Purifi PTT1.3T04-HAG-01 33mm Alu dome tweeter review

  What and why am I testing?

My face-to-face encounter with today’s review subject proved to be genuinely long-awaited - both literally and figuratively. I first learned of its “birth” - more precisely, the creation of the first working prototype - as far back as December 12, 2022, during a conversation with one of its “parents,” Lars Risbo.

The news felt almost biblical in scale - reminiscent of the Annunciation to the Virgin Mary by the Holy Spirit heralding the conception of a new life (a messiah?) - the world’s first PTT tweeter.

Since then, several key milestones have marked its journey:

• 18.05.2023 - the tweeter prototype made its public debut as part of a two-way loudspeakers at the Munich High End 2023 show
• 09.05.2024 - the tweeter evolved beyond its 3D-printed components, acquiring fully cast aluminum rear chamber and waveguide assemblies with a meticulously optimized profile and an additional “coherer.” It assumed its final production form and was showcased at the Munich High End 2024 show
• 15.05.2025 - the tweeter subsequently entered the market as a full-fledged production model
• 19.01.2026 - just three years later, the tweeter has finally arrived on my test bench - ready, as it were, to undergo interrogation and reveal the whole truth about itself

So, today we’re getting acquainted with the first tweeter from the Danish company PURIFI featuring a 33 mm aluminum dome - the PTT1.3T04-HAG-01 - which completes the company’s vertical expansion of its product lineup. Their portfolio now includes all the types and sizes of drivers needed to build serious three- and even four-way loudspeakers, all using diaphragms made from the same material.

As you probably already know from my previous reviews (PTT4.0X04-NFA-01, PTT6.5X04-NFA-01, PTT6.5X08-NFA-01, PTT6.5X04-NAA-08, PTT6.5M08-NFA-01A, PTT8.0X04-NAB-02, PTT10.0X04-NAB-02), when it comes to PURIFI products, it’s immediately clear that we’re dealing with something highly sophisticated - a product that carries the latest technical innovations and delivers the highest performance standards.

I must admit that, for me personally and for my measurement setup, testing any of this company’s drivers is a serious challenge when it comes to obtaining accurate, reliable, and repeatable measurements. Sometimes, the distortion levels of their cone drivers fall even below the sensitivity threshold of my setup. Well, let’s see what trials the dome tweeter has in store for me.

I would like to express my sincere gratitude to PURIFI for kindly providing the PTT1.3T04-HAG-01 tweeter for testing, and personally to their team members Lars Risbo and Morten Halvorsen for taking the time to discuss technical solutions and assist with logistical issues.

You can read the history of the PURIFI company here.

  The datasheet

The datasheet is exemplary, offering an exceptionally detailed description of technical parameters and characteristics, supported by a variety of measurements - on-axis and off-axis frequency responses, harmonic distortion versus sound pressure and voice coil current, distortion dependence on signal level, and even intermodulation distortion. The measurement conditions are clearly specified. In addition, the datasheet provides practical recommendations for the tweeter’s use and installation nuances. It’s the best driver datasheet I’ve ever seen! I’m starting to think the team at PURIFI should stop calling their documents datasheets and start calling them “databooks”.

A careful analysis of the datasheet gives every reason to believe that we are dealing with an extraordinary tweeter, featuring an extremely low voice coil inductance of just 0.013 mH, a low resonance frequency of 625 Hz, and, importantly, a very high linear excursion of ±1 mm, along with an exceptionally wide and consistent sound dispersion across a broad frequency range. The combination of low inductance and large excursion hints at a very capable motor and a highly linear moving system - a fact fully confirmed by the accompanying graphs.

Honestly, the graph of current harmonic distortion in Fig.10 particularly amazed me - the harmonic levels fall between -100 and -140 dB. Though, one could just as easily spend endless time admiring the sound-pressure distortion graph in Fig. 2. Excellent work!

   The package and construction

A pair of tweeters arrives in a sturdy glossy corrugated cardboard box, finished on the outside in the company’s signature style. The drivers are secured with cardboard inserts featuring shaped cutouts. Simple, compact, and, most importantly, very reliable.

