• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

PURIFI finally did a fully purified passive speaker design! The SPK 16 prototype is here - with a PTT tweeter

I thought you were joking, looks like you really had problems with vertical shift.


Philharmonic BMR Monitor directivity:

1738668572104.png

Horizontal -6dB beam width 160° (+/- 80°)

1738668619214.png

Vertical -6dB beam width 40 ° (+/- 20°)

Reference:

Sitting at a typical 10ft (~3m) with +/- 20 degrees you have +/- 3.64ft (~1m) vertical shift from the reference axis.

If you are sitting at arms length to the speaker, eg. 80cm, then +/- 20 degrees is about +/- 30cm (1ft) vertical shift.

TL;DR?
Setting up a speaker such at the design axis is near the main listen position is important, whether it’s a dome, cone, or planar device driver.
 
Last edited:
Seems to me the use case for the smaller tweeter is in a 3-way with a small mid, whereas the bigger WG would be for 2-ways with medium size midwoofers.

And IMO it would be a huge upgrade from a long, skinny wiggling tinfoil strip - a speaker with wide horizontal dispersion that doesn't turn in to a dull bore when you stand up.
Those wiggles can produce some beautiful music. And at any normal far-field listening distance, they sound just fine standing up. Hopefully the Purifi's will sound as good and play louder for those listeners so inclined. But on a more serious note--interchangeable face plates sound like an excellent idea if that's the only difference between the various models. Lars--is that a practical idea?
 
I came to understand that a somewhat gradually declining power response up to higher frequencies is preferable for a speaker to not sound overly bright. If the purify tweeter now has a rather steady off axis behavior - would that mean that a tilted down on axis frequency response is needed to sound pleasant?
 
And at any normal far-field listening distance, they sound just fine standing up.

Respectfully, that simply wasn't my experience with your standmounts at Axpona last year. But from what I heard seated I would love to hear them with a tweeter, such as this Purifi, that offers similarly wide horizontal dispersion without the vertical compromises. Maybe you could give the upgraded models the suffix "JD5." :)

I came to understand that a somewhat gradually declining power response up to higher frequencies is preferable for a speaker to not sound overly bright. If the purify tweeter now has a rather steady off axis behavior - would that mean that a tilted down on axis frequency response is needed to sound pleasant?

IMO, no, unless they're placed in a hall of mirrors. Air absorbs HF pretty well, and there's energy loss with reflections.

We basically have a system set up on each end of our home. The main speakers in one use ring-radiator compression drivers with very small exits and a well engineered phase plugs/waveguides for basically constant 100ish degree dispersion up to 20kHz. The other uses beryllium dome tweeters in shallow elliptical waveguides with no phase plug or "cohererer," and the standard narrowing HF directivity. Both are designed for flat on-axis response. The biggest difference in the treble between them, honestly, is that the former maintains a subjectively correct spectral balance across a wider seating area.

The Purifi does have quite a bit wider dispersion (~140 deg compared to ~100 deg) so that might change things. Or not. FWIW, I always thought the old Mirages that were very wide dispersion and tuned with a marked on-axis downwards tilt so that the sound power was similar to conventional speakers, sounded dull.

607mirageomd.1.jpg



Note that Grimani tunes their super-wide-horizontal-dispersion horn to flat on axis as well.
 
IMO, no, unless they're placed in a hall of mirrors. Air absorbs HF pretty well, and there's energy loss with reflections.
I think this depends a lot on the level of the DI which does relate to the width of the pattern amongst other things. But can also be personal preference or adaption. I fall very much into the trained listeners camp in the Harman research where I want the in-room to be down at high frequencies with most speakers.

My experience with constant DI dipoles is that they absolutely needed an on axis downslope to sound right and to not be too bright.

Earl Geddes put a slight downslope in his Summa speakers which was also to stop them being too bright. They have a very flat but high DI.

Omni, dipole and flat DI horns could all be called constant directivity but they tend to need to be treated differently to sound balanced.
 
I think this depends a lot on the level of the DI which does relate to the width of the pattern amongst other things. But can also be personal preference or adaption. I fall very much into the trained listeners camp in the Harman research where I want the in-room to be down at high frequencies with most speakers.

My experience with constant DI dipoles is that they absolutely needed an on axis downslope to sound right and to not be too bright.

Earl Geddes put a slight downslope in his Summa speakers which was also to stop them being too bright. They have a very flat but high DI.

Omni, dipole and flat DI horns could all be called constant directivity but they tend to need to be treated differently to sound balanced.
In addition, can it be that the whole recording-mixing-mastering chain over the last 50 plus years has become accustomed to declining power aka downslope hf?
 
my experience with our wide dispersion tweeter is that it simply gives more air, sense of space and nerve to the sound. definitely does not sound bright to me. the difference is mainly above 10kHz. I have listened in different rooms even with concrete walls (our demo room i Munich)
 
Those wiggles can produce some beautiful music. And at any normal far-field listening distance, they sound just fine standing up. Hopefully the Purifi's will sound as good and play louder for those listeners so inclined. But on a more serious note--interchangeable face plates sound like an excellent idea if that's the only difference between the various models. Lars--is that a practical idea?
hi, yes the faceplates (aka waveguides ) are just fastened with 4 screws so interchangeable. Some customers will do their own design and some want screws from the rear (visible screws are getting increasingly unpopular these days). So there will likely be many variants over time.
 
In addition, can it be that the whole recording-mixing-mastering chain over the last 50 plus years has become accustomed to declining power aka downslope hf?
Yes absolutely, as it impossible to consider dropping all the music that already exists, a speaker that uses a directivity that differs sufficiently from industry norms in production is likely to require a bit or a lot of modification for it to sound "right" with the music that already exists.

