Some speaker driver measurements...

This point is very easy explained and experienced - compare a 8kHz square signal with an 8kHz sine signal. Can you hear a difference?
https://onlinetonegenerator.com/

Why? These signals LOOK completely different?
That's the same reason why most people interpret impulse responses wrong 😏


A 8kHz square has it's main wave at 8kHz and the next one at 24kHz - which we can't hear. So they sound identical.
The same with THD - when you have H3 at 8kHz - you can't hear it cause the sound will happen at 24kHz.

You would need a steep 20kHz filter (or 16-18kHz more realistically) to interpret an impulse response right. AFTER the filter the sine and square wave would look exactly the same. Do we use such filters in audio tech ... mhhhh ... 😉
When comparing impulse responses of midrange/woofers you would need to put it after a filter to cut the part of the frequency range you never will use. Some would be surprised how boring these comparisons would get.
Can you show a correct interpretation of an impulse? Makes it more understandable for me what is wrong vs what is right.
 
I compared a notch in series to a Seas DXT tweeter 1:1 non-blinded during measurements and noticed a clear difference with fast sine sweeps up to~48kHz (96kHz sampling frequency) fully exciting the resonance. Without the notch there was a sharp buzzing note in the reverberation after the measurement that disappears completely with the filter applied. Difference was very obvious. Same test with a Bliesma T25A (breakup @34kHz) not so...

1741712523741.png


@IamJF personally I like the DXT very much, very neutral sound if applied/used right.
I have also here the TAC/GB. Spectacular deep and clear transparency, but also somehow artifically coloured a bit. Liked it for a while, switch back now to the DXT because I'm a bit tired of hearing it - maybe this is what you do not like...
 
So, if one have DSP available, one could add an indictor in series with HF driver, calculated to be a first order LP at say 15k (-6dB) and in DSP set a PEQ as; f=15k, Q=0,5, G=6dB and enjoy a decreased level of distortion in the whole pass band? In this case there would be some loss after 15k of a dB or 2... I didn't think that through... (either?) 🙂

//
 
@IamJF personally I like the DXT very much, very neutral sound if applied/used right.
I have also here the TAC/GB. Spectacular deep and clear transparency, but also somehow artifically coloured a bit. Liked it for a while, switch back now to the DXT because I'm a bit tired of hearing it - maybe this is what you do not like...
For me these tweeters show a lot of detail at high frequencies - but not natural detail and they "push it in your face". This can help for mixing when you adapt to it (Genelec) but I never could get used to it and my ears get tired quickly. I owned a speaker with the DXT tweeter (but without notch) - not for me.

A friend of mine built a lot with these Seas tweeters and I always told him about Bliesma. Finally he built something with the T25A and after listening his reaction was "Oh ... I understand now". Now he also uses the T25B in his studio monitors ...

But that's probably too off topic, I will stop my little rant here 🤓
 
In an effort to try to understand how to interprete distortion measurments and impact of a parallel RLC in series with a driver the following measurements with ARTA at 315mm on-axis of the PTT4.0M08-NAC04 mounted in box and at 2.83 Vrms:

The driver without RLC:
PTT4.0M08-R-315mm-OnAxis-DIST-2825mV.png


RLC with 30Ohm:
PTT4.0M08-R-RLC30-315mm-OnAxis-DIST-2825mV.png


RLC with 45Ohm:
PTT4.0M08-R-RLC45-315mm-OnAxis-DIST-2825mV.png


RLC with 60Ohm:
PTT4.0M08-R-RLC60-315mm-OnAxis-DIST-2825mV.png


My interpretation: The cone-breakup has a spl peak at 8.2kHz, and applying an RLC tuned at that frequency gives a reduction at 8.2 , also 8.2/3 and less so at 8.2/5 .
Also that the R value in the RLC has significant effect on the SPL and Distortion reduction at 8.2kHz, not so for Distortion reduction at 8.2/3.

Whether this is audible will be the next test, to give an idea of the active filtering that will be applied:
PTT4.0M08-R-RLC-XO2.png

Note: the phase is not shown, but the filter that will be applied is linear phase.

Interested in other interpretations ;-)
 
Some obvious observations, but still...:

  • DSP filters can't suppress driver own created harmonics (aka distortion) - anything going on in the pass band that generates 2:nd 3rd etc will playout outside the passband.
  • DSP filters can't suppress down stream electronics to generate and playout distortion in the passband.
  • DSP filters can't suppress down stream electronics to generate and playout distortion in the stop band - analog filters will.
  • DSP filters can't suppress the rubbish sent back by a driver to the driving amp - in the stop band, analog filters will (by its increasing impedance towards the amp (?))