In hand, the tweeter feels substantial (weighing 660 grams) and solidly built, with proportions that I find very harmonious. Let’s take a closer look at all the visible structural elements:

• The mounting flange, 4 mm thick (measured at 4.2 mm), features six fastening holes along its outer edge, each 3.5 mm in diameter. Around each hole are 2 mm deep countersinks for screw heads up to 6.5 mm. At the center of the flange sits the “coherer” - PURIFI’s own term - consisting of two concentric rings supported by three arched spokes.
  The flange has a complex profile and, together with the coherer, functions as a waveguide. The coherer’s parameters (width of the rings, radial and axial placement) and the waveguide profile were carefully determined through meticulous FEM optimization of their combined performance. A useful side benefit of the coherer is that it protects the diaphragm from accidental mechanical contact.The flange and coherer form a single structural component, cast from aluminum and coated with a dark gray micro-textured matte powder finish, matching the color and texture of other PURIFI drivers. A soft foam rubber gasket is attached around the perimeter on the rear side.The flange has a diameter of 104 mm, with a minimum mounting hole diameter of 92 mm. This leaves a mounting surface just 6 mm wide - narrow enough to present a challenge when using standard threaded inserts for wood panels.
• The mounting flange is attached from the rear with four screws to a black thermosetting plastic ring, which serves as the support for the tweeter’s moving system and the terminal block.
• Immediately adjacent to the black plastic ring sits a large ring-shaped neodymium magnet, 80 mm in diameter and 5 mm thick, coated in black powder paint.
• Directly behind the magnet is a massive cast aluminum rear chamber cover. The cover is completely inert when tapped and features the same dark gray micro-textured powder coating as the mounting flange.
  It has a complex, non-axially symmetrical shape. At its center is a deep conical recess, likely intended to diffuse the internal sound wave emanating from the rear of the diaphragm.On the side of the cover opposite the terminal block is a rectangular glossy plastic label with a barcode and the tweeter model. The rear features a ring-shaped micro-textured plastic label with the company logos. Above the “i” in “Made in Denmark” there is a small dimple - the outlet of a thin drainage channel designed to equalize pressure inside and outside the tweeter.
• Behind the coherer, the dome-shaped diaphragm is visible - convex, silver-white, 33 mm in diameter, 8 mm tall, and made of deeply anodized aluminum.
• The diaphragm’s surround is extremely hard to see, as it is covered by the coherer and matches the flange in color and texture. Nevertheless, I determined that it is very narrow, concave in shape with an approximately triangular profile, and made not of fabric (as in most tweeters) but of a very soft polymer. This unusual combination of geometry and material is likely responsible for the feature PURIFI calls “Negligible Surround Radiation Distortion”.
• The terminal block looks solid and matches the driver’s level of build quality. It is made from thermosetting plastic with gold-plated spade terminals pressed in.
• Using a magnifying glass, one can see the ends of the lead wires emerging from the plastic ring and soldered to the terminals. The lead wire consists of a tightly twisted bundle of fine wires, forming a miniature spiral - a construction I had not previously encountered.
• The build quality is exceptional - absolutely flawless. No glue spills, scratches, dirt, stains, chips, gaps, or misalignments. Perfect!

So, we’ve examined all the visible structural elements, which leave no doubt that we’re dealing with a serious and carefully thought-out design. Unfortunately, there is no available information on the tweeter’s internal construction, though knowing PURIFI, I’m certain it’s far from simple.

To shed some light on the inner workings of this still “black box”, I reached out to Lars Risbo, one of the co-owners and developers at PURIFI Transducer Technology, and he kindly indulged my curiosity. Here are a few excerpts from our conversation:

Wow! These comments clarify a lot, and my suspicions have been completely confirmed. The design is indeed thoroughly thought out - not a single detail has been overlooked. Features such as two copper and one aluminum ring, two folded transmission lines in the rear chamber, a new type of lead wires, and an N55 neodymium magnet (currently the most powerful commercially available magnet!) are things I have yet to encounter in any other tweeter!

