ASR Directiva Open Source Speaker Review

[ctrl]

I fail to see how you can get more expensive than that?

Try an diamond membrane tweeter from Seas, Bliesma or Accuton for example.

 

[voodooless]

Try an diamond membrane tweeter from Seas, Bliesma or Accuton for example.

Oh my… € 6800 for the Seas . Only must be really desperately rich to go for this.

 

[preload]

Yes , this is a good tweeter to play 44,1 or 48 kHz sampling material. Not for 88 or 96 kHz . The resonance at 27 kHz will have distortion peaks at multiples of that frequency. That will cause a rise in the distortion profile at 13.5kHz for the second-order, 9kHz for the third-order and 5.4kHz for the fifth-order.
Something that can be seen at 96 dB distortion measurements at 13,5 kHz .

Can you explain that further? Harmonic distortion products, which are multiples, would be at higher frequencies (ie 54khz for H2). Is the existence of distortion products lower than the fundamental your own theory or have you seen it described somewhere? I'm asking because I am curious. Thanks.

 

[ctrl]

OK, know any speakers that uses it? I'm curious what the final cost would be for such a speaker.

This always depends on the tweeter used.
A B&W 2-way loudspeaker with diamond tweeter is already available for 8000€ per pair.

805 D4 Stand-mount Speaker | Bowers & Wilkins

The Bowers & Wilkins 805 D4 stand-mount loudspeaker is the most compact model in the 800 Series Diamond range. Discover more.

www.bowerswilkins.com

 

[eddantes]

@Rick Sykora @ctrl Gents, congrats on what should rightly feel like an achievement! Those are impressive looking charts!

I did have a (possibly stupid) question: Would offsetting the tweeter to one side have a moderating effect on that dip caused by vertical reflections?

I realize such a change would introduce "other things"... but I am curious.

 

[ctrl]

Would offsetting the tweeter to one side have a moderating effect on that dip caused by vertical reflections?

A few centimeters of offset shouldn't really make much difference.

And if the tweeter were to be offset vertically at 45° to the woofer ("diagonal"), for example, then it is very likely that the horizontal measurements would already show the effects of the diagonal cancellations.

Update: Horizontal the different distances of the tweeter to the baffle edge will make a difference.

 

[eddantes]

A few centimeters of offset shouldn't really make much difference.

And if the tweeter were to be offset vertically at 45° to the woofer ("diagonal"), for example, then it is very likely that the horizontal measurements would already show the effects of the diagonal cancellations.

Thanks for the response. It appears that such a change would make marketing people happy but have no meaningful effect. Cheers.

 

[changer]

Am really anxious to try a larger speaker and am looking forward to the next iteration. I hope many of you are as well!

Yeeees, a cost friendly JBL 708p-ish speaker with a 3D printed axisymmetric OS-SE waveguide and DSP!

 

[PeteL]

This always depends on the tweeter used.
A B&W 2-way loudspeaker with diamond tweeter is already available for 8000€ per pair.

805 D4 Stand-mount Speaker | Bowers & Wilkins

The Bowers & Wilkins 805 D4 stand-mount loudspeaker is the most compact model in the 800 Series Diamond range. Discover more.

www.bowerswilkins.com

Thanks.

 

[McFly]

A few centimeters of offset shouldn't really make much difference.

And if the tweeter were to be offset vertically at 45° to the woofer ("diagonal"), for example, then it is very likely that the horizontal measurements would already show the effects of the diagonal cancellations.

Update: Horizontal the different distances of the tweeter to the baffle edge will make a difference.

What if the tweeter and woofer were say 15mm lower from the top of the baffle, just to minimise thin areas to apply veneer to if one were finishing a set?

 

[changer]

Everytime you change the distance of a driver to the baffle edge, this changes the diffraction behavior and the correction filters that where developed for the former driver placement will possibly fail to correct the new. This is how it goes. You must measure again.

 

[Chromatischism]

Yes , this is a good tweeter to play 44,1 or 48 kHz sampling material. Not for 88 or 96 kHz . The resonance at 27 kHz will have distortion peaks at multiples of that frequency.

Just because the sampling frequency is that high, doesn't mean there is content there.

