There are Asian mines and suppliers of BE. TAD still makes BE diaphragms, from a different source and technology. And so does Radian and JBL so far, supplied by Materion.
Materion has changed some of they're BE foil products, thickness etc.
And the whole line of "hifi" mfgs are decimated.
Likely have been warned in advance too.
They can develop new diaphragms with different thickness foils.
Use different suppliers.
Or vapour deposited BE, but that makes the prices skyrocket.
So there are options, just might not be as convenient for Scanspeak etc.
So that is what one is left with.
I would speculate that the beryllium foil was used in several industries, loudspeakers being the smallest. As those other industries changed, their demand for Be foil dropped. Eventually the supplier was not selling enough foil to economically sustain the production. The loudspeaker market is not big enough, by itself, to justify this product line.
I have seen this same situation numerous times in the defense industry, where production quantities are very small compared to other industries.
I have seen this same situation numerous times in the defense industry, where production quantities are very small compared to other industries.
It is still used in several industries, one of its main uses is in X-ray machines, mammography, and other medical applications.
Making sure that we do not get exposed to unnecessary radiation under medical care, that's not going to change anytime soon.
Materion is not doing very well financially, on the stock market etc. lately. That is the most likely reason.
So one sheds off products that have lower production volumes and profitability.
They're stock values and ratings are not very good to say the least. It's just the basics of profitability.
NGK Berylco, is still going strong and has supplied Pioneer /TAD, and Yamaha for decades already, for they're vapour deposited diaphragms. Through they're Japanese department.
They also make BE foil in the required thicknesses etc.
Hopefully they find other sources after a while and things go back to 'normal', because there are other sources for raw materials and products.
Just a question if they are willing to do it in a small enough scale for audio manufacturers, at a reasonable price.
Medical use carries a premium in price, vs our uses.
Last edited:
Comparing Bliesma A and B tweeters directly, free field on axis (eliminating the differences in directivity due to different dome profiles), both filtered to flat FR on axis, to me there is a distinct difference in timbre. The B has a weird softness to it, that I actually find annoying and reminds me more of silk tweeters. A matter of taste, but I actually prefer the sound of the cheaper A variant.
All in all, over a couple of decades of using and building speakers, it has (literally) paid off for me to go into any comparison without any preconceptions, especially the worst preconceptions / assumptions "more expensive is always better" and "more complicated / elaborate is always better".
There's some more to it, that I can't reveal due to NDA / privacy rights, but I am quoting this post because it is for the most part accurate. My personal opinion would also be that there should be a public statement, not just from Bliesma, but from all manufacturers that currently suffer from this issue. We're their fricking customer base after all - not good to keep us in the dark.
@stoneeh All other things being equal, the difference between a good Alu/Mg alloy dome and one made from Be is due to the down-modulated breakup components smearing the pistonic range. Ive always advocated the use of an HF notch on most metal domes to pull this sharp resonance mode down 10 - 15 dB. It hinges on the data obtained from the publshed specs and impedance plot as to how much, how wide and at what frequency to attenuate. I know there are many skeptics here who don't think the major HF break peak can do harm to the rest of the audible pistonic band covered by the tweeter. I tried this on myself in a blind A/B test with an Audax Ti dome and was rather surprised how it affected the entire HF pass band. In the case of the Audax dome, the peak was over 15 dB @ 25.5k and a Q of 10. With the peak notched out, the sound stage became more holographic and tangible, plus the treble sounded more like that of a higher end fabric dome with more extension. Once the filter was bypassed, the HF sounded a little rougher, spitty, less accurate and overly sibilant. The better Alu/Mg domes weren't as much affected by the breakup, but you could still hear the effects of it.
I havent had a chance to compare Be to Al/Mg or Ti in this sort of fashion, but the sonic impression I get from a Be tweeter compared to a higher end metal dome is the Be tweeter always sounds a little duller and smoother without any harshness. This is alot like how the other metal domes sound with the notch filter inserted. The only soft dome I've ever heard which can hang dynamically and level wise with the T25B is the Morel TSCT1044 / 1104. This one sounds like a very good metal dome without the HF breakup.
So yes, there is a distinct difference between the sound of a Be dome and a Ti or Alu/Mg dome if they're put side by side. They're just not as easy to replicate the difference on when not listening to critical music
I havent had a chance to compare Be to Al/Mg or Ti in this sort of fashion, but the sonic impression I get from a Be tweeter compared to a higher end metal dome is the Be tweeter always sounds a little duller and smoother without any harshness. This is alot like how the other metal domes sound with the notch filter inserted. The only soft dome I've ever heard which can hang dynamically and level wise with the T25B is the Morel TSCT1044 / 1104. This one sounds like a very good metal dome without the HF breakup.