  Impedance frequency response

The diagrams below show the frequency response of the impedance magnitude at various scales:

The measured resonance frequency was Fs = 676.3 Hz, slightly above the specified 625 Hz. For a tweeter that will be used well above its resonance frequency, this difference is of no practical significance.

The shape of the resonance peak is asymmetrical, which, as you may have inferred from Lars Risbo’s comments, is due to the operation of the coiled transmission line in the rear chamber.

Thanks to the extremely low voice coil inductance (0.013 mH), the impedance rise with increasing frequency is minimal, with the ratio of the impedance at 20 kHz to the minimum impedance at 4 kHz being just 1.15. This is achieved through the use of two massive copper rings in the motor, and it positively affects the reduction of nonlinear distortion.

The impedance curve is very smooth across the entire operating frequency range and perfectly matches the datasheet. Even when heavily zoomed in, it remains exceptionally “clean,” with no signs of parasitic resonances in the moving system.

The impedance characteristics indicate a well-engineered motor and a perfectly balanced moving system.

  On-axis frequency response (at 315 mm)

Below is the smoothed axial frequency response of the tweeter, measured in a test baffle at a distance of 315 mm from the microphone:

The tweeter’s frequency response and the measured sensitivity at 3.5 kHz (95.5 dB) perfectly match the datasheet.

The waveguide adds a few decibels of acoustic gain, resulting in a slight lift in the frequency response over several octaves, peaking around 3.5 kHz. This makes the response curve slightly “humped,” which makes it somewhat tricky to define the average sensitivity numerically - though, is it really necessary? If we evaluate the real sensitivity based on input power rather than voltage, considering the low DC resistance of the voice coil at 3.5 Ω and the slightly misleading effect of the waveguide, I would tentatively rate the tweeter’s sensitivity as average to above average.

The frequency response is smooth across the entire operating range, and the hump at 3.5 kHz caused by the waveguide can be easily corrected even in a passive crossover.

At 14 kHz, there is a small “bump” of about 1 dB, a characteristic also reflected in the datasheet. Its origin is difficult to explain, but I am confident it will be of no practical significance to anyone.

Starting at 15 kHz, the frequency response gradually rises, reaching a maximum boost of around 15 dB at the diaphragm’s primary resonance at 27 kHz. Beyond that, the diaphragm operates in a break-up mode, yet well-controlled mode, with the response holding at a fairly constant 92 dB up to 43 kHz.

Upon first measuring the on-axis response of the PTT1.3T04-HAG-01 tweeter, I immediately thought of the Scan-Speak D2908/7140 beryllium dome tweeter, which exhibits a very similar frequency response and also uses a small waveguide. The only difference lies in the frequency and amplitude of the diaphragm’s main resonance.

Below are the smoothed on-axis frequency responses of the two tweeters from the matched pair, measured in an enclosure at a distance of 315 mm from the microphone:

The frequency responses of both units are perfectly matched across the entire range up to 14 kHz. Not bad for tweeters that are not specifically matched in pairs at the factory.

Beyond that, they diverge slightly, but the maximum difference nowhere exceeds 2.4 dB. I consider this an indicator of good consistency.

  Off-axis frequency responses (at 315 mm)

Below are diagrams of off-axis frequency responses - conventional and normalized, in which the axial response is taken as a reference, and the off-axis ones reflect only the difference with it:

All tweeters lose output with increasing frequency and off-axis angle, but what truly matters is how they do it. In most cases, tweeters without waveguides exhibit a two-knee type off-axis response: first, the curves remain bundled together; then, at the first bend (the first “knee”), they begin to fan out at varying rates up to a certain frequency (the second “knee”), after which they diverge much more rapidly. Examples include the VIFA NE25VTS-04, Accuton C25-6-158, Audax TW034X.

Tweeters equipped with various types of waveguides or horns also typically show either a two-knee pattern (Satori TW29TXN, Scan-Speak D2908/7140, Morel CAT378) or a single-knee behavior (SEAS T29X001, Wavecor TW030WA11).

Such behavior leads to a monotonic increase in the directivity index (DI) with frequency, with the rate of increase changing at the knee points. In other words, the radiated sound power decreases monotonically with frequency, which can disturb tonal balance, since the perceived balance at the listening position is largely determined by the spectrum of the diffuse sound, that is, the spectrum of the radiated sound power.

