So my answer is yes!
Better, lower distortion high frequencies. in the audible range.
Frequency response with sine waves, is different than musical wave forms
I have a pair of MP-02 Plasma tweeters that I sometimes use in summer (windows open) that I believe are capable of going to 100khz
I cross these over (3rd order) to my ML Monolith III’s at 12khz, and yes you can hear the difference, if I had to explain it’s sound, it would be an extended delicacy, with more air there, you can really notice it on harpsichord it just floats in front of you, like you can reach out and touch it.
I remember reading about tests where people could distinguish music with bandwidth opened to 50kHz vs 20kHz. It is because our hearing operates in time domain more then frequency domain. Music is processed as wave and not as bunch of frequencies. Also, higher frequency limit ensures lower phase shift in audible band improving harmonics summing. My amplifier has -3dB @ 70kHz and it is causing -22 degree phase shift at 20kHz. Limiting amplifier to 20kHz would cause the same phase shift at about 6kHz affecting the sound, even if your hearing only goes to 15kHz.
I like the extension of a Tannoy ST-100 super-tweeter on top of an Esotar 330D tweeter that has good HF extension of itself. (In the Merlin application the Esotar is down several db at 20khz.) But while the ST is set high at a nominal 18khz, in actuality its third-order crossover rolls down into the more audible mid-treble range. It’s conceivable that a ST is a glorified hearing aid for aging audiophiles with attenuation in their own HF response.
The ST does open up perceived treble extension and air. But put your ear close to it and all you’ll hear is faint chirping.
^^ Its not the ability to hear to 20KHz or beyond that is why you can hear differences in bandwidth outside of the 'audio region'. Please read Kijanki's post above.
When you have a rolloff at or near the audio passband, its audible as phase shift down to 1/3rd to 1/10th the cutoff frequency, depending to a certain extent to the steepness of the rolloff. Phase shift is converted by the ear/brain system (just as distortion is also) to tonality. The result is a tonal coloration. This is why bandwidth is so important.
Sorry to be late to the party Veroman, but better than never?
While I think that just about everything already said is true, there is another aspect to consider: the crossover.
As Ralph (atmasphere) points out, rolloff causes phase shift, which is very bad. More generally, all RLC or variant circuits cause phase shift. All analogue filters consist of RLC circuits. All crossovers, whether in the speaker or in an analogue "electronic" crossover, use filters, and so cause phase shift. Hence all analogue crossovers cause phase shift.
If you are limiting the super tweeter (to protect it from higher energy low frequencies) with an analogue high pass filter, it too will cause a phase shift. That is one important reason why full range ESL's sound different from any other speakers - they have no crossovers.
Hence super tweeters might sound better to some people in some systems, but the benefit comes at the cost of phase shift. I suspect that the most bang for the buck is in quality caps and resistors and inductors in the crossovers; that is, if you don't spring for ESL's. YMMD
"Hearing" concerns the conscious notion that we are aware of sound, and the unconscious response to sounds that we may not recognize as such. Our hearing mechanism is mechanical- electrical. Specifically sound activates the mechanical elements of our auditory system which convert mechanical movement to an electrical impulse sent to the brain. Our brain recognizes impulses within a specific bandwidth as "sound", but does not consciously provide a response to impulses outside of the bandwidth range. But- our auditory mechanism responds to sound outside of the normal range of hearing, and our brains still receive an electrical impulse from the out of range stimulus. While we are not conscious of a "sound" generated by out of bandwidth signal, we do acknowledge a response at the subconscious level.
3 landmark studies proved that we can respond to content above 20khz. In the early 1950s the US Army tested the hearing of soldiers and determined that while we can not "hear" content above 20khz, we are aware of it when it is removed from program material. Soldiers reported that something had changed, or was missing from the content, but they could not fully describe what was missing. Just that is was "something". This study and findings were reaffirmed by CalTech in the 90's, and by Swedish researchers about 10yrs ago.
Just about everyone who has reviewed a super tweeter reported that even though they could not "hear" the super tweeter, nor could the pinpoint specific sounds as coming from the super tweeter, felt that it added something beneficial to the overall content. And they definitely could tell when the super tweeter was silent or removed.
Whether or not content above 20khz is important to you, it can be a factor in accurate reproduction of recorded music. Brass and string instruments produce overtones that extend beyond 20khz and can reach up to 40khz. 16/44 digital introduces a brick wall filter at 22khz that abruptly cuts off the harmonics, and causes phase anomalies and odd order distortion to be reflected back into the audible range. How ? Picture a pool of still water, the surface looks like glass. Then drop an object into the center. Ripples will emanate in concentric circles from the point of impact, and flow outward until they come into contact with the wall of the pool. Upon impact with the wall, the ripples are reflected back towards the center, but are out of phase with the ripples still coming from the center. Some of the ripples will cancel each other, some will enhance each other, but the overall effect will be very noticeable.
HiRez digital allows for the filter to be moved out to the point where it no longer interferes with extended instrument harmonics. 24/96 extends it to 48khz, and higher sampling rates move the wall further outside the audible range. Listeners almost universally report that hirez sounds smoother, more natural, and more analogue like compared to 16/44 Red Book.
A long winded answer, but I hope it helps.
I’d wager any audibly perceived difference for the better with the addition of a "super tweeter" has little to none to do with frequencies extending 17-20kHz, but rather that the tweeter in question is better at handling the frequencies that are actually audible in the upper octaves. Equally important (even more so, I’d say), is the reproduction of the lower spectrum of the tweeter’s working area, and how this meets the with the midrange. To my ears most, if not all direct radiating dome tweeter lack energy so to, perhaps paradoxically, make them sound less like tweeters, and more like something that’s a natural extension of the midrange. Mention of frequencies that extend way beyond 20kHz, together with their claimed importance, seem more like marketing ploy than anything of real significance.
I heard the Sopranino, by ENIGMAcoustics, a free-standing, self-contained, plug-and-play, self-biased electrostatic super tweeter at CES in 2013. It was being used with the Magico S5 speaker and Pass Labs XA 100.6 monos. With the tweet out of the circuit, the female vocalist was centered in the apparent space, slightly forward with great imaging. Plugging in the Spranino changed the image from a small hall to a large theater and moved your apparent seat back about 20 rows. Incredible! And no, I have no idea how it achieved that effect, but I have to agree that Super Tweeters seem to affect the sound and the image below 20Khz..
The audio extremes challenge the Fourier-centric idea of 'sonic recognition'. My experience with really hi-end tweeters like RAAL, Fostex, etc. only operating 8K+ tell me these have more impact on spatial presentation than serious spectral significance. They simply 'finish the job' of presenting an instrument or voice. Cymbals can be spectrally completed with 10K+, but 65K - really? I'm not sure on that... But if it sounds better to you - GO FOR IT!
so i finished my project. the 1.5 cap is a bit too much into the high mids. am getting a 1.0 to replace it. the improvement is noticeable. air is the best word to describe it. cymbal sizzle much more delicately delivered. i mounted them in their own open baffles and they perch on stands above the ob full range baffle