So e.g. Purifi is on wrong track for reducing distortions?
By the way, what type distortion are you talking about? There are many, like frequency response and polar response irregularities, HD, IMD, resonances, compression, these are all distortions and possibly with different roots. So saying that a lower Mms combined with more motor force leads to lower distortion does not seem to be a well-established rule.
I think it's not hard to see that the story is not that simple.
Last edited:
I don't think Purifi is wrong, they look at other aspects of distortion as they think it's more important. The BL does not have to be low, it has to be constant over the whole excursion of the coil (no drop in BL when reaching xmax). Idem with the cone, you need a stiff and damped surround, not a heavy cone. And their cones are not heavy or the magnets weak, their cones are relative light and their magnets are not with the highest BL for their size, but are strong (and mainly well balanced).
I peronally think distortion has a lot of origins, and both methods help to reduce it. But speakers are alwasys a compromise and a balancing excercise to find the best setup. Focussing on one factor alone does not result in a good sound, not with speakers and not with speaker drivers.
I peronally think distortion has a lot of origins, and both methods help to reduce it. But speakers are alwasys a compromise and a balancing excercise to find the best setup. Focussing on one factor alone does not result in a good sound, not with speakers and not with speaker drivers.
What is Purifi angle?
I didn't think the statement about motor to mass ratio lined up with the Data anyway. You have to be careful how you take these words of the AI. Thats why I said take with a grain of salt, it takes a lot of work to fix the AI's logic at times, I even struggled to keep the Data identical in order to create these graphs. So the AI got the distortion interpretation backwards, so disregard its comments about lighter cones and distortion.
I'll attempt to get some comments from it again, remember, take with a grain of salt, lol This is content created by the data from the charts, or at least its supposed to be. I think The charts are accurate but I am having issues with getting AI to interpret the data, not that we can't do it ourselves.
Purifi doesn't seem to prioritize low moving mass at all.
Purifi 6,5" paper cone midwoofer Mms = 24,4g
SEAS Excel 6,5" paper cone midwoofer Mms = 13,7g
Purifi 10" metal cone woofer Mms = 118g
Scan-Speak Revelator 10" metal cone woofer Mms= 51g
A heavy cone would dampen anything that was an active signal..... Actually a increasing the weight of the cones, dampens all signals.... The thing is, we can increase the signal of the source to compensate. So as long as Qts is low, how is a heavy cone not the better choice? If Damping effect of increasing mass is linear or linear enough, then the heavier cone with low Qts is the winner.
Similar to lighter cones or middle weight cones. Low (or more likely low enough) Qts is always desirable, no matter it's low or high Mms.
The kinetic energy probably has a "breaking point" despite BL and cone-material , where one cannot avoid major negative effects-
And i think Purifi kept their Mms under that "breaking point"
Agreed. Mass provides Resistance to non active signals. I think the interesting thing is that Mms is not linear. Some aspects of Mass is linear though.
at 2mm or less, whats the difference between 2 identical loudspeakers with the variation that A has low Mass vs B that has high Mass.... but they have Identical 0.2 Qts. I meant to say that "A heavy cone would dampen anything that was Not an active signal" and it dampens active signal as well, but active signal can be increased to compensate.
Last edited:
How do you know that this isn't the removal of excess decay in the HF of the driver?
Right and wrong are 'moral' concepts > don't exist in nature.
Put simply, in nature everything is about alignment (balance, equilibrium, harmony etc.).
The difference with our linear perception and way of thinking (programming, wiring) lies in the cyclical, complex non-linear dynamics outside our '(observable) reality'.
This can be illustrated with an example related to audio (frequency, vibration), the work of the Swiss physicist Hans Jenny:
From a recent paper titled 'Interface compatibility and hysteresis in shape memory materials are affected by lattice distortions from applied stresses':
"EM Feygin · 2025 — Significant effort has been put into designing shape-memory materials that can survive many cycles without functional or structural fatigue."
This creates a bridge to Purifi.
Last year I had the pleasure of attending a demo of the Kii SEVEN in a private setting, with Bruno present.
Afterwards there was ample opportunity to ask questions.
Bruno clearly exhibits the characteristics of a genius. He speaks softly but firmly.
He explained, among other things, his modus operandi, in which he puts certain things on paper in advance (prior to the development of algorithms, running sims etc.).
We discussed a number of topics, including hysteresis.
Unlike the majority of the disciples of the SINAD cult (on that other forum), Bruno knows what really matters.
He also explained why he no longer posts on such forums.
Although Bruno himself works according to fairly rigid mind/science methodologies, he does have an eye/ear and appreciation for other approaches.
This is an excerpt from an article on high-end audio in the largest Flemish business newspaper:
Audiophilia nervosa
"The concept of high-end audio is indeed vague," says Brussels audio engineer Bruno Putzeys. 'A speaker costing 2,000 euros can completely play a speaker costing 20,000 euros off the shelves.' Putzeys is mainly known in the Netherlands and the US. After securing a number of patents at Philips, Grimm Audio and Hypex, he co-founded Kii Audio, with headquarters in Apeldoorn. Kii does not offer absolute high-end, but complete systems for 10,000 euros. But qualitatively they are strong. And they sell about 500 speakers a year.