So yes, there is a distinct difference between the sound of a Be dome and a Ti or Alu/Mg dome if they're put side by side. They're just not as easy to replicate the difference on when not listening to critical music
@profiguy: I considered this more or less a non-issue, but I did in fact try a notch filter on the A.
My on-paper assessment would be that subharmonics of the ultrasound membrane breakup could potentially affect the passband, BUT we must remember that the breakup is only of influence when excitated by the input signal. The breakup of the Bliesma A tweeters is around 30 kHz. As long as the signal source is a digital one with the usual 44,1 or 48 kHz, the signal ends at about 20 kHz. So there's simply nothing there to excitate the membrane resonance. One could argue that harmonic distortion components do, but these are so extremely minimal with the Bliesmas even at elevated SPLs that I don't think you can argue they're relevant.
As for the practical part, I used the procedure documented here: https://purifi-audio.com/blog/app-notes-2/low-distortion-filter-for-ptt6-5x04-naa-11 . The membrane resonance can only be properly electrically dampened via a passive serial notch filter (or low pass, as a matter of fact). Passive parallel notch filters, as well as active ones, have less or no effect.
I used my own impedance and acoustical measurements. Manufacturer data might not be accurate enough for every tweeter, due to series deviation.
My listening impressions would be that I heard no difference. Fwiw, I employ the same type of passive circuit in my mid frequency drivers, and I hear noticeable difference there. But with the Bliesma A(luminum) HF driver, none.
Not judging on your impression / test, just relaying mine. Anyway, overall an interesting and relevant point.
My on-paper assessment would be that subharmonics of the ultrasound membrane breakup could potentially affect the passband, BUT we must remember that the breakup is only of influence when excitated by the input signal. The breakup of the Bliesma A tweeters is around 30 kHz. As long as the signal source is a digital one with the usual 44,1 or 48 kHz, the signal ends at about 20 kHz. So there's simply nothing there to excitate the membrane resonance. One could argue that harmonic distortion components do, but these are so extremely minimal with the Bliesmas even at elevated SPLs that I don't think you can argue they're relevant.
As for the practical part, I used the procedure documented here: https://purifi-audio.com/blog/app-notes-2/low-distortion-filter-for-ptt6-5x04-naa-11 . The membrane resonance can only be properly electrically dampened via a passive serial notch filter (or low pass, as a matter of fact). Passive parallel notch filters, as well as active ones, have less or no effect.
I used my own impedance and acoustical measurements. Manufacturer data might not be accurate enough for every tweeter, due to series deviation.
My listening impressions would be that I heard no difference. Fwiw, I employ the same type of passive circuit in my mid frequency drivers, and I hear noticeable difference there. But with the Bliesma A(luminum) HF driver, none.
Not judging on your impression / test, just relaying mine. Anyway, overall an interesting and relevant point.
Thanks for the tip for good tweeter. I always little bit overlooked Morel drivers, they are usually very expensive and measured performance is not good. But there are some good looking models, and I see TSCT1044 measures really well. Unfortunately TSCT1044/1104 are the most expensive tweeters Model offers, damn.
T25S might be similar in sound character....
SS already announced this over 6 weeks ago:
https://www.scan-speak.dk/scan-speak-newsletter-aug23-alert/
Others will announce when they are ready/exhausted their Be supply…
https://www.scan-speak.dk/scan-speak-newsletter-aug23-alert/
Others will announce when they are ready/exhausted their Be supply…
I’d speculate you are correct in your assertion that different driver materials sound different, but not for the reasons of differing off axis response profiles measured at specific frequencies……this where the objective argument for measurements falls apart………complex harmonic content is NOTHING like target or tone sweep response…..bending modes from accompanying pass band response modes drastically change the performance.
@stoneeh Your findings with the Alu version domes is understandable and it does matter which Alu alloy they use, so fair to say not all Alu domes are the same regarding this as well as the type and shape of dome used. If you were to run "pure" Alu (not an alloy), you'd run into fatigue issues with cracking and the like, so there's no way around this necessity to use a stronger combination of metals. This is why I refer to these domes as Alu/Mg, being the most common alloy used, among a few other minor ones mixed in.
