ambience to the ear without additional confusion.
I have personally never been able to integrate plasma or ribbon tweeters myself. I do remember hearing stacked Quads with the Kelly/Decca ribbon sounding very compelling in a VERY large room.
One post drew my attention: beliefs are not acceptable - we only should consider the facts. This must be the most ignorant statement that I have ever read. If there were such persuasive "facts", all speakers would be built the same way. We have so much diversity because there are many solutions to playback enjoyment. Sure, there is science behind many of the solutions, but there is no one speaker that is considered by all to be the "reference". The demands are too diversified and even those very intimate with live sound do not agree.
My answer to the original question: No, ribbons are not the best. In certain situations they can be the "best" choice but that is a very loaded statement. Many speakers out there with ribbons have really miserable integration with no hope of ever working properly. In my world 99% of the speaker is the intentions of the designer not the architecture chosen.
Last edited:
Facts...
I dont think any one has presented any facts....yourself included.... @rowuk
One fact:
A dome, a cone, a planar panel structures all have break up modes, unless they are driven by EMF across the entire diaphragm.
Orthodynamic Planar, electrostatics, and Ribbons, plasma, are the only devices to do so. AMT perhaps also, although its operation is not so simple, and I suspect generates other 'artifacts' that must be minimised.
Supposition: due to the above fact, break up is ameliorated almost completely in those devices, result, cleaner sound.
Fact: a typical ribbon or plasma tweeter has a radiating surface/boundary, that is an order of magnitude lighter than any other diaphragm, presenting the same uniform movement, without break up.
Supposition: transient response is far more psycho acoustically relevant than low order THD.
Opinion: the brake up of a typical 20mm 25mm or 32mm dome, is offensive to my ear, and has never been conquered.
The only "near perfect" dome, is a hard dome.
The above facts and supposition/opinion, form the logical basis for my preference for Ribbons. (Having never had the pleasure of hearing ESLs, Planars, or Plasma)
I dont think any one has presented any facts....yourself included.... @rowuk
One fact:
A dome, a cone, a planar panel structures all have break up modes, unless they are driven by EMF across the entire diaphragm.
Orthodynamic Planar, electrostatics, and Ribbons, plasma, are the only devices to do so. AMT perhaps also, although its operation is not so simple, and I suspect generates other 'artifacts' that must be minimised.
Supposition: due to the above fact, break up is ameliorated almost completely in those devices, result, cleaner sound.
Fact: a typical ribbon or plasma tweeter has a radiating surface/boundary, that is an order of magnitude lighter than any other diaphragm, presenting the same uniform movement, without break up.
Supposition: transient response is far more psycho acoustically relevant than low order THD.
Opinion: the brake up of a typical 20mm 25mm or 32mm dome, is offensive to my ear, and has never been conquered.
The only "near perfect" dome, is a hard dome.
The above facts and supposition/opinion, form the logical basis for my preference for Ribbons. (Having never had the pleasure of hearing ESLs, Planars, or Plasma)
Last edited:
Well, it is a Fact that the longer the vertical source is the more directional it becomes at very high freq.s. (..assuming there is no waveguide like what RAAL does.)
It is Fact that for a given Sd that true Ribbons have much lower Mms (..than domes/compression drivers).
It's also a Fact that Ribbons have almost nothing in the way of suspensions (excluding the surround design of Viawave). Note: the Bohzen design isn't really a "ribbon", though it is a nice driver.
Finally, it is a Fact that I mentioned all 3 already in this thread. 😉 (..though I didn't previously say why the ribbon had vertical spl-loss.)
It is Fact that for a given Sd that true Ribbons have much lower Mms (..than domes/compression drivers).
It's also a Fact that Ribbons have almost nothing in the way of suspensions (excluding the surround design of Viawave). Note: the Bohzen design isn't really a "ribbon", though it is a nice driver.