"A small manufacturer usually makes the same thing for a higher price, partly because he purchases much more expensively," he says. 'The customer often only looks at the price, but usually the difference is not in the sound, but in a very beautiful, fine finish, like a piece of furniture.'
'A rare exception is Wim Vanderstraeten. He doesn't work scientifically at all, but he has so much patience and such good ears that he offers excellent value for money," says Putzeys. Vanderstraeten has been building speakers for 35 years, mainly in the price range of 2,000 to 5,000 euros. He is popular with professionals. A number of classical musicians love it.
Brain tricked
Loudspeakers costing 5,000, 10,000, 50,000 or 100,000 euros: can you hear the difference? "I find fifty thousand euros very difficult to justify," says Putzeys. 'At the English-South African Vivid Audio there is so much engineering that there is something to be said for it. But most speakers in that segment disappoint me. Loudspeakers have been built for over a century without major revolutions. The most important evolution is the active loudspeaker. It has a built-in amplifier, so you can connect it directly to your CD player, TV, laptop, etc. But actually it has also been around since the 1930s. A loudspeaker in no way reproduces the real sound of a concert. The wave movement in the air doesn't even resemble that in the original setting. We trick our brains all the time. All philosophies about the ideal speaker miss that most important link: the brain. It remains a psycho-acoustic illusion with limits. You're never really convinced that the musician is in the room. But we can minimize the 'suspense of disbelief', among other things by displaying all pitches with the same sensitivity.'
In any case, Putzeys has the impression that he is already coming quite close to the ultimate sound. 'At the same time, I know that high-end involves a lot of voodoo nonsense, alternative medicine that has no benefit and has ruined the entire market. For example, cables costing 10,000 euros exist. It makes you think: how is it possible?'
We want to know if he still hears the music himself. Can he still get carried away? 'That works, but I have to turn my switch: disconnect from technical listening and sink into the music. The best way to do that is to just listen to a complete album. That's complicated, by the way, with the vinyl revival: you don't have a remote control. I force myself into it'.
When my brother asked whether he also appreciates tube amps, Bruno replied: "Yes, some of these clearly reflect the results of a long-term empirical optimization process".
Last edited:
Does not know exactly what you mean by that.
All i have is all my listening and testing of speakers and different drivers last 25-30 years.
It was my opinion, does not need to be others.
I have never heard a heavy cone/125+ gram play good midrange (400-1200 hz, and my 15 inch JBL 2215h papercone with Mms at 98 gram is the best quite " heavy" driver ive heard up to around 8-900 hz.
But again my opinion
At sub-sub-sub (etc.) domain levels, things become more and more complicated.
We can focus on 'Klippel' articles, but I don't find it useful in this regard.
As Bruno rightfully pointed out, the basics haven't changed since the 1930s.
An original Lansing Iconic beats many, if not most, modern high-end speakers in creating (a believable) psychoacoustic illusion.
A fast driver needs a light cone. Or does it? | PURIFI
Description
This 6.5" is a member of the ultra low distortion woofer family from PURIFI. Compared to the renowned PTT6.5X family, the W-version has higher sensitivity and is well suited for applications that doesn't require extended excursion.
The driver incorporates the easily recognizable PURIFI NeutralSurround suspension and ultra-linear motor system PURIFI PureDrive technology working in concert to elevate performance to a whole new level despite of it’s compact size.
The PTT6.5W04-NFA-01 features:
Vibroacoustic optimized cone and dust cap in proprietary paper fiber mix
Negligible Force Factor Modulation (FFM) and Surround Radiation Distortion (SRD)
Ultra Low Magnetic Hysteresis Distortion (MHD)
Best mix of volume displacement and midrange efficiency
Uncompromised midrange performance
Designed and manufactured in Denmark
Electrical & Acoustical Parameters
DC resistance 3.7Ω
Minimum impedance above resonance, Zmin 4.15Ω
Frequency for minimum impedance, fmin 310Hz
Maximum impedance, Zo 88Ω
Voice Coil inductance @ 1kHz 0mm, Le 0.32mH
SPL @ 2.83Vrms/1m, 300-800Hz 90.3dB
SPL @ 1W/1m, 300-800Hz 87.5dB
Thiele/Small Parameters
Resonance frequency, fs 33Hz
Mechanical Q factor, Qms 6.3
Electrical Q factor, Qes 0.27
Total Q factor, Qts 0.26
Equivalent volume, Vas 28.9L
Effective piston area, Sd 132.7cm2
Effective piston diameter, D 13.0cm
Force factor, Bl 7.7N/A
Mechanical resistance, Rms 0.70kg/s
Moving mass, Mms 21.1g
Suspension compliance, Cms 1.10mm/N
Typical Box Tunings
Typical corner frequency, sealed B2 alignment 89.7Hz
Box volume, sealed B2 alignment 4.5L
Typical corner frequency, bass reflex PTT alignment 48.4Hz
Box resonance, bass reflex PTT alignment 50.7Hz
Box volume, bass reflex PTT alignment 12.7L
Port diameter, bass reflex PTT alignment 5.6cm
Port length, bass reflex PTT alignment 22.9cm
Typical corner frequency, 2xPR PTT alignment 51.0Hz
Box volume, 2xPR PTT alignment 11.0L
PR mass per PR, 2xPR PTT alignment 51.5g
Now let's Klippel...
