Advantages of beryllium?

@phusis

I believe many people are able to hear what you describe with metal or rigid material versus damped materials. Equally there are others who dont seem to pick up on it or be sensitive to it. Maybe some of us listen more closely to timbre (the tonal content and the way it decays). The rigid drivers measure very well and have a wider usable bandwidth which gives the speaker better frequency range specs - so they have strong merits. My point is that this extra bandwidth comes at a price - the in band performance is not as clean on a waterfall. I still have not seen a better measurement on a tweeter than the Excel Millenium soft dome made with a doped sonolex fabric - Harbeth use this tweeter and Harbeth midrange is a damped design too that also just happens to be highly regarded for mid range quality - of course I believe this is no cooincidence and that transducer design and material is very important.

@erik_squires

LOL The author is entirely correct about smiley EQ curve speakers with the standard industry midrange scoop ....however this is INDUSTRY WIDE - you can hardly blame Stereophile or JA for what consumers want and buy in huge quantities. Junk food is bad for you too but people love it and but it in large quantity! This has been called mid scoop or BBC dip but it is highly prevalent - B&W dominate this style sound.

If you don’t like industry standard smiley EQ (presumably because you don’t yet need hearing aids) then get ATC which are flat (but actually sound midrange forward compared to most everything else.)

@bar81

+1 Exactly how I read JA last paragraph. Full of euphemistic BS whenever the POS SOTA ultra expensive audio jewellry measured performance is poor in certain aspects. He is also careful not to compare products and be overly enthusiastic when performance in a reasonable price product is exceptional - usually just saying he is impressed. After all a reviewing magazine needs advertising revenue and cooperation from the industry - so it can't pick winners and it's job is to cheer on every latest product. So kudos that JA still manages to convey his views to those who are discerning enough to translate that last paragraph.

@Koost_amojan

You are correct. The cost of the speaker is no reflection on the quality - especially the tweeter.

The Excel Millenium Sonotex soft done tweeter has the best waterfall I have seen. It is expensive at around 200 euros per unit and therefore very few manufacturers use it.

Since most speakers are fairly good in frequency response the waterfall plot plot is far and above the most important speaker measurement. Another key plot is the dispersion horizontally.

Unfortunately not many folks understand the importance.

@randy-11

Interesting link. I think they sum it up very well. Be is better in the very top octave than Aluminium or Titanium (10 to 20 KHz)

Note that on the plots Be is not necessarily the best from 3 to 10KHz (the really important range musically for the tweeter). This is the point I am trying to make about internally damped drivers - better performance over a narrower frequency range. If the sound from 10KHz to 20KHz is most important to you then Be is the way to go (at the expense of more resonance at 3KHz to 10KHz)

As as far as I am concerned there is not so much musically in the 10K to 20KHz range - so I prefer a tweeter that performs better from 3KHz to 10 KHz.

@kosst_amojan

Obviously some implementations are better than other. My statement pertains to the problem of lack of internal damping in most metallic drivers (magnesium being perhaps a notable exception).

I count 5 resonances on the Focal Aria tweeter lasting up to 1.5 msec. These resonances are much much longer than the wavelength of the sounds that tweeter emits (an eternity in terms of PRAT) and will definitely color the sound in the way I described.

The resonances are multiple as a rigid disc has multiple resonant modes. I know this for sure as I have large collection of Sabian, Zildjian and Paiste cymbals and they shimmer with all kinds of non harmonic tones. A cymbal is an exaggerated example but the same principle stands.

If you want to hear musical timbre you need a driver that is critically damped - being inert it just gets out of the way once the desired movement is executed.

2 msec of waterfall hash on the first example I gave is really going to affect everything: the timbre of transients on percussive instruments (twang of guitar strings) to the articulation of sbilance on vocals.

The high but very narrow or sharp resonance peaks sometimes seen on JA plots is something to do with his measurement setup - you can ignore those - they look too narrow to be real effects.

FWIW the best waterfall plot I have seen, apart from the one on the Joseph Audio speaker linked above, is on a Quad electrostatic. So for those people who can hear what a difference a good electrostic speaker makes audibly in timbre then you can appreciate how a rigid tweeter can be coloring the sound in the way I describe.

Here is a waterfall plot just for the Seas Excel Millenium fabric dome tweeter.

http://www.diyspeakers.net/snippets.org/ldsg.snippets.org/graphics/seas/e006w.gif

@mmeysarosh

The KEF R700 does measure well but the tweeter still has some hash and is not nearly as clean as the Seas Excel Millenium tweeter - see the outstanding lack of resonances (much better than any metal dome) in the waterfall plot for this speaker with the Seas Sonotex dome in their Excel Millenium tweeter (around 200 euros each!)

https://www.stereophile.com/content/joseph-audio-rm33si-signature-loudspeaker-measurements-part-2

@koost_amojan

You need to look for a clean waterfall with fast (damped) decay and no hashy stuff or resonances. The main resonant ringing is not normally in the audible band - my concerns are the additional resonances and lack of damping in rigid domes. The titanium dome on the JM Lab Utopia is up to 2 msec across the treble - this is a lot of vibration compared to something mostly clean or down 20 dB after 0.5 msec.

Highly rigid and light but poor internal damping.

Personally I don’t like the splashy sound of drivers of this type design (metal and highly rigid). They have great bandwidth that makes for impressive measured performance but I find the sound is "splashy" due to the way rigid materials vibrate naturally (like a bell vibrates and rings after an initial hit but a damped material like a pillow does not).

Splashy is a good term - as in when you splash the water it makes a lot of sound after the initial splash. Acoustically this means the driver imparts its own sound to the timbre whereas an internally damped cone material is much more inert - contributing much less coloration after the sound stops.

I prefer damped designs even though they tend to have a narrower bandwidth and can suffer from breakup and therefore require more careful design and larger more expensive drive motors. Damped cones sound much more natural and faithful to the original tone/timbre of recorded instruments even if they are not as linear on a speaker frequency plot.

Here is an example of a titanium tweeter - look at the ringing in the waterfall plot in the treble !!!

https://www.stereophile.com/content/jmlab-utopia-loudspeaker-measurements-part-2

Here is an example of Be - similar problem in the treble but very much better than titanium

https://www.stereophile.com/content/focal-maestro-utopia-iii-loudspeaker-measurements