The right way to notch the tweeter breakup is using a combination series, parallel circuit, forcing the signal to devide with an additional group of parallel components to the tweeter instead of the driver impedance dictating how much voltage drops across it after just a series filter circuit before or after the HP filter. This also dampens the tweeter electrically at the VC should any outside mechanical vibration or shock happen to excite the diaphragm externally.
This may sound trivial, but we've taken Ti dome tweeters and subjected them to the sound of ringing a small hand bell loudly and empty glass bottles being struck, which was in the harmonic range of exciting the dome resonance and could be measured with a scope at the tweeter terminals at a surprisingly high level. I'm not saying this is a critical thing, but it does prove that even though the drive signal to the tweeter doesn't contain any high frequency harmonic components, the tweeter is quite happy to resonate at this breakup frequency without any input drive signal.
The other thing which can excite the resonance is HF / RF noise getting through the amplifier, despite the typical output zobel used at the speaker terminals of most amps. There are a few other ways HF can enter the signal chain despite the music not containing these frequencies of around 25 khz. Transients from a phono stage fed by an MC cartridge is one way, depending on the design of the RIAA EQ. If you're lucky to own older analog recordings with true HF content past 20k, it only takes a cymbal crash or louder triangle to aggravate the tweeter breakup. If you've ever heard what this is like when it does happen, it sounds horrible, like a sharp crack leading the played note. The upper end resonance of a MM cartridge can be high enough if not terminated properly with the right impedance at the preamp or wrong cable capacitance or combination of the two. This can aggravate tweeter resonance too.
I know this all sounds like splitting hairs, but it can affect how well the tweeter performs if its prone to HF resonances and enough signal in the 25k range gets through - the highest likelihood of this happening is with percussion instruments, specifically cymbals. You cant directly hear it happening, but it can spoil the listening experience, being one of those things which makes your system sound unexplainably bad at certain times, sort of like AC motor induced noise getting into your signal chain during the day and being absent at night. Some people will never notice it if they're not perceptive to it or can't hear it due to their system not being able to reproduce it. Yes, the brick wall filter in your DAC will get rid of it at 44.1k Redbook, but a high resolution 96k recording transfered from 30 ips analog tape can contain enough information up at those frequencies to light up the tweeter.
The right way to notch the tweeter breakup is using a combination series, parallel circuit, forcing the signal to devide with an additional group of parallel components to the tweeter instead of the driver impedance dictating how much voltage drops across it after just a series filter circuit before or after the HP filter. This also dampens the tweeter electrically at the VC should any outside mechanical vibration or shock happen to excite the diaphragm externally.
This may sound trivial, but we've taken Ti dome tweeters and subjected them to the sound of ringing a small hand bell loudly and empty glass bottles being struck, which was in the harmonic range of exciting the dome resonance and could be measured with a scope at the tweeter terminals at a surprisingly high level. I'm not saying this is a critical thing, but it does prove that even though the drive signal to the tweeter doesn't contain any high frequency harmonic components, the tweeter is quite happy to resonate at this breakup frequency without any input drive signal.
The other thing which can excite the resonance is HF / RF noise getting through the amplifier, despite the typical output zobel used at the speaker terminals of most amps. There are a few other ways HF can enter the signal chain despite the music not containing these frequencies of around 25 khz. Transients from a phono stage fed by an MC cartridge is one way, depending on the design of the RIAA EQ. If you're lucky to own older analog recordings with true HF content past 20k, it only takes a cymbal crash or louder triangle to aggravate the tweeter breakup. If you've ever heard what this is like when it does happen, it sounds horrible, like a sharp crack leading the played note. The upper end resonance of a MM cartridge can be high enough if not terminated properly with the right impedance at the preamp or wrong cable capacitance or combination of the two. This can aggravate tweeter resonance too.
I know this all sounds like splitting hairs, but it can affect how well the tweeter performs if its prone to HF resonances and enough signal in the 25k range gets through - the highest likelihood of this happening is with percussion instruments, specifically cymbals. You cant directly hear it happening, but it can spoil the listening experience, being one of those things which makes your system sound unexplainably bad at certain times, sort of like AC motor induced noise getting into your signal chain during the day and being absent at night. Some people will never notice it if they're not perceptive to it or can't hear it due to their system not being able to reproduce it. Yes, the brick wall filter in your DAC will get rid of it at 44.1k Redbook, but a high resolution 96k recording transfered from 30 ips analog tape can contain enough information up at those frequencies to light up the tweeter.
Yes you are certainly right, harmonics would most likely be completely different. I wouldn't leave that off the table.
But for the sake of the argument, for people who believe harmonic distortion in drivers doesn't matter, or that they are masked etc...
