. . . or Space Ship
Yep,
as Brian mentioned, they make their own baskets too. Many of these baskets
are actually plastic? They injection mold whatever basket they need at the
time, and change the plastic mixture depending on what they want. Sound
cheap? Absolutely. They’re quite strong too, and have better damping than
metal. I asked Mark if I could jump up and down on one of these to test his
theory. His eyes darted a little, but acquiesced. I jumped as high as I
could, and landed squarely on a woofer basket appropriate for a 6 inch
woofer, and the darn thing didn’t budge. Somewhat miffed, I committed myself
to eventually break it with repeated stomping and grunting, but I think I
could have done the same with a stamped steel or cast aluminum frame as
well. I just have a particular knack for breaking stuff. But, the point
remains, it showed tight control of a cost-effective method with good
performance results.
Once they've got their cones nice and pointy, they put on their own rubber
surrounds too.
The back plates of their fancier drivers, like those used in their Signature
line, are made by using electricity to burn out, so to speak, the shape that
they want from the solid piece of metal. You can't machine some of the cuts
that are needed to make a mold (example: square hole) so this Electron
Discharge Machine (EDM) is the way to go. Pretty cool.
Similarly, Paradigm makes their own spiders, dust caps (put on dead center,
like their surrounds, with a precision gizmo), crossovers (they wind their
own inductors to insure proper values), and with the exception of the
aforementioned Signature Series, enclosures.
Can I comment on the enclosures that they make in house? Building drivers is
impressive. It differentiates the smaller companies from the big guns. But
it was the enclosure manufacturing that dropped my bottom mandible. The
efficiency and quality of their enclosure (cabinet) manufacturing is utterly
freaking NUTS! They start with pre-laminated sheets of MDF (or particle
board for the most affordable models), cut it in bulk with this huge saw,
roll it over to a CNC machine that cuts out the shapes three at time in a manner
of seconds, send it over to another CNC machine that cuts out the driver and
terminal/crossover openings, and then the end results all fold together to
be finished with a combination of hot glue (for immediate fastening) and
wood glue (for long-term strength), with smaller models folded automatically
by machine! When I heard Brian describe it, I said to myself,
"Clever, whatever . . ." When I actually saw it in action, it burnt noodle.
I
remember Brian mentioning that they test their speakers within a 1 dB window
from their target response. Mark showed me just that. Just to clarify, this
means that they keep the 'reference' tweeter and crossover assembly that
represents the results they want to calibrate respective measuring devices
against these standards. Then EVERY tweeter and crossover assembly must have
a response that fits within a 1 dB window compared to that real reference, i.e.,
± ½ dB. That's not to say that the speaker itself will be flat in
frequency response in any given direction ± ½ dB, but rather that any
given speaker which you purchase will be within
± ½ dB from the target
response for that model. Compared to the rest of the industry, this is
phenomenally good.
They
don't actually test woofers, I guess because variations are far more
difficult to hear at lower frequencies, and the real results in a customer’s
home are more location-dependent (due to room boundaries) than anything the
woofer’s going to do, but I’m just guessing.
I watched their people testing
crossovers and tweeters, and in every instance, the response was easily
within the window. Not that it should really be so surprising, given the
effort put into maintaining consistency in the parts construction from the
get go.
I found it even more reassuring when I saw evidence that some parts do in
fact fail, indicating that this testing wasn’t just for my benefit. I came
across a whole box of tweeters that didn’t pass and were marked for
disposal.
Given that these were drivers created with great pains to be consistent, it
makes me shudder to think how many speakers are out there, made by
manufacturers who don’t test every product, let alone have stringent
manufacturing practices, that even if identical in model, parts, and design,
still aren't "voice-matched" in a true and practical sense. Can you imagine
how poorly a pair of loudspeakers would perform creating a 'stereo' image
when one speaker doesn’t truly match the other? I remember John Dunlavy (a
high-end loudspeaker engineer who used the OEM method for drivers, and after
performing testing of every driver consistently rejected more than half of
them) commenting that most of the speakers out there, and even some
highly-regarded products, had exactly this problem.
Click Here to Go
to Part IV.