The off-axis responses of the PTT1.3T04-HAG-01 remain tightly grouped up to 1.3 kHz, after which they split and begin to decrease monotonically with frequency at nearly the same rate across all angles, all the way up to the diaphragm’s primary resonance at 27 kHz. The normalized off-axis curves show the first knee at 1.3 kHz, where the responses begin to fan out, and by 4 kHz they reach a plateau extending all the way to 15 kHz.

Within this region, the curves run almost perfectly parallel, their relative levels remain unchanged. Throughout the plateau, the directivity index remains constant, hence the term Constant Directivity. The radiated sound power spectrum, and therefore the tonal balance at the listening position, remains stable. This behavior is the result of the optimally engineered waveguide and coherer design.

From approximately 15 kHz onward, the waveguide gradually loses control over directivity, the second knee appears, and the responses begin to diverge. The range above 15 kHz is of limited significance for high-quality sound reproduction.

Beyond its constant directivity, the tweeter also exhibits impressively wide dispersion up to the upper limit of the audible band. The manufacturer specifies an angular coverage of 140 degrees at –6 dB across a broad frequency range. That’s remarkably impressive for a tweeter with such a large diaphragm and such compact overall dimensions!

  Harmonic distortion (at 315 mm)

Shown above are the harmonic distortion plots measured at 2.83 V and 11.2 V, corresponding to sound pressure levels of 95 dB and 107 dB at 3.5 kHz, respectively. The measurements were performed on-axis at a distance of 315 mm from the tweeter to the measurement microphone.

To prevent power overload and excessive diaphragm excursion during distortion measurements, a second-order analog passive high-pass filter with a cutoff frequency of 800 Hz was used. Therefore, in these graphs we analyze the frequency range from 800 Hz and above. The use of a passive analog filter instead of the more commonly employed digital one was necessary to reduce the overall nonlinear distortion of the measurement setup itself.

At all volume levels and across the entire frequency range of interest, the second harmonic, the most euphonical in character, clearly dominates. As expected, it increases proportionally with applied voltage. The second-harmonic curve is fairly flat and smooth, without sharp peaks or dips. Only above 14 kHz does it begin to rise, due to the influence of the diaphragm’s main resonance. Even at 11.2 V, the second harmonic does not exceed –40 dB at 900 Hz. Impressive!

The third, fourth, and fifth harmonics above 3 kHz remain at extremely low levels even at 11.2 V and at 2.83 V they are barely worth mentioning. As for the slight rise of the fourth and fifth harmonics above 5 kHz, I have strong reasons to suspect that this is caused not by the tweeter, but by my measurement setup. What I had feared before testing has indeed happened - the bottleneck in distortion measurements turns out to be my own setup!

Below 3 kHz, the higher-order harmonics increase rapidly as frequency decreases, which is entirely typical and expected. This is related to the rapid increase in diaphragm excursion, which is inversely proportional to the square of frequency. Motor and suspension nonlinearities also increase and begin approaching their limits. For example, at 2.83 V and 1 kHz, the diaphragm excursion of the tested tweeter is 0.145 mm - not insignificant compared to the maximum linear excursion of 1 mm. At 11.2 V, the excursion already reaches 0.87 mm! At 3 kHz, these figures would be roughly nine times smaller.

Yet even with this rise, distortion levels below 2 kHz remain lower than those of any tweeter I am familiar with.

What more can be said? The numbers and graphs speak for themselves. Across the entire frequency range, not just for a particular harmonic at a specific frequency, the PTT1.3T04-HAG-01 is, to date, the lowest-distortion tweeter I have ever measured. Bravo!

  Voice coil current harmonic distortion

Despite its simplicity, this type of measurement is a very useful tool for assessing the linearity of a driver’s motor. The diagrams above show the harmonic distortion measured at 2.83 V and 11.2 V, corresponding to sound pressure levels of 95 dB and 107 dB at 3.5 kHz, respectively.