  • DSP filters can suppress resonances in the pass band(s).. that trigger distortion downstream and thus reduce playout level of these harmonics.
  • DSP filters can suppress frequencies in the stop band.., that trigger distortion downstream and thus reduce playout level of these harmonics - this was the case above...

Would that be correct and complete?

//
 
The distortion reduction 3th harmonics at 2.6 kHz would that be real at that frequency range?
Yes, that 2.6kHz peak in HD3 would be real acoustic output, but the output is at 8kHz 😉

2.6kHz is on the pass band so gets full input signal as your low pass filter is higher, around 3-4kHz. Obviously you have cone moving at frequencies on the pass band. Also the power amp would be after the DSP. Both the amp and driver would generate some distortion products including harmonics, which extend way beyond the pass band and any that lands on the cone resonance will get amplified further.

3rd harmonic is likely from driver motor and it appears in current through voice coil. Increasing series impedance by using RLC filter or just series inductor reduces the current so reduces acoustic output of this sorta distortion.
 
  • Like
Reactions: jzl
For me these tweeters show a lot of detail at high frequencies - but not natural detail and they "push it in your face". This can help for mixing when you adapt to it (Genelec) but I never could get used to it and my ears get tired quickly. I owned a speaker with the DXT tweeter (but without notch) - not for me.

A friend of mine built a lot with these Seas tweeters and I always told him about Bliesma. Finally he built something with the T25A and after listening his reaction was "Oh ... I understand now". Now he also uses the T25B in his studio monitors ...

But that's probably too off topic, I will stop my little rant here 🤓
Interesting this exchange. I can give my return of experience. I never listen a Bliesma T25... But about the SEAS metal tweeter I use 3 of them 27TBCG < 27DXT < 27TAC . In this order it is about clarity / neutrality. I never heard a difference with a notch above 20khz, perhaps I am death ? Never have listening fatigue. They should be used above 2khz. Note I always tune them by instrument good phase tracking and the response in the range 2kHz-10kHz +-1dB, in a 2dB band. Treble could be fatiguing if you have too much. The 27TAC had the less flat curve after filtering, it has a "hole" of 2dB between 2-5kHz. For the price nothing to complain.
Note a tweeter should be well match with the good woofer to extract the best. For me it is not really about details but fidelity of the sound reproduction...
 
Hard to compare the DXT and TAC/GB independent of the application, because they have different polar patterns - the DXT is a bit special with its waveguide.
Just today I switched back from TAC/GB to DXT with this speaker, xover for both is ~2kHz LR4 (equalized individually, of course):

1741972289488.png


TAC/GB appears more transparent with somehow really spectacular transparency, but DXT sounds more inconspicuous and more neutral and long-term enjoyable for my taste, at least as I was able to apply and tune them within this speaker. (Same series notch sized to ~27kHz as the breakup frequency for both is nearly same).

I have also Bliesma T25A here running in another LS and yes, it's of course the better tweeter (as I would expect for double the price), but not by much - the "micro resolution" is sligtly better compared to both Seas tweeters as it seems to have somehow little more "pixels/bit resolution", hard to describe...

Have a Audax TW025A20Mg on the shelf and I'm really curious if it can match the T25A which I (and others who listened to it) percept as nearly perfect (I'm not interested in overpriced Beryllium or Textreme drivers when such nice drivers can play so for a realistic price...).

After using a Monacor DT-25Ti, older Seas aluminum dome model and then the DXT in the past I had completely lost interest in soft domes which I used mostly before. Will give now the new Scan Speak D3004/602200 a chance (have it also here) as this is the soft dome with the best FR on & off axis behaviour for a reasonable price I could find (distortion is also very nice - Wavecor TW030WA23/24 is on par, can play louder but has more narrow directivity > 10kHz)
 
Last edited:
  • Like
Reactions: IamJF
(I'm not interested in overpriced Beryllium or Textreme drivers when such nice drivers can play so for a realistic price...).
Agree but depend on your project, I replaced a 27TAC by a TW29TXN-8 just to improve the clarity of the details in the high treble. The crossover frequency was 1.5kHz LR4 acoustic.The woofer is the W22NX001. But I think if I have a higher point of crossover 2.5kHz, it will be useless to replace the 27TAC.This tweeter is pure badass.
It is more a usablity rangé of the tweeter, sometime you pay more to have a better motor.
My best treble is reproduce by a mu10rb open baffle with 2 x 27dxt (rear/front) at 3.5kHz@LR2. Not the more neutral sound but very realistic and enjoyable.
 
  • Like
Reactions: ianbo