Natural Sound NS17 Review (Speaker)

[jerryzhang]

Sure, I will be happy to test that one. Just click on my name and start a conversation when ready and I can provide address information.

Thank you for being here and answering questions.

Thank you so much for your reply! But I still to my oversea distributor first because oversea logistics too complicated to me. I will prepare it asap.
Hope you like it! The report of NS15 as below:
This is FR curve：

This is Harmonic distortion curve ( I think it is much better than NS17):

This is impedance curve:

and this is 30 degree of axis FR curve (green curve):

 

[jerryzhang]

Amir's measurement also shows that the speaker is very close to 87dB at 1kHz - see green circle below.
View attachment 144250

Unfortunately, the sensitivity of a loudspeaker is not determined according to a standardized procedure. But the sound pressure level at 1kHz, is obviously not suitable to provide a realistic value.
More realistic is the average value of a certain frequency range - for example from 40-1000Hz or 300-3000Hz,...

I have not enough experience for speaker, so I discuss with the test engineer about sensitivity. Another reason is the woofer of NS17 original FR curve as below, the SPL is 85.8db and close to 87db of 1000Hz. I didn't realize it was a problem at that time.
Anyway, I realize NS17 is not easy to drive soon, after I try to test over 10 amplifiers. Finally, I provide independent pre-amplifier and rear amplifiers with 300 watts. So, it is not a problem when they select amplifier now. And I provide 2 pre-amplifier for chose, one is tube pre-amplifier for better tone colour, another one is transistor pre-amplifier for better dynamics and detail.

 

[pjug]

I have not enough experience for speaker, so I discuss with the test engineer about sensitivity. Another reason is the woofer of NS17 original FR curve as below, the SPL is 85.8db and close to 87db of 1000Hz. I didn't realize it was a problem at that time.

Some might consider the sensitivity a problem, but I think it is mostly that many of us don't like the way you are specifying. If you use an average as suggested so that the specification is more honest, then I doubt you will hear much complaining about that. There are other small speakers that Amir has measured that have significantly lower sensitivity than yours.

 

[jerryzhang]

The original FR of woofer of NS17 as below. it is seen very smooth from 100 to 2000 Hz only about 4 db range. But it has a high peak on 3000Hz, almost 10 db. We had to development a very complicated frequency divider to suit this problem. Original crossover frequency is 2000 Hz, but I found treble is little worse than 2500 Hz.
I think too complicated frequency divider cause some problem such as not easy to drive and and valley of FR curve and harmonic distortion issue between 1000-2000 Hz is higher than before.
If move the peak to 4000 Hz for avoid the issues, but f0 should move to 45 Hz or higher (final f0 should be about 43 Hz). Another way move crossover frequency move to 1800 to 2000 Hz, then frequency divider will be simplify too. It is a difficult choice.
Any suggestion?
Anyway, I need to work on it asap. I hope the new measurement method from here will give me the way. I am very regret to don't know the methods that time.

 

[BYRTT]

.....I don’t think I’ve seen a 1” tweeter change so much in dispersion, you could ride the ER & SP curves like a slide.....

Looks Focal Aria 906 had close to same dispersion tendency, but where Focal dialed in a full baffle step loss correction filter for woofer domain then Natural Sound did only half or so a baffle step correction filter for woofer domain ..

 

[jerryzhang]

Some might consider the sensitivity a problem, but I think it is mostly that many of us don't like the way you are specifying. If you use an average as suggested so that the specification is more honest, then I doubt you will hear much complaining about that. There are other small speakers that Amir has measured that have significantly lower sensitivity than yours.

I agree now! I design earbuds before speaker, sensitivity of earbuds always of 1000 Hz. I will correct it when I update NS17.

 

[ctrl]

MZKM said: .....I don’t think I’ve seen a 1” tweeter change so much in dispersion, you could ride the ER & SP curves like a slide.....

Looks Focal Aria 906 had close to same dispersion tendency,

The early off-axis frequency response drop is likely due to the acoustic lens.



An ideal tweeter without acoustic lens shows the following behavior in infinite baffle:


An acoustic lens with 7mm depth, the off-axis frequency responses drop is pronounced:

 

[Chromatischism]

This seems like a treble response that is not easy to fix with EQ.

Interesting use of the anechoic chamber, though. Surely that takes a ton of man hours.

 

[jerryzhang]

The early off-axis frequency response drop is likely due to the acoustic lens.

View attachment 144323

An ideal tweeter without acoustic lens shows the following behavior in infinite baffle:
View attachment 144324

An acoustic lens with 7mm depth, the off-axis frequency responses drop is pronounced:
View attachment 144325 View attachment 144326

which one is better? Acoustic lens, if necessary, maybe I can update my tweeter next generation.

 

[BYRTT]

This seems like a treble response that is not easy to fix with EQ.....

EQ filter in left side column below made Amir some okay happy under listening test, EQ filter in right side column is farfield optimized and on paper transfer curves to look alot like what Focal Aria 906's out of box curves was looking ..

 

[ctrl]

which one is better? Acoustic lens, if necessary, maybe I can update my tweeter next generation.

It is not possible to make a general statement about which tweeter is better suited.
A tweeter completely without acoustic lens or small waveguide radiates very broadly up to a certain frequency, the further behavior is then strongly determined by the cone material.

Your loudspeaker designer can simulate different tweeters using BEM simulation or something similar to see if there is a better alternative.
The tweeter radiation should also not be too wide, since the 6'' woofer does not radiate very wide. Because the woofer is installed "behind" the baffle a high crossover frequency with a flat filter slope (which would "soften" the transsition from woofer to tweeter) is not possible because of the interference caused by the woofer cutout.

 

[jerryzhang]

It is not possible to make a general statement about which tweeter is better suited.
A tweeter completely without acoustic lens or small waveguide radiates very broadly up to a certain frequency, the further behavior is then strongly determined by the cone material.

Your loudspeaker designer can simulate different tweeters using BEM simulation or something similar to see if there is a better alternative.
The tweeter radiation should also not be too wide, since the 6'' woofer does not radiate very wide. Because the woofer is installed "behind" the baffle a high crossover frequency with a flat filter slope (which would "soften" the transsition from woofer to tweeter) is not possible because of the interference caused by the woofer cutout.

Thank you very much for your reply! We need to do some experiment to test which way to improve it.

 

[BYRTT]

Thank you very much for your reply! We need to do some experiment to test which way to improve it.

Think remember from @ctrl example that the real smoothest or linear one to omni naturals is the flat plane, the small acoustic lens example do increase directivity down lower in frequency to improve cooperation with the beaming woofer but up higher in frequency it cost some non optimal "objective" criss cross patterns ..

 

[ctrl]

If one wants to have a look at this complex of topics in detail here are a few links (first two are in German, but the pictures are self-explanatory and there is deepl.com for the translation).

Tweeter cone shape and dispersion pattern
Part two

What does an small acoustic lens or front plate step do


The influence of the diaphragm material can be observed very well with the identical 1'' Bliesma tweeters. Just compare the normalized frequency response measurements on hificompass.com.
Bliesma T25S-6
BlieSMa T25A-6
BlieSMa T25B-6
BlieSMa T25D-6