To prevent power overload and excessive diaphragm excursion during testing, a second-order analog passive high-pass filter with a cutoff frequency of 800 Hz was used. Therefore, in these graphs we analyze only the frequency range from 800 Hz and above. The use of a passive analog filter instead of the more typical digital one was necessary to reduce the overall nonlinear distortion of the measurement setup itself. I should note that, for the first time, I extended the lower limit of the distortion scale down to –120 dB, rather than the usual –100 dB!

Voice coil current nonlinearity directly reflects the nonlinearity of the mechanical force driving the diaphragm, since this force is related to current by the simple relationship F=B×L×I, where B is the magnetic flux density, L is the length of the voice coil wire within the magnetic gap, and I is the current. Therefore, in the frequency range where the contribution of moving-system nonlinearities becomes negligible, it is practically impossible to achieve sound pressure distortion levels lower than the current distortion.

The second harmonic dominates throughout, but its level in the current is significantly lower than in the sound pressure. This indicates that the motor’s linearity exceeds that of the suspension.

The behavior of the remaining current harmonics mirrors what was observed earlier in the sound pressure measurements. Once again, I encountered the limitations of my setup. The measured fourth and fifth harmonics reached the “floor” of my measurement system at –110 dB. This becomes evident when comparing my 2.83 V distortion plot with the manufacturer’s corresponding measurements (Fig.10 in the datasheet). In their data, the fifth harmonic in the 2–5 kHz range sits at around –120 dB, while the fourth harmonic reaches –130 to –140 dB!

My own experience confirms that there is no reason to question the credibility of PURIFI’s measurements.

I would describe the level of current harmonics of all orders as extraordinarily low for a dome tweeter. I have never encountered anything like this before!

  Step response

The step response is essentially the mirror image of the frequency response, from which it is derived through a unique mathematical transformation. It is one of the standard methods for analyzing linear systems, allowing the system’s response to external excitation to be examined in the time domain rather than the frequency domain.

In this case, we observe a typical tweeter step response. The rise time is very fast and is determined by the tweeter’s upper cutoff frequency, which for the PTT1.3T04-HAG-01 exceeds 40 kHz.

The system returns to rest with slight oscillation and a small negative overshoot. This behavior is characteristic of tweeters exhibiting a low-frequency roll-off described by a second-order polynomial with relatively low pole Q. In simpler terms, this is typical of tweeters with a gentle slope of frequency response around the resonance frequency Fs. Compare, for example, with the step response of a tweeter with a steeper frequency response around Fs - Scan-Speak R3004/602000.

The oscillatory behavior visible on the decaying portion of the curve is caused by the diaphragm’s main resonance at 27 kHz. Once suppressed by a notch filter in the crossover, this ringing will leave no audible trace whatsoever.

  Waterfall

The waterfall plot is another tool for linear system analysis. It illustrates how the frequency spectrum of the response decays over time and often helps reveal hidden resonances that may be difficult to detect using other types of measurements:

Within the audible frequency range, the waterfall decays very quickly - after the second time slice, nothing remains. No hidden parasitic resonances were detected. The slight irregularity in the 2–4 kHz range is a measurement artifact and is not related to the tweeter’s actual behavior.

In the ultrasonic range, we observe a long decay tail from the diaphragm’s main resonance at 27 kHz and a shorter one around 40 kHz, both decaying to an acceptable level within approximately 2.5–3 ms. There is no reason to be concerned about the long tail, as it lies beyond the limits of human hearing. Purists may suppress it with a notch filter, as recommended in the datasheet, after which it will leave no visible trace in the waterfall plot.

  Listening impressions

As usual, after completing the objective testing program, I moved on to subjective evaluation. The tweeter was first listened to simply in-hand, directly without filtering, and then through a second-order digital high-pass filter with crossover frequencies from 1 to 3 kHz. I carefully studied its sonic signature.

After that, still listening in-hand, I compared it with several top-tier tweeters from other manufacturers that I have available. Finally, the tweeter was tested as part of a miniature two-way loudspeaker system based on the PTT4.0X04-NFA-01 midwoofer. These cabinets have hosted a great many tweeters over time! I clearly remember every possible variation in their sound signature.