 

[richard12511]

Yes , this is a good tweeter to play 44,1 or 48 kHz sampling material. Not for 88 or 96 kHz . The resonance at 27 kHz will have distortion peaks at multiples of that frequency. That will cause a rise in the distortion profile at 13.5kHz for the second-order, 9kHz for the third-order and 5.4kHz for the fifth-order.
Something that can be seen at 96 dB distortion measurements at 13,5 kHz .
Here is a comparison with Genelec 8330 .View attachment 157504
View attachment 157503

But with real music(including high res), won't the 27kHz fundamental be significantly lower in spl? It varies, but looking at most music, it's often -30dB at 20kHz, with 27kHz being even lower. I think the only reason it shows as a problem here is because this is not actual music, but rather a sine sweep that plays 27kHz at the same spl as it does 100Hz. I would think those peaks almost completely disappear with real music(assuming 27kHz really is the fundamental).

Seeing these side by side is a good way to highlight the value(at the build price) of these speakers. 8330 already has great distortion performance, and yet the Directiva is still considerably better. Directiva also wins the spinorama battle, side by side. Not saying the Directiva is necessarily the better speaker(GLM is a great thing), but it definitely seems to measure better.

 

[richard12511]

This always depends on the tweeter used.
A B&W 2-way loudspeaker with diamond tweeter is already available for 8000€ per pair.

805 D4 Stand-mount Speaker | Bowers & Wilkins

The Bowers & Wilkins 805 D4 stand-mount loudspeaker is the most compact model in the 800 Series Diamond range. Discover more.

www.bowerswilkins.com

Given the price of the Seas diamond tweeter, that's actually a good price.

 

[musicforcities]

do you do you know any known designs of a 2 way bookshelf speaker pair, DIY or commercial, that would have a larger than 1.4K BOM cost before any active or passive electronics,

I am sure there are in materials and especially if you add a nice finish to a commercial speaker.

as pointed out above someplace, it is actually more challenging to design a small two way that tests this well thanks a larger speaker.

For example, take the Mirage MRM -1 I previously mentioned. They came out in 2000 but 2021 retail would be about $3500 a pair at least. A pair. The nicer finishes would add a few hundred to that. They have 1.5” cnc corian front baffles, three layer construction with 3/8 welded steel in the middle, custom titanium tweeters housed inside a sealed section of the cabinet, a custom high excursion carbon/mica infused poly woofers with stainless machined phase plugs and machined frames, multiple part magnet. Those sorts of things were not as common in 2020. Also a crossover with 18 elements. We can ignore the luxurious overkill of the four Cardas binding posts…cabinet jewelry (they look great but I’d prefer speakon any day. Abd the thick metal gold plated jumper bars are craptastic in that they very easily work loose and don’t make great contact even in the best case. I wish they had only one set. But I can’t bring myself to rewire them….I digress). Sales volume was very low, as they are for such things, driving the per unit cost of drivers etc up if they are not right off a rack (and these were not). So I could see them approaching 1.4K BOM at today’s prices. I am sure there are current examples, but you get the idea.

Of course, as you pointed out, the retail price of the Directiva would be far higher than the BOM. the materials are not the majority of costs of production For commercial. one must factor in R&D costs plus labor, marketing and other production costs. the breakeven/cost per unit price point will easily exceed 1.4K with drivers and materials likely no more than 1/3 of that cost. Now add retail mark up….

I think you are correct that the Directiva would retail in the region of $4000 -4500 a pair of commercial. Maybe a bit less if it was seen as a loss leader. I would think volume of sales is so small at such price points mass-production is a bit of a misnomer. And profit margin per unit would be considerably higher than mass market stuff.

Of course one would wabt to get the matching stands…which would be where all the real profit margin would be…lol

 

[PeteL]

I am sure there are in materials and especially if you add a nice finish to a commercial speaker.

as pointed out above someplace, it is actually more challenging to design a small two way that tests this well thanks a larger speaker.

For example, take the Mirage MRM -1 I previously mentioned. They came out in 2000 but 2021 retail would be about $3500 a pair at least. A pair. The nicer finishes would add a few hundred to that. They have 1.5” cnc corian front baffles, three layer construction with 3/8 welded steel in the middle, custom titanium tweeters housed inside a sealed section of the cabinet, a custom high excursion carbon/mica infused poly woofers with stainless machined phase plugs and machined frames, multiple part magnet. Those sorts of things were not as common in 2020. Also a crossover with 18 elements. We can ignore the luxurious overkill of the four Cardas binding posts…cabinet jewelry (they look great but I’d prefer speakon any day. Abd the thick metal gold plated jumper bars are craptastic in that they very easily work loose and don’t make great contact even in the best case. I wish they had only one set. But I can’t bring myself to rewire them….I digress). Sales volume was very low, as they are for such things, driving the per unit cost of drivers etc up if they are not right off a rack (and these were not). So I could see them approaching 1.4K BOM at today’s prices. I am sure there are current examples, but you get the idea.

