Selah Audio Purezza Review

[yourmando]

Wow, another fantastic review. I really appreciate all of the time it must take to do these reviews.

That's an amazingly smooth early directivity index from a rectangular 2 way box speaker design. Looks like driver size, selection & crossover design were enough for a smooth directivity handoff. No fancy waveguides needed.

From the nearfield driver measurement chart, looks like the woofer was crosses over to the ribbon at 2.4khz or so? If so, great job on not really making a blip in that region on the polar plots (at least horizontally).

Looks like @Selah Audio nailed it!

 

[yourmando]

And you are correct that the Selah kit doesn't include cabinets or flat packs? The equal price difference from what I can see is around $350 less for the BMR. At your fully assembled price of $1700, it is probably easier to go that route.

Looks like kits do not include cabinet flat packs, but he can provide a quote:

 

[mitchco]

You can see the prediction follows pretty reasonably to what is measured in the seated position, aside from the major modal dip at ~125Hz and the dip between 2-3kHz.

Great job man! Thoughts about that dip between 2-3 kHz? PS. Those Purifi woofers are indeed amazingly low distortion. I like how they can be cranked with no hint before they give up - so unlike most conventional drivers...

 

[hardisj]

Some of you may notice that this review's content is a touch different than my previous reviews. I am no longer providing a dedicated compression and THD test result. The reason for this is two-fold:
1) I have replaced my THD results with Harmonic Distortion gifs that show not just the THD but also the 2nd-5th orders of distortion. Some folks were asking for the nth-order components and it made more sense to me to provide the HD data in this manner, broken out fully, in lieu of the (now redundant) THD-only plot. Additionally, I have provided this in gif form at varying levels. The gif version was chosen because it illustrates how the drive level (from -6dB, 0dB (2.83v), +6/8/10/12/14dB) alters the output distortion.
2) The dedicated compression graphic also is gone because it was redundant with the Max SPL compression graphics. Actually, the Max SPL compression graphics were much better; they are a more complete and realistic set of compression data than the compression graphic I was providing before. There is no need in providing both. So, I dumped the old one.


I'm not asking for opinions. I like it the way it is now and it is a little easier on me as well. I'm just letting you know in case you were wondering.

 

[hardisj]

Great job man! Thoughts about that dip between 2-3 kHz? PS. Those Purifi woofers are indeed amazingly low distortion. I like how they can be cranked with no hint before they give up - so unlike most conventional drivers...

Thanks!

I'm not sure. I always look at the NF data to see if the drivers are breaking up, and while my evaluation of the Purifi woofer a la carte shows mild breakup >2kHz, the NF data indicates this has been resolved through the crossover. That leaves me then to assume that the dip is more a culprit of edge diffraction at ~3.5kHz. But that's just a guess. I'd have to model things to see. There are folks here who can make quick work of that, while it would take me having to remember what program to use and how to do it. Been years since I've modeled a full loudspeaker.

I forgot to mention that once I saw the data, I realized the horizontal axis HF response seemed to flatten out more when turned about 10-20° off-axis. So, I did some listening with them in that configuration, though, I can't honestly say I could hear enough difference that couldn't have been influenced by the fact I already knew what to expect.

 

[hardisj]

Thanks for the feedback, folks! I appreciate it. One never knows how the effort will be received. So, the kind words are a relief and serve as motivation to keep going.

- Erin

 

[yourmando]

@Selah Audio, that BZ Labs CQ76B tweeter looks interesting.

Can you comment on why you chose this ribbon tweeter? When I search on it, I find very little info, and mostly mentions from you in your own designs!

I’d be curious to hear how you heard about it and your thought process in using it with some of your designs.

 

[xarkkon]

Great super detailed write up as always! Given the glowing review of these speakers, were there any parts of if that you thought could be better?

Following the earlier messages, in what way are better than the BMRs? As Dennis pointed out, price is a factor, but I'm interested to know how the BMRs fall behind to these.

With the passive radiator on these (funny they don't mention it on their website, didn't realise there was one till your review), how do these compare to the Buchardt S400s?

For future reviews, would you consider doing a quick comparison vs other speakers that you've reviewed before? For someone with a pair of Buchardts looking for a next DIY project, that comparison might give a much clearer idea of what to expect from one of these. Thanks!

 

[HooStat]

Boy, these look great. I think I would opt for the more balanced bass profile at the expense of some sensitivity. Can always lower the bass with EQ as part of room correction. But this is a very nice job, both on the speaker design and on the review.

 

[jhaider]

I’m impressed the racetrack PR can keep up with the woofer.

 

[Martin]

I wonder why the Purifi passive radiator wasn't used, what changes would be required to use it, and it's effect on the speakers' bass performance. The look of that funky surround on the Purifi woofer is growing on me.

Martin

 

[Chromatischism]

Great review as usual.

Very interesting design decision to use that passive radiator. I remember sending him nearfield measurements of mine last year which might have piqued his interest. I'd be curious to hear more about why he ended up choosing it over a vented design. My guess is it allowed for a smaller cabinet which matched well with the active woofer, or even to allow closer placement to the front wall as he alluded to with his BSC. Basically, the same reasons it was chosen by Buchardt Audio. Rick may have other reasons.

That's an amazingly smooth early directivity index from a rectangular 2 way box speaker design. Looks like driver size, selection & crossover design were enough for a smooth directivity handoff. No fancy waveguides needed.