Finally, it is a Fact that I mentioned all 3 already in this thread. 😉 (..though I didn't previously say why the ribbon had vertical spl-loss.)
Last edited:
I agree ScottG.
Well to be more accurate, it is a fact that the larger the source emitter becomes, the more directional it becomes with increasing frequency.
If the emitter has an aspect ratio greater than 1, then the largest dimension directivity increases relative to that plane of that dimension.
Aside: I enjoy my small ribbons, and have not spent time listening to anything approaching a line source. I dont believe a 2 inch ribbon can be considered as such.
Ribbons dont have suspension, mainly because they do not require the radiating surface to be rigid, and do not rely on any type of pistonic movement for their analysis or reproduction. In fact, flexibility is key to their operation.
The entire radiating surface is driven, and bows convex or concave, and as Mandrake cleverly pointed out, this is their failure mode, when overdriven, they are stretched.
But is is opinion, if the vertical restriction in directicity is good, or bad.
My opinion, is that is is good, especially outside a acoustically 'dead' room.
If i lived in an anechoic space my opinion might be vastly different
Well to be more accurate, it is a fact that the larger the source emitter becomes, the more directional it becomes with increasing frequency.
If the emitter has an aspect ratio greater than 1, then the largest dimension directivity increases relative to that plane of that dimension.
Aside: I enjoy my small ribbons, and have not spent time listening to anything approaching a line source. I dont believe a 2 inch ribbon can be considered as such.
Ribbons dont have suspension, mainly because they do not require the radiating surface to be rigid, and do not rely on any type of pistonic movement for their analysis or reproduction. In fact, flexibility is key to their operation.
The entire radiating surface is driven, and bows convex or concave, and as Mandrake cleverly pointed out, this is their failure mode, when overdriven, they are stretched.
But is is opinion, if the vertical restriction in directicity is good, or bad.
My opinion, is that is is good, especially outside a acoustically 'dead' room.
If i lived in an anechoic space my opinion might be vastly different
Last edited:
I think I prefer the circular ones visually, but the rectangular face plate actually fits in better with the new design I'm working on.
The existing circular ones in the 2 way are just flush mounted onto the main baffle in a conventional way, as a result I have a 2nd order all pass filter incorporated into the tweeter crossover in addition to the high pass and L-Pad to add enough broadband delay to allow me to use a symmetrical 4th order linkwitz-reily crossover and still achieve very good phase tracking across the overlap region (at least on axis) despite the significant acoustic centre offset.
This was done for expediency sake as the cabinet already existed with the ribbon tweeter replacing other tweeters that had previously been tried, and the alternatives were a tweeter pod on the top to set the tweeter back (and a new front panel on an already finished cabinet) or "bending" the phase using asymmetrical slopes which I did not want to do. So I gave the All Pass filter a try for the first time in a passive network.
It actually came out well and integration is excellent, however for the 3 way I want to go the tweeter pod approach so I can set the tweeter back physically to get mechanical acoustic centre alignment and eliminate the all pass section in the crossover, which is a lot of components saved.
I've used these tweeters set back sitting on the top of a cabinet before and it works well. They have enough vertical directivity that diffraction off the top corner of the cabinet doesn't seem to be an issue if you smooth the transition area. (Or even if you don't, really)
One problem with the current tweeters on a conventional full width panel is that there is still a small amount of treble diffraction from the left and right sides of the cabinet despite the waveguide. It's much smaller than what the same cabinet produced with a dome tweeter but it is noticeable and causes a little bit of ripple in the polar response slightly off axis. And because the cabinet is 39cm wide the ripple is fairly tightly bunched, making toe in a bit fussy.
The idea of the pod with the rectangular faceplate tweeter is to make the pod only slightly larger than the tweeter itself and round the edges so that the diffraction edge to the sides of the tweeter is very close instead of 15+ cm away..
