On modeling vs. prototyping. "On the front end there's a good bit of modeling. That's really part of a weekly circular routine. Once the design, performance and price targets are loosely pinned down with the form factor, non-modeling takes over. It's not always the same routine but with targets set, we next look at old and new ideas against style, function and what the material pallet might be. Now materials can be independently modeled to see properties which then informs the super-structuring and driver mounting. We usually do this in non-exotic software like Fusion 360, Vectorworks and plug-in-less SolidWorks. Once the solid is detailed, we throw it through more sophisticated software. Plug-ins for SolidWorks have been our go-to over the last decade but there are some handy features in Comsol that can speed things up. Hardware and computing power are a real issue when going deep and usually this is totally unnecessary.
Soul VI, 16Ω Soul Supreme and Omen Def Supreme for another size reference.
"I personally use VectorWorks and Fusion 360 because they're fast and output virtual parts which are easily made into machine code. I want to see actual things in my hands more than work them to death in software. Others here dive deeper into the modeling of ideas. Rick Henthorne loves to work at the particle level of wave mechanics, Layne Reed likes to model acoustic behaviors with his top-shelf thermodynamics plug-in. Layne is also ace at statics/dynamics and has those modules at hand. On the electrodynamic side, that's best modeled with outcomes which already fully reflect reality. Software is amazing nowadays. As a kid I thought we had it good when we could model port and box tuning and it would match predictions. Computing and modeling are critical to get stuff right with less cost. Working virtually with software shortcuts is also much faster than working it out in math; especially when all that has already been done in one form or another. Occasionally there are new math problems however you must slog through.
"Once the design team of three or less works with a real thing, there are actual sounds which inform where to go next and what to improve. Now hard measurements come into play. Some things are easier seen on-screen like electrical/acoustic reactivity within an impedance curve or Smith chart. Magnitude and impulse response plots help expose small details that might get buried in the excitement of a new thing we're chucking music and sounds through. I'd say that more than half the 'discoveries' and 'improvements' occur solidly in the experience/listening domain but the multi-faceted driver/cabinet design phases, modeling and using the math and experience of all those who came before is the real totality we all stand on. Without it I don't think that personal experience and elbow grease would yield much more than a touch of this or that – and often not in the direction of making good sound."
Vintage Marantz with modern Rupert Neve and vinyl.
My next question segues back at the promo video. There Sean had said that without a firm goal as a personal idea on the ideal sound, he'd not know where to go Had his idea on that sound changed over the years? What does he listen for? Does he use particular sounds or music as reference because he's very familiar with them as live sounds? Is it a more intuitive recognition of 'now it's right'? This gets to the crux of speaker design. If you won't recognize when you're done because result overlays intended target, you work in the dark. The target can be measured, subjective or a combination. A reference is simply mandatory. What is Sean's?