From Erin's review:
Bottom Line
T/S Parameters and Linear Excursion:- My measured linear excursion is limited to about 6.9mm one-way. This is based on 10% distortion limits. This jives with the spec provided if one were to take the graphics and apply the displacement limits (75% for Cms). The product spec sheet states a linear xmax of 10mm one-way linear likely defined by the physical overhang of the voice coil [VC length (23.7mm) - gap height (4mm)].
- This is good linear excursion for a speaker this size. However, the Scan Illuminator 18wu I tested in October 2011 still takes the cake in the 6-7 inch midwoofer linearity category with a Klippel verified 9.1mm one-way excursion (also limited by compliance).
- My specs fall well within line of the manufacturer’s spec. Not much else to add here.
Frequency Response:
- The average sensitivity is measured at about 88.1dB from 300Hz to 1000Hz.
- On-axis response linearity is ±1.5dB within 100-2kHz with some mild bumps here and there.
- On-axis response linearity is ±3.0dB within 66-2.5kHz.
- Breakup is kept to a very respectable +4dB above the mean SPL at 3kHz. This is quite good.
- Off-axis response shows nice linearity until about 4kHz (evidenced by the 60° measurement showing a +4dB resonance at ~3.9kHz).
Distortion and Compression:
- This is the best speaker I have tested (I don’t have the same distortion data for the Scan 18wu I mentioned above). An incredibly low 1% THD above 70Hz at 100dB output!
- The dominant mode of distortion is 2nd-order. But, really, it is so low… who cares what order is the contributor. 3rd-order distortion is 10-20dB lower than 2nd-order.
- The maximum SPL tests are limited by the -2dB compression threshold; namely by the 500Hz region. Looking back at the HD and IMD measurements, there appears to be something in this region causing higher 3rd-order distortion HD. I’m not seeing traces of it in the impedance, though, which is the first place I normally look for problems “in band” (as opposed to out of the typical bandpass region).
It's undoubtedly a superior midwoofer, notwithstanding the undesireable 3rd-order distortion around 500 Hz.
For comparison, THD plots of 2 highly regarded midwoofers from the recent past that are still available in some shops for roughly 1/7th of the Purifi's retail price:
SICA 6H1.5CP-8:
PEERLESS HDS-P830869:
I know of a few midwoofers that also show increasing third order harmonics, only around 200-300 Hz, which is less bothersome imo.
The fact is that those drivers are up to 18x cheaper than the Purifi.
Why do I find this borderline naive and preposterous? He doesn't solve the most basic microphone/recording problems. (As benchmark, is same-person binaural recording/playback SOTA already near perfect?)
So the "ultimate sound" will remain a "in no way real" reproduction, an "acoustic illusion".... Then what's the point of his life-long pursuit?
Well personally, I think reproduced sound/music has always been "realistic" enough, since the "beginning", for the intended purposes. (Imitating another animal's call, for example.)
Last edited:
The cone of the Purifi is actually relatively heavy:
- Force factor, Bl 7.7N/A
- Moving mass, Mms 21.1g
SB ACOUSTICS SB17MFC35-8 (measured):
- 6.27
- 12g
Sica 6,5 H 1,5 CP:
- 7.02
- 13g
Last edited:
There's a marketing aspect to this.
If you only focus on small desktop monitor-style speakers and hi-fi drivers, you won't get lifelike reproduction.
Even though a compact monitor with Purifi drivers may sound 'bigger' than it actually is, the experience is (still) completely different from a large
101 dB 2-way system.
There are no substitutes for cubic inches and efficiency.
An ultimate torture test is the reproduction of large (and heavy) orchestral works.
"You're never really convinced that the musician is in the room. But we can minimize the 'suspense of disbelief', among other things by displaying (~reproducing) all pitches with the same sensitivity."
Personally, I've experienced a convincing 'band playing in the room' illusion a few times.
"You're never really convinced that the musician is in the room. But we can minimize the 'suspense of disbelief', among other things by displaying (~reproducing) all pitches with the same sensitivity."
Personally, I've experienced a convincing 'band playing in the room' illusion a few times.
So all the distortion datas Klippel shows us is not that distortion (or lack of) what you are talking about? We still can't measure why a big ol' speaker still better than a nowadays small speaker? Simply by the old ones makes more dB/W with light (low Mms) drivers and utilizes larger boxes? That's all we need then for a lifelike sound?