This pushes the diffraction frequency up higher to frequencies where the driver is more directional, (less energy reaching the diffraction edge at frequencies where it matters) reduces the polar density of any diffraction ripple that does slip through (less fussy about toe in as ripples in the response will occur over a wider angular offset from on axis) and also reduces the time delay of that diffraction to the point where it is in the "fusing" region instead of starting to be detected as a separate event from an imaging point of view.
That's the theory anyway, it is yet to be put to the test. 🙂
For this very narrow pod approach, the narrower the tweeter the better so the circular version would have added unnecessary width which I don't need and reduced my flexibility in coming up with the optimal size and shape pod.
Last edited:
So what would you put on the floor to establish that? The wall and ceiling solutions I can think of btw.
Don't want to spoil it, but history shows that a lot of flat drivers actually needed damping carefully applied in order to reduce unwanted resonances and/or to improve linear behavior. So there's no general truth in that I guess. And you steer away very swiftly from the challenge to fix the membrane in it's rest position, which brings up all kinds of trouble regarding the linear movement of the membrane as a whole. There's no free lunch, alas.
I'm crossing mine over to an 8" full range driver (albeit at 3Khz) and have managed to achieve pretty much seamless integration to the point where it sounds as coherent as the full range driver on its own, and the tweeter does not stand out as a separate source.
But a lot of time and work went into the crossover to achieve that, it wasn't easy. Both drivers have 4th order filters hand optimised to give an actual 4th order Linkwitz-Reily response acoustic response, (including two notches at 2Khz and 4Khz on the full range driver to get the response accurately into shape around the crossover point) the tweeter crossover has an all pass filter to add "linear" delay across the overlap region to compensate for acoustic offset, (so the two filters can phase track within +/-30 degrees across the entire 2 octave overlap region) and also shelves the top end down slightly to compensate for the lift at 10Khz the tweeters have. The levels between the two drivers have been carefully matched to within about 0.2dB.
Because the ribbon tweeter is more directional than the typical dome tweeter I think it integrates better with a midrange driver that has a reasonable amount of directivity of its own. In this regard it integrates much better with a large full range driver than a dome tweeter does. (I've tried two different domes and a horn loaded dome on the same speaker before settling on the ribbon and the ribbon integrated the best with the horn dome second best and normal domes a distant last due to a too abrupt change in directivity at the crossover)
For sure, ribbons are not the best in all ways. They have limited power handling and SPL capability and are more easily damaged. (Even a vacuum cleaner trying to suck some dust off the front of one will destroy it) Small ribbons like the G2 are only suitable for living room sized systems, you certainly wouldn't use them on anything higher SPL than that like a PA system or outdoors. (You can't use them outdoors due to wind damage anyway)
Larger/longer ribbons can produce a lot higher SPL and go lower in frequency but start to sacrifice in other areas - the taller you make them the higher their SPL and power handling becomes but the more directional they get. A 50mm tall ribbon is about the longest you want for vertical directivity - once you get to 100-150mm they're far too directional vertically but still far too short to act as a line source.
If you make the ribbon wider the variations in flux density across the gap increase and you start to get bending modes in the diaphragm due to unequal field strength, while not "cone breakup" per se this is a deviation from the ideal of a flat planar piston that moves back and forth as one without any bending motion, so wider ribbons start to lose that purity and clarity of a narrow ribbon.
But if your design can work within the limitations above (the sensitivity/power handling vs directivity trade off) then I do think that ribbon tweeters are the best sounding tweeter.
Although it may be a matter of taste, I believe there are advantages to having asymmetric directivity - e.g. a different pattern horizontally than vertically.
Sidewall reflections can add spaciousness to the sound, but if the reflections are excessively strong (omnidirectional speaker) or have a much darker nature (low treble in the sidewall reflection due to a horizontally wide driver) that doesn't give a good result, and it's one reason that I don't use a full range system as a daily driver even though they can sound great for specialised listening.
