2020, Chapter 6:
Tales Told By A Shotgun
Or, “Why do we do both great-measuring products, and ones that would have measurement aficionados screaming in horror?”
Or, “Why op-amps…and tubes?”
Weeeeeellll…
Let’s TL;DR:
Same reason Ford does the Mustang GT350 and the F-150.
Same reason Chebby does the Corvette and the Silverado.
Same reason Mercedes does the AMG GT and the S-Class.
And so on.
Not into car analogies?
Same reason Kershaw does everyday, outdoor, multi-function, work, and designer knives from the single digits to three digits in price.
Same reason Specialized does bikes from $350 to $8000+ in mountain, road and e-bikes, in single-speed to dizzying ranges of speeds.
Same reason there are manual, calibrated, and automatic espresso tampers, ranging from one to four digits in price.
Bottom line, different people like different things, so we do a lot of different things. Just like a whole lot of other companies, in a whole lot of different industries.
Simple as that.
Yes, seriously.
“But What Ties You Together?”
Aha. Great question.
Or wait, let me rephrase. The question really is, “Well, if you just make a whole bunch of different crap, how do you keep from devolving into an evil megacorp just pushing for more and more and more and more sales, devouring our wallets, our planet, and the entire universe?”
Simple. Here’s the commonality:
We make fun, affordable audio components that we like…and we hope you like too.
See the connection?
Fun.
That’s it. Yes, that’s all. Seriously.
We’re having fun coming up with a whole lot of cool stuff. If you like it (and you’ve liked quite a bit of it), we make more. If you don’t, we stop making it.
We recently gave ourselves more permission to have fun with the various Thunderdome products, where, instead of second-guessing ourselves, wondering what you’d like, going through icky focus groups, or otherwise hamstringing ourselves with paralysis-by-analysis, we decided to introduce multiple products and see what you like.
And, you know what? It’s working out fabulously.
“Wait a sec,” someone says. “You also said, ‘Affordable.’ How the hell do you justify that when you make crap that sells for like $2K?”
Let’s go back to another analogy:
Same reason the new mid-engine Corvette is affordable, when compared with cars that sell for 3-4x its price.
Also, let’s be real. We sell exactly one product for more than $2K, and that’s Yggdrasil. Ragnarok 2 starts at $1499.
Still think that’s ridiculous?
Compare Yggdrasil’s manufacturing cost (with four $64 DAC chips, DSP with custom filter code that took 5 person-years to complete, USB with custom USB receiver code that took 2 person-years to complete, an insane choke-input power supply, a completely nuts seamless-look chassis, etc.) to other stuff on the matket, and yeah, it’s expensive, but it’s not priced like a car or a house, which is what some others night charge. Sure, it may not be your cup of tea, and yeah, you might prefer our best-measuring DAC, Modi 3, at $99, and that’s totally fine, but Yggdrasil’s not overpriced.
Same with Ragnarok 2. Compare it to other integrateds recently reviewed in Stereophile, some with
DAC modules that cost more than Ragnarok 2 itself, and then get back to us on how it’s a crappy value.
Still not convinced?
Sure. Cool. That’s why we do tons of products at $99-299 that are designed specifically to maximize performance per dollar. Fulla 3, Hel, Magni 3, Magni Heresy, Modi 3, Asgard 3…all really solid choices.
“What about Vali 2?” astute readers might ask.
And yeah, that’s a great option too. But it’s more power-limited than the Magni twins, and its distortion may raise eyebrows. But it’s a great way to see for yourself if there’s something in this “tube sound” thing, and if you want to pursue it further…for $149.
And that’s where everything comes together. Some people make think Magni Heresy and Modi 3 are the best thing ever…and some might prefer Vali 2 and Modi Multibit.
And that’s perfectly fine.
Same way it’s perfectly fine we like to experiment with both kinds of technology—both cutting-edge integrated and delta-sigma and old-skool tubes, discrete, and multibit. Bottom line, we’re open to anything.
There are only two things we don’t do…one because we can’t, and one because we choose not to.
- The one we can’t do? Super inexpensive products, say at Chinese domestic market pricing. We can’t do, say, a balanced tube hybrid amp for $49. Now, there’s nothing inherently wrong with this. We’ll leave the value analysis of these products to the purchaser. But the fact is, we can’t do this with US-based labor, manufacturing, and support prices.
- The one we won’t do? Car- and house-priced products. Again, nothing inherently wrong with this, assuming you have money to burn. But it’s not us. We’re simple. We design a product, figure out how much it costs to make, and use a standard multiplier to calculate its retail price. This has led to some people saying we are “not asking full value,” which we believe is code for “Sitting back, stroking your chin like Dr. Evil, wondering just how much you might be able to soak the customer for.” So, to be frank, **** that.