For starters, it makes no sense to compare harmonic distortion for one driver at 85dB, vs a driver at 83dB.
So at the very least, frequency responses must be matched to compare harmonic content.
And to be clear, I HAVE measured the PTT6.5X04 NFA (pulp fibre), and the PTT6.5X04 NAA (alu) in the same cabinet, in 360 degrees.
And after matching the on-axis precisely by DSP, in their intended passband ie. 40Hz to 2.0 KHz, all of the seventy other frequency response measurements (Hor 10, 20, 30... 330, 340, 350, Vertical 10, 20, 30... 330, 340, 350) didn't match each other.
So yes, the vibrational / bending / eigenmodes are likely very complex and they also vary, dependant on voltage drive level, in 3D space.
But I wouldn't go so far as to say that the objective argument for measurements falls apart.
More like "we don't quite know why it sounds different" In fact, we may be needing to do different measurements, not just acoustic measurements.
eg.
How non-acoustic measurements were used to discover how the Stradivarius violin sounds different to other violins:
https://onlinelibrary.wiley.com/doi/10.1002/anie.202105252
Last edited:
T25A is a totally underrated tweeter and really great for a low price!
I also know a few people which prefere the more "detailed"/"direct" sound of normal metal domes. Many of the drummers btw
.But in my opinion they overdo this a little - it's a little to much for being natural. T25B has the same details but less "in your face" and indeed, I would say it sits between metal and silk. It reminds me to a ring radiator but without the strong beaming at high frequencies.
All this "hearing" is quite questionable!
What you have on CD or other digital media is : PROCESSED MUSIC !!!
Have you ever listen to REAL INSTRUMENTS ???
Obviously not , or your hearing/brain memory is defective!
Calibrate your biological hearing device (aka EARS) with UNPROCESSED MUSIC :
https://www.madisoundspeakerstore.com/test-cds/drums-bells-loudspeaker-evaluation-cd/
Every time I put a CD into the player I think to myself WTF have they done in the recording STUDIO ??? WHERE is the SOUND going to? Eaten by snakes?
Where's your motor (answer at 00:55sec) :
What you have on CD or other digital media is : PROCESSED MUSIC !!!
Have you ever listen to REAL INSTRUMENTS ???
Obviously not , or your hearing/brain memory is defective!
Calibrate your biological hearing device (aka EARS) with UNPROCESSED MUSIC :
https://www.madisoundspeakerstore.com/test-cds/drums-bells-loudspeaker-evaluation-cd/
Every time I put a CD into the player I think to myself WTF have they done in the recording STUDIO ??? WHERE is the SOUND going to? Eaten by snakes?
Where's your motor (answer at 00:55sec) :
I have that Cd along with many others like it, including my own recordings. Theres one instrument far more difficult to record and reproduce - the piano. Its percussive, melodic and has tons of overtones which decay at various rates. The harp itself partially resonates at much higher frequencies than we hear, but nonetheless is part of the total sound package, which is extremely complex. Then there are all those hammers and strings being struck, most with 3 strings per note. Its a beast of an instrument that will shame most speakers in existence.
To my ears, the better of the metal domes (and some composites) will be the most faithful at reproducing the actual midrange and treble part of sound contained in the piano spectrum (including the transients and overtones), aside from the timing issues created by the crossover itself, which is a whole different affair on its own.
There are only a handful of decent AMTs which offer the resolution and low distortion performance of a good metal dome. Most AMTs have far more distortion in the lower mids than decent domes have. The few good ones are made by Mundorf and Beyma. The far biggest problem with AMTs is the lack of vertical dispersion. That in itself is a deal breaker for me.
Be tweeters address the bulk of the issue causing down modulated distortion from breakup by raising it further out of band than other metal diaphragm materials. Its the best compromise between a soft dome and a Be one.
To my ears, the better of the metal domes (and some composites) will be the most faithful at reproducing the actual midrange and treble part of sound contained in the piano spectrum (including the transients and overtones), aside from the timing issues created by the crossover itself, which is a whole different affair on its own.
There are only a handful of decent AMTs which offer the resolution and low distortion performance of a good metal dome. Most AMTs have far more distortion in the lower mids than decent domes have. The few good ones are made by Mundorf and Beyma. The far biggest problem with AMTs is the lack of vertical dispersion. That in itself is a deal breaker for me.
Be tweeters address the bulk of the issue causing down modulated distortion from breakup by raising it further out of band than other metal diaphragm materials. Its the best compromise between a soft dome and a Be one.