A constant directivity profile in the horizontal plane fits the needs of most rooms quite well I think - it reduces the sidewall reflection to a manageable level but keeps a good spectral balance in the reflection.
While horizontal reflections are considered to be "good" up to a point, vertical reflections (off the floor and ceiling) are almost universally considered to be bad, and they are not processed the same way in the brain. They tend to cause tonal imbalances and smear the image location instead of being heard as a distinct source because they come from the same yaw angle as the original sound.
So more directivity in the vertical plane than the horizontal plane is beneficial in my opinion, and 15 years listening to speakers with more high frequency directivity vertically than horizontally seems to bear that out.
You don't want it so directional vertically so that the sweet spot is too small though. A 50mm tall ribbon seems to be a good middle ground here while the long 150mm ribbons are far too narrow vertically.
This is largely true - a ribbon is directly driven across the entire radiating diaphragm unlike most other types of drivers and in theory this means a a uniform moving piston with no breakup modes.
In practice they are not perfect though for a couple of reasons:
1) The magnetic field strength in the gap is not completely uniform across the entire area of the foil. This means slightly more force is imparted by the electromagnetic field on some parts of the ribbon than others, as a result some parts will move back and forth slightly more thus the ribbon as a whole bends slightly with each cycle.
This bending isn't the same as uncontrolled propagation of bending modes along a cone or dome though which travel like a wave and form resonances at spot frequencies, it's more of a standing wave where the wave pattern stays in the same location at different frequencies - eg it just moves back and forth slightly more at the sides than the middle.
Optimising the magnetic gap design to make the field as uniform as possible is a large part of designing a good ribbon tweeter, and those with narrow magnetic gaps (and narrow foils) usually perform a lot better in this regard as it gets harder to make a uniform field in a large gap.
2) The air and mechanical loading on the foil varies at different locations on the foil. The foil is clamped at the top and bottom so obviously can't move as much right next to where it is clamped as it can near the middle. On the G2 the foil is 70mm long between clamps but only the middle 50mm is actually visible through the faceplate with 10mm hidden at the top and bottom - presumably this is done to block radiation from the top and bottom parts of the foil where it is clamped.
Air will rush around the edge of the foil through the gap between the foil and magnetic pole, this will have a slight air loading effect at the edge.
One other negative of a closed back ribbon tweeter is that there is a small cavity resonance between the ribbon and the back of the tweeter - some like the G2 have a felt pad glued to the magnet structure behind the foil to damp most of this but there can still be a small resonance somewhere between 10-20Khz, however this is easily corrected with an RLC notch if you want to.
All in all though, it's much closer to the breakup free ideal piston radiator than a traditional tweeter, and the proof is in the waterfall plots.
I think a lot of things are more important than very low order THD in all honesty.
As I mentioned in my earlier post perception of non-linear distortion has a threshold - for a given frequency, SPL and masking situation you can hear distortion above a certain threshold and can't hear it below that threshold, especially if there are masking factors as there usually are in music. It's a pretty binary thing - you hear it or you don't. Once it's low enough that you don't hear it, reducing it further can be measured but not heard.
Linear distortions on the other hand like Amplitude response errors are very audible and you can continue to make further and further gains on how good it sounds as you get the response flatter and smoother. In fact the closer a speaker is to flat and neutral the smaller the errors you can discern.
If you have something that's really lumpy and un-flat (with peaks and dips in the 10dB range through midrange and treble) you'd be struggling to hear a 1-2 dB adjustment.
But listen to something that's already flat within say +/- 1dB and you can hear a broadband shelf adjustment of 0.2dB quite easily, for example tuning the L-Pad on a tweeter as it changes the character of the sound and imaging quite a lot.
I've heard some good soft dome tweeters before, but they never sound convincingly real. They sound "HiFi", and sibilance is usually a problem. Done right a ribbon tweeter can cross the uncanny valley and start to sound spookily real.
But the integration to the rest of the speaker has to be dialled in perfectly for that to happen.