Yes, this distance is acoustically rather small (under a quarter wavelength) and the drivers smoothly sum and blend into one another.
The intention has never been to make a time coincident speaker ala Thiel and Dunlavy. The benefits of this approach (preserving waveshape), of which the audibility is debatable, is outweighed by the big concessions in power handling, distortion, directivity and lack of supression of driver break-up.
Modern DSP systems can accomplish linear phase high order filters but this requires a fully active system and the associated issues with amplifier, DAC/ADC quality, digitizing analog sources etc. I particularly like the DIRAC mixed-phase stuff, which does this sort of time domain correction.
Well, yes, the closer that we can get to the thoretical ideal, the better, but speakers deviate orders of magnitude higher from the ideal than something like an amplifier or DAC. The question then becomes, which errors to you prefer and the whole art and science of the design of those compromises comes into play.
In any case, part of the reason that i enjoy loudspeaker design and the advance of the technology is that they are such imperfect devices relatively speaking and there are some many opportunities, as a designer, to improve things.
What I also meant is, what percentage of speaker design is final voicing and tweaking crossover etc. when theory, parts and rough crossover design are done. So far I assumed In speaker design, relevance of / time consumed with final voicing and fine tuning also is much higher than with amps/DAC’s, even with the best parts/concepts on hand. Some characteristics and their degree of quality even seem to be counterrotating and hard to combine when I think of some speakers excelling in certain and loosing in other characteristics…but probably these are just effects attached to certain concepts.
I sent in a question to “Ask Paul,” but you might be THE GUY to answer the question.
I’ve heard several good sounding designs recently that use ribbon mids and/or tweeters in a more or less “conventional” tower enclosure. My question to Paul was what would be gained in a line source design using PS Audio’s new drivers (beyond running the individual drivers more gently)? From what I heard at the RMAF, the dispersion of the re-designed AN3’s is pretty remarkable.
Also, it seems the latest ribbon midrange drivers have more dynamics than full-size electrostats and planar magnetics in spite of having less radiating area. Wondering if that is, in fact, true and if so, why?
AMT‘s are practically folded surfaces, unfold them and see how the surface increased.
With planar foils, practically the complete surface faces the direction of movement and is pushing against the perpendicular (strongest) counter forces of the air.
Due to the folded construction of the AMT‘s only the percentage of surface facing towards the direction of movement is pushing agianst the air. While the percentage of surface that is not facing the direction of movement, is exposed to much less forces (which are not perpendicular to their surface) but yet contributes to the dispersion.
Due to their more compact design, it is easier to drive them with more power per surface area. The magnetic field can be concentrated to that smaller space allowing stronger magnetic fields.
Consequently: more power, less pushback => more dynamics.
Very well explained!
And it fits to the practical experience why not only few planar speaker lovers arrive at such AMT or generally ribbon fitted dynamic speakers.
For planar lovers many conventional speakers pay their dynamic advantage with too little resolution, speed and other disadvantages…most of this changes with such ribbon designs and one recognizes that 80% of music‘s soul (and especially its live character) is dynamics when the rest also fits.
I knew that the AN3 design initially featured an AMT folded ribbon driver for the highs, but the original driver in the first AN3 prototype (the Bohlender-Graebener Neo 10) is a planar-magnetic ribbon that (I don’t think) employs the folded structure of an AMT design.
So wondering how Chris’ new design of the midrange driver improves on this, and if this has changed the thinking on the line source designs originally proposed for the AN2 and AN1.
Oh I thought the AN3 coax driver was composed of AMT’s, but when I just watched the video above it seemed to be composed of straight ribbon drivers. Anyway it looks indeed more the ‘D Appolito philosophy than line array. Perhaps the AN3 is just too small for placing an efficient line array and that the line array will be applied in the taller AN2 and AN1.
I have a feeling these are going to be special. I heard the ‘beta betas’ and thought they were pretty darn good. I am betting the redesign pushes things over the top
We jokingly call it the 90/90 rule, where the first 90% of the project takes 90% of the time and the last 10% of the project takes another 90% of the time
Yes, I think that what you’re saying is true regarding the final subjective character following the engineering concept. There is a lot of variability as to the execution of a given concept though, and there are always examples of really good performing and bad performing versions of a given technology, often relating to the art and craft of things.
Before my involvement with the project, PS Audio was working on custom AMT drivers. Unlike a cone / dome drivers, since planars and AMTs aren’t axisymmetric, each uses custom tooling and assemblies and are a bit of a “science project” to get the performance and manufacturing consistency where you want it.
My personal history is about 20+ years working with planar drivers, with my family owning Speakerlab in Seattle, where David Graebener of Bohlender Graebener, Eastech, Wisdom Audio was the engineer there when I was growing up. He and Jim croft designed the Carver ALS driver, and later Speakerlab Auricle and BG RD series drivers. I later went on to work at BG for 5 years and worked with Igor Levitsky there.
Carver and BG supplied Genesis (when Paul and Arnie were there) and obviously prior to that the Infinity EMIT, EMIM etc. were SOTA planar type drivers and so we’re still really trying to leverage that know-how and make some products highlighting the best characteristics of thin film drivers.
My work at Adire Audio before that, working with wold-class driver design engineers has informed a lot of the cone transducer work, where I’m trying to bring some technology and value there to our other custom drivers.
Anyway, that’s kind of where I’m coming from as a transducer guy, but we’ll obviously want to use the best tool for the job. Planar tweeters are quite expensive to do right (in part because of the huge investment and understanding in the parts and assembly of domes) and so in inexpensive speakers, I still would consider domes, if it is the best available at the budget.
Yes, we’re still doing line array AN2 and AN1 designs and have been working on those and I’m very excited about those projects. It’s really a dream gig for me and have been thinking about what I have wanted to do here for a long time.
Nice! So what improvements in SQ or spund presentation are you working to achieve with a line source design vs what is in the AN3’s I heard at RMAF? (those sounded pretty.darn.good).
Well, thanks for asking. I think that, moving forward, I would want to get close to finishing the product and then market them (talk about all of the details etc.) rather than the other way around
Line arrays, in general offer much more dynamic capabilities and effortlessness in that regard. They also offer better spacial SPL consistency. Since you are listening in the nearfield at mid and high frequencies, depending on the size of the array and listening distance, the SPL drop with distance is halved, and so the “sweet spot” can be much wider.
This wording makes me think the author is a marketing type or just loves confusing the public. “The AN3 on display at RMAF 2019 is the final pre-production prototype.”