Unfortunately, due to the very large required mounting cutout diameter, the tweeter could not be installed in any of the existing enclosures. Therefore, it had to be mounted on top of the cabinet. Everything was measured, simulated, and integrated using the universal  MatriXover-II crossover.

Several days of intensive listening and meticulous analysis led to the formation of clear and stable impressions, which can be expressed through the following psychoacoustic attributes and one question:

• Tactility, Physical presence, Relief
• Instrument Separation
• Depth
• Softness and Velvet Smoothness
• Juiciness of the High Frequencies
• Airiness
• Neutrality
• Effortlessness
• Is this really aluminum?
   

In my view, these attributes dominate over the others and best characterize the sound of the PTT1.3T04-HAG-01. Let’s examine them in more detail.

Tactility, Physical presence, Relief — yes, this is the first thing that comes to mind during the initial listening sessions. The PTT1.3T04-HAG-01 is the kind of tweeter that adds flesh to flat skeletons and two-dimensional projections of performers and musical instruments. They become palpable, three-dimensional, holographic. You can not only hear them — you can almost see them.

Instrument Separation is outstanding and is one of the tweeter’s strongest qualities. Instruments literally spread apart in space, with air forming between them.

Depth - the soundstage gains a tangible sense of perspective and subjectively expands both in width and depth. When listening to a single tweeter in-hand, depth is difficult to assess, though hints are present. In a stereo loudspeaker setup, however, everything becomes immediately clear.

Softness and Velvet Smoothness — a touch of honey in vocals never hurts. This is not about adding sugary silk coloration, not at all. The natural timbre is fully preserved, yet even the most shouty, piercing, harsh, or intrusive voices become softer, rounder, and more pleasant. Brass and strings benefit greatly from this velvet touch, while metallic brass instruments retain their bite and solidity, which is no small achievement.

Juiciness of the High Frequencies — any metallic percussion sounds not dull or flat, but juicy, crystalline, and sculpted, as if floating in a cloud of sound as distinct sonic objects you almost want to reach out and touch.

Airiness — thanks to its wide dispersion pattern, there is plenty of “air” in the top octave. It does not protrude forward or burn like a laser beam, but evenly fills the entire sound field in both width and depth, lending the presentation a realistic liveliness.

Neutrality — the tonal balance is not shifted toward either the cold or warm side, and there is no pronounced coloration. No veil, no greyness, no excessive richness, sweetness, or dryness.

Effortlessness — even when listening in-hand, even without any filtering (i.e., directly connected), and even at high volume levels, there is a sense of relaxed, free, and unconstrained performance at lower frequencies, without shrillness, harshness, or chuffing (as sometimes happens with tweeters featuring wide surrounds). I believe this impression is due to the low resonance frequency and large linear excursion, reinforced by low distortion.

Is this really aluminum? — yes, this question kept coming to mind. I am very familiar with the typical sonic character of aluminum tweeters and have never considered myself a fan of them. I thought I could identify them blindfolded in any setup. But with the PTT1.3T04-HAG-01, things are different. I even caught myself thinking that if someone offered me a wager to identify the diaphragm material in a blind test, I would probably decline. If any of the typical “aluminum” traits are present (increased brightness, dryness, timbral leanness), they are present only to a minimal degree. It seems to me that the softness, velvet smoothness, and tactility largely mask its family “aluminum” traits.

If asked about Resolution (timbral), Dynamics, Speed, and Density of the PTT1.3T04-HAG-01, I would answer the same way Rolls-Royce responds when asked about the horsepower of their engines - “adequate”. That means the tweeter is not a record-breaker in any of these parameters, but they are at a high level and certainly will not become a weak link or bottleneck in the overall performance. There are no absolutely perfect tweeters that simultaneously possess the highest possible scores in every sonic attribute, at least, I have yet to encounter one. A reasonable compromise is always essential.

Tweeters with a “rich” or “full-bodied” character, and I would place the PTT1.3T04-HAG-01 closer to that category, are never ultra-fast or ultra-dense. By “rich,” I mean the previously described traits: low-reaching capability, tactility, relief, softness, and velvet smoothness. This is not a complete list of defining attributes, but it is sufficient in this context.