From the nearfield driver measurement chart, looks like the woofer was crosses over to the ribbon at 2.4khz or so? If so, great job on not really making a blip in that region on the polar plots (at least horizontally).

Indeed, however another thing a waveguide would do is to broaden the high-frequency dispersion. Notice that the ribbon starts to beam above 10 kHz whereas a waveguide would maintain a more constant directivity throughout the whole tweeter range to 20 kHz. Tradeoffs. Example with waveguide (Buchardt S400):

Boy, these look great. I think I would opt for the more balanced bass profile at the expense of some sensitivity. Can always lower the bass with EQ as part of room correction. But this is a very nice job, both on the speaker design and on the review.

I had similar thoughts since I use XT32 to flatten and linearize the bass region but maybe it doesn't matter to us as much. Perhaps making the speaker work well for those not using room correction is the right approach.

I wonder why the Purifi passive radiator wasn't used, what changes would be required to use it, and it's effect on the speakers' bass performance. The look of that funky surround on the Purifi woofer is growing on me.

Just a guess – likely because you'd have to use 2 of them to get enough area which would be challenging to fit and not increase the size of the speaker.

 

[Dennis Murphy]

You beat me to the punch. I recall you testing the Purifi.

Would there be a difference with say using two SB Ceramic woofers in a floor stander compared to the Scan?

And you are correct that the Selah kit doesn't include cabinets or flat packs? The equal price difference from what I can see is around $350 less for the BMR. At your fully assembled price of $1700, it is probably easier to go that route.

I'll soon be getting a lot more hands-on experience with the Purifi in a properly sized box. I'll be using two of the 8-Ohm version in a small floor stander along with a new "Airblade" 180-degree AMT radiator that can be crossed around 1200 Hz. Maybe that will make a Purifi believer out of me. I will be bringing out a tower version of the BMR, but it will use the same 8" ScanSpeak Revelator woofer I used in my Philharmonic 3. I've just never found anything that comes close to it, and it likes transmission-line loading.

 

[yourmando]

I wonder why the Purifi passive radiator wasn't used, what changes would be required to use it, and it's effect on the speakers' bass performance. The look of that funky surround on the Purifi woofer is growing on me.

Martin

Maybe the racetrack passive radiator allowed for a smaller box?

Hificompass has a clever design that uses 2 of the Purifi passive radiators on opposite sides to reduce vibrations, plus a diagonal sound barrier. But it does look bigger.

https://hificompass.com/en/projects/2-way-systems/puri-bliss-bewg

 

[yourmando]

Indeed, however another thing a waveguide would do is to broaden the high-frequency dispersion. Notice that the ribbon starts to beam above 10 kHz whereas a waveguide would maintain a more constant directivity throughout the whole tweeter range to 20 kHz. Tradeoffs. Example with waveguide (Buchardt S400):

View attachment 81924

True. The polar response and sound power directivity index do show beaming above 10kHz, but what stood out to me is that the early reflections directivity index remains quite smooth from 0 to about 5db. It looks about as good as the Buchardt early reflections DI!

 

[Ilkless]

Indeed, however another thing a waveguide would do is to broaden the high-frequency dispersion. Notice that the ribbon starts to beam above 10 kHz whereas a waveguide would maintain a more constant directivity throughout the whole tweeter range to 20 kHz. Tradeoffs. Example with waveguide (Buchardt S400):

View attachment 81924

Actually, that's more down to the fact that Buchardt uses a 3/4" dome rather than the typical 1" dome. Yes, with the right throat design and driver, as we see in some of the Joseph Crowe waveguides on this forum, it can expand, but that's more of the exception than the norm where top-octave dispersion is substantially determined by the radiating diameter of the driver itself because of the small wavelengths at those frequencies. Tools to expand dispersion include phase shields and diffraction-based waveguides (Eg. SEAS DXT).

Overall, that is some exceptional spin performance with a 6.5" midwoofer and direct-radiating ribbon. Really impressed by what Rick managed to pull.

 

[Chromatischism]

Overall, that is some exceptional spin performance with a 6.5" midwoofer and direct-radiating ribbon. Really impressed by what Rick managed to pull.

I'd love to hear them.

 

[Ilkless]

I'd love to hear them.

To keep the dispersion mismatch within +/-1dB between these 2 types of drivers is just excellent XO design.

 

[Selah Audio]

Wow, another fantastic review. I really appreciate all of the time it must take to do these reviews.

That's an amazingly smooth early directivity index from a rectangular 2 way box speaker design. Looks like driver size, selection & crossover design were enough for a smooth directivity handoff. No fancy waveguides needed.

From the nearfield driver measurement chart, looks like the woofer was crosses over to the ribbon at 2.4khz or so? If so, great job on not really making a blip in that region on the polar plots (at least horizontally).

Looks like @Selah Audio nailed it!

Thanks for your positive comments. One of my design goals was to show what could be achieved without a waveguide.

 

[Poseidons Voice]

Erin,

Take a bow!

Dennis M and Rick S.,

The idea of a non waveguide based speaker, either 2 or 3 way with careful attention to smooth directivity transitions between drivers in the mid to higher frequency range is growing on me. As is the goal for wide dispersion.

Thanks for the options!

Best,
Anand.