I think it is fair to say that it depends, but there are definately occasions where for some people a very constant directivity matched to a flat on axis will not be their preference.

The balance between bass and treble can be used interchangably quite often, something that sounds bright could be fixed by giving it more bass, or something that sounds bass light could be fixed with a treble reduction. Just another thing that makes a yes or no answer hard to give.
 
we mention laws of physics here. How is this suddenly considered marketing BS?
It would be perfectly valid if the four poles corresponded to two different,separate amps.
As long as the same amp sees the same speaker load number of poles may be beneficent impedance-wise to a tiny degree.
But to be independent it has to be separate amps I think (and even that has beaten to death here if not in an active configuration)

I mean that I can't see the difference between connecting the two ways of a speaker either at the speaker binding posts or at the amps binding posts.
(I assume that the 4 poles correspond to two connections of course,as other companies modules do too)
 
Last edited:
It would be perfectly valid if the four poles corresponded to two different,separate amps.
As long as the same amp sees the same speaker load number of poles may be beneficent impedance-wise to a tiny degree.
But to be independent it has to be separate amps I think (and even that has beaten to death here if not in an active configuration)

I mean that I can't see the difference between connecting the two ways of a speaker either at the speaker binding posts or at the amps binding posts.
(I assume that the 4 poles correspond to two connections of course,as other companies modules do too)
ah i see. you completely miss the point: we talk about nonlinearity currents arising from the driver back EMF. These currents can cause voltage to appear on other drivers if the amp has non zero output impedance or if there is a common impedance (common on cable). if you run separate cable all the way to a very low impedance point such as the point on the ET amp where the feedback is taken then you have virtually eliminated the common impedance and one driver will not affect the other.
 
ah i see. you completely miss the point: we talk about nonlinearity currents arising from the driver back EMF. These currents can cause voltage to appear on other drivers if the amp has non zero output impedance or if there is a common impedance (common on cable). if you run separate cable all the way to a very low impedance point such as the point on the ET amp where the feedback is taken then you have virtually eliminated the common impedance and one driver will not affect the other.
Ok,let me understand that perfectly clear.
The benefit you're talking about comes both from the 4 pole connections PLUS bi-wiring all the way to the speaker?

Will it be the same if the 4 pole connection stops at the amp's (outer) binding posts as it usually happens or it has to go full bi-wire to the speaker?

(you can imagine the can of worms your answer can open,I'm playing the devil's advocate here)
 
yes to the first part (run 2x2 pipes all the way to the amp).

if the 4 poles join to just two at the back of the amp then we get the internal wiring as common impedance and we get more cross talk. the best is to run 4 poles all the way to the PCB.

All of this may of course not be audible but could be measured.
 
Here's what a typical hard dome tweeter looks like, on a cabinet with large facets for diffraction control:

1738770263909.png


1738770296519.png


Here's what happens when you remove the standard faceplate, and install a custom waveguide:
1738771092123.png


You lose the bulge around 3-4K Hz, or "directivity error" as some call it... but the trade-off is that it narrows the beam-width.
The above measurement was measured on a very large quasi-infinite baffle, and you can see that the -6dB beam-width is about +/- 45°.
On a typical cabinet, say 8" wide, it still be just a few degrees different.

Now here's Purifi's:
1738770359833.png



It will sound spacious due to the wide +/- 70 ° beam-width, all the way out to 20KHz. And it will have life-like dynamics, hitting over 100dB at the 10ft listening position, per tweeter/speaker.

amirm won't need to draw a beam-width line in his reviews, nor will be find any compression taking sweeps at 106dB, let alone 96dB
He'll even need to upgrade his microphone to measure miniscule amounts of distortion.
 
Last edited:
It will sound spacious due to the wide +/- 70 ° beam-width

I can't wait, only other dome I know that does this is the one on the behringer 2030p, constant all the way to the top. I'm kinda curious why no other company (afaik) has managed this. Plenty of well controlled stuff out there but it's narrow.
 
Here's what a typical hard dome tweeter looks like, on a cabinet with large facets for diffraction control:

View attachment 426240

View attachment 426241

Here's what happens when you remove the standard faceplate, and install a custom waveguide:
View attachment 426245

You lose the bulge around 3-4K Hz, or "directivity error" as some call it... but the trade-off is that it narrows the beam-width.
The above measurement was measured on a very large quasi-infinite baffle, and you can see that the -6dB beam-width is about +/- 45°.
On a typical cabinet, say 8" wide, it still be just a few degrees different.

Now here's Purifi's:
View attachment 426242


It will sound spacious due to the wide +/- 70 ° beam-width, all the way out to 20KHz. And it will have life-like dynamics, hitting over 100dB at the 10ft listening position, per tweeter/speaker.

amirm won't need to draw a beam-width line in his reviews, nor will be find any compression taking sweeps at 106dB, let alone 96dB
He'll even need to upgrade his microphone to measure miniscule amounts of distortion.
sounds like you had measuremented this tweeter, i remember you had a project about peerless CD with a big wave guide ,this both about DI control, but purifi's is smaller and look like a usually room setting speakers , which one would you choice if budget not a question?
 
This man, who obviously did an excellent job, stated




And shows


https://www.diyaudio.com/community/...he-easy-way-ath4.338806/page-818#post-7867796


https://www.diyaudio.com/community/...he-easy-way-ath4.338806/page-818#post-7867844


 
Back
Top Bottom