Of course, as you pointed out, the retail price of the Directiva would be far higher than the BOM. the materials are not the majority of costs of production For commercial. one must factor in R&D costs plus labor, marketing and other production costs. the breakeven/cost per unit price point will easily exceed 1.4K with drivers and materials likely no more than 1/3 of that cost. Now add retail mark up….

I think you are correct that the Directiva would retail in the region of $4000 -4500 a pair of commercial. Maybe a bit less if it was seen as a loss leader. I would think volume of sales is so small at such price points mass-production is a bit of a misnomer. And profit margin per unit would be considerably higher than mass market stuff.

Of course one would wabt to get the matching stands…which would be where all the real profit margin would be…lol

It’s all hard to tell and I get your point, I don’t have experience in speaker designs, but I do in electronics. In general, and yes there are exceptions, but, in General, the real price of a commercial product is more like 6 times BOM cost, anything less you are losing money, but direct sales can remediate a bit of that. Like Schiit for exemple. Hi volume can also shave some of that factor, but again, ultra Hi end don’t do volume.. That was my only point. In this case BOM cost is not 1400. because with this cost you don’t have a finished product, at the very least it’s missing a crossover, and as designed, you are missing amplification, ADC, DSP, and DAC for 4 channels of audio. In my opinion, and again I absolutely have nothing against that and I value this kind of performance, but I stick to it, translated as a sellable products, we are looking at 10k$+ active bookshelves, In other words, very Hi end territory, and as for DIY, you save some, but not as much as one would think, we´ll have to know how much the Kits come at, but from what I get, the kit will still be only drivers, encolure, and connectors, so something that needs electronics in order to be consider full active speakers. Again, not a critic of this design, I’m all for it.

 

[Tangband]

But with real music(including high res), won't the 27kHz fundamental be significantly lower in spl? It varies, but looking at most music, it's often -30dB at 20kHz, with 27kHz being even lower. I think the only reason it shows as a problem here is because this is not actual music, but rather a sine sweep that plays 27kHz at the same spl as it does 100Hz. I would think those peaks almost completely disappear with real music(assuming 27kHz really is the fundamental).

Seeing these side by side is a good way to highlight the value(at the build price) of these speakers. 8330 already has great distortion performance, and yet the Directiva is still considerably better. Directiva also wins the spinorama battle, side by side. Not saying the Directiva is necessarily the better speaker(GLM is a great thing), but it definitely seems to measure better.

Yes thats true.
The 8330 is one of the best measuring loudspeakers in that pricerange , but the Directiva is better at some points , and thats an achivement.

Suggestion : A DIY:er can also go for a closed box Directiva, with Linkwitz transformer.
The purify drivers can take the extra strain. It also makes the Directiva better suited to mix with subwoofers if its closed.

 

[Arc Acoustics]

The woofer is superb, but the tweeter simply isn't.
Directivity is not that good, and one of my least favoured tweeters because of the way it sounds, I genuinely hate that thing if I'm honest.

So, choosing "as a product" by the measurements and my personal impression of the tweeter, I still will choose one of Neumann over Directiva.
They sound neutral at least, and there is a clear difference between the loudspeaker which is designed by professionals who can design the directivity properly or not.
If I spend on that woofer in my DIY project, I will choose (subjectively) a far preferable tweeter and design a better waveguide for it.

 

[Tangband]

Just because the sampling frequency is that high, doesn't mean there is content there.

Thats true. But it might happen in some recordings.

 

[Tangband]

Can you explain that further? Harmonic distortion products, which are multiples, would be at higher frequencies (ie 54khz for H2). Is the existence of distortion products lower than the fundamental your own theory or have you seen it described somewhere? I'm asking because I am curious. Thanks.

Read more here.

www.audioexcite.com » SEAS H1499 27TBCD/GB-DXT