I can reassure everyone who is very concerned about the significance and audibility of the 27 kHz resonance - I have not been able to detect any parasitic influence from it. Of course, my hearing is not as good as a bat's, perhaps yours is better. Therefore, the need for a 27 kHz filter plug is irrelevant for me.

The tweeter is absolutely genre-agnostic. I detected no genre preference whatsoever. It plays everything!

  "How to use" recommendations

Based on the very low harmonic distortion and its frequency profile at the bottom of the operating bandwidth, I would recommend using the PTT1.3T04-HAG-01 tweeter in high-end multi-way loudspeakers with a crossover frequency starting from 1.5 kHz. Use at lower crossover points is also possible, but will depend on the required maximum SPL and your tolerance for nonlinear distortion.

Any “electrical” filter slopes are possible, even first-order slopes at the lowest crossover frequencies - the tweeter is fully capable of handling such implementation.

Thanks to the waveguide controlling the dispersion pattern, the tweeter can be successfully integrated even with relatively large midwoofers or midrange drivers up to 8 inches in diameter.

The tweeter mounting surface is only 6 mm wide - narrow enough to create difficulties using standard threaded inserts for wood panels. A solution is to use a stepped insert hole: the first mounting plane is 92 mm in diameter at a level of -4.5 mm, while the second is 81 mm in diameter and a level of -13 mm from a frontal plane. This leaves more material for the screws and thread  inserts.

I can recommend the following pairings for building stunning modern loudspeakers with very clean, transparent, dense, and dynamic sound while maintaining relatively compact dimensions:

Two-way "WT" configuration with aluminum midwoofer:
TT6.5X04(08)-NAA-08
PTT1.3T04-HAG-01

Two-way "WT" configuration with paper midwoofer:
PTT6.5X04(08)-NFA-06
PTT1.3T04-HAG-01

Two-way "WTW" configuration with aluminum midwoofers:
PTT6.5X08-NAA-08
PTT1.3T04-HAG-01

Two-way "WTW" configuration with paper midwoofers:
PTT6.5X08-NFA-06
PTT1.3T04-HAG-01

Three-way "WMT" configuration with aluminum woofer and midrange:
PTT10.0X04-NAB-01
PTT6.5M08-NAA-08
PTT1.3T04-HAG-01

Three-way "WMT" configuration with aluminum woofer and paper midrange:
PTT10.0X04-NAB-01
PTT6.5M08-NFA-01
PTT1.3T04-HAG-01

The three-way configurations listed above are capable of filling a room up to 40 m² with full-range sound, reaching down to 26 Hz, and delivering ultimate audio quality.

  What is the price and where to purchase it?

At the time of this review’s publication, the PURIFI PTT1.3T04‑HAG‑01 tweeter can be purchased directly from the manufacturer or from the following online retailers, with prices starting from around €504 per unit (ex‑VAT):

• https://www.madisoundspeakerstore.com/
• https://www.soundimports.eu/
   

  Summary

Testing has proven that the saying “You must spoil before you spin” is completely inapplicable to PURIFI’s first tweeter. On the contrary, as always, we are presented with a very mature product featuring a meticulously designed construction, unique technical solutions, impressive measurements, and excellent sound - a tweeter fully engineered from a clean sheet. All measured characteristics and parameters perfectly match the datasheet. Consistency between the two tweeters is good. Bravo, PURIFI!

Now, when it comes to building a loudspeaker system entirely with PURIFI drivers, all using diaphragms made from the same material, there are no obstacles! In my experience, both factors contribute to achieving a coherent and homogeneous sound. I would very much like to build and hear such a system myself, and I hope that in the future I will have the opportunity to make it a reality.

What I Liked:

• Very large linear excursion ±1 mm
• Smooth frequency response
• Low resonance frequency
• Wide radiation dispersion with constant directivity
• Extremely low harmonic distortion in sound pressure
• Extremely low harmonic distortion in current
• Usable down to 1.5 kHz or even lower
• Excellent sound quality
• Outstanding build quality
   

What I Didn’t Like:

• A rare case where there's nothing to complain about

 

More extended measurements can be found here

Yevgeniy Kozhushko/18.02.2026