Beyond those two things, expect us to explore the entire universe of audio…and see what comes out of it. We’re having fun. Maybe you’ll like some of what we come up with. We certainly hope so.
“But You Shouldn’t (or Should) Be Doing…”
Okay, now let me anticipate some of your objections to this “fun from any angle” philosophy. Most of this boils down to a Vulcan/Victorian mantra of “You should not be doing this because…reasons.” Nothing against Vulcans or Victorians, but when I think “fun,” neither of them come to mind. Not gonna be inviting any of them to a party anytime soon.
Enough dissembling. Let’s dive in.
“You shouldn’t be doing tubes/discrete/multibit because it measures like ass,” is the first objection. And yeah, tubes, discrete designs, and multibit DACs typically may not measure as well as the best of the op-amp/delta-sigma based products out there, but to call it “like ass” obscures the reality. It also closes off one fun avenue of audio investigation. It’d be like if Willie Wonka found a new tributary of the Chocolate River and decided not to explore it. Or if he’d decided not to take the Great Glass Elevator into space (not making this up, read the books.) And closing off investigation is anti-science. Fact is, discrete can measure better than op-amps, especially when used in low-noise or current-output applications such as in a phono preamp or DAC I/V converter. These are usually heroic approaches, though, with high complexity, high power dissipation, etc…but discrete can be better. Tubes, yeah, well, the overall loop gain is low, so it’s gonna be hard to get great measurements. Multibit…again, with heroic designs, performance can be high (and maybe not so heroic, heh heh…the golden age of multibit may yet be ahead of us). So we continue to explore all avenues, to see where it leads us, to have some fun, and (hopefully) come up with stuff that you like, too.
But that’s not all. Oh no. There’s a ton of different opinions out there:
“You shouldn’t be doing op-amps and delta-sigma, because measurements aren’t all that matters,” is the other objection. Some people have an inbuilt aversion to anything that uses op-amps, pronouncing them all crap. To that, we’ll say that they probably haven’t heard the latest stuff, which can be very good. Same with delta-sigma DACs. To say you can’t get good sound and good measurements is super-silly. The design may not be super-challenging; you may have to work hard to differentiate yourself from other companies using the same chips, but to say it can’t sound good is just plain wrong. Plus, measurements of all types of our gear have reaped benefits across the board—lower noise, higher performance, higher power…and we’re committed to it enough that we’ve held up production of an important new product because it didn’t meet the specs it was supposed to be capable of. Not that the 6-8dB change would be audible, but because it meant something was amiss with the overall design. Measurements aren’t evil. They’re a tool, to be used with blind listening, to improve all products.
But it gets deeper. Lots of people tell us we should be doing something.
“You should have a more expensive line,” is what we hear most often. To which we say, “No ****ing thank you.” Because we suspect that much of the car- and house-priced stuff needs to be shipped with its own psychologist, to help continually convince the purchaser they made the correct choice. Plus, it’s not our bag. We don’t really know how to make stuff much more expensive, unless we’re getting into super-crazy chassis work (which doesn’t improve the sound) or completely insane stuff like ovens to control operating points (which we haven’t needed). So no. We don’t need a more expensive line. If you want to spend more money, there are plenty more companies to choose.
“You should do power conditioners/decrapifiers/linear power supplies/cables/insert-other-audiophile-nervosa-object here.” Again, no. We don’t do power stuff. We do a few short cables, because 6” cables are kinda hard to find. We did a USB decrapifier when USB ports kinda sucked, but that time has passed. We don’t get into nervosa stuff. Again, it’s not us. Again, there are plenty of companies that will sell you these products.
“You should do a broader line that includes…” yeah, and our line is already plenty broad enough, when you get right down to it. We understand that we can’t meet everyone’s needs, and that if you’re looking for a big-power amp or a multichannel amp or a whole bunch of other stuff, our line may not meet your needs. We’re sorry. We’ve grown about as fast as we can responsibly, with no investment and beginning in a garage about 10 years ago. Maybe our line will expand to include your needs someday. We’ll see where our designs take us!
So What Does This Mean For the Future?
In short, it means we’re going to keep exploring both avenues—highly integrated, good-measuring products and discrete, tube, and multibit stuff that doesn’t measure as well. We’re going to have fun while doing it. And we’ll be bringing fun, high-value stuff to you that we find interesting, and hope you do, too. It really is as simple as that.
“But that sounds like shooting in the dark,” some might say. “What do you expect from such a scattershot approach?”
In short, we expect better results across the board. By not closing off any avenues of product development, we’re open to the widest array of possibilities…including some really cool surprises we might not have been expecting.
And, no matter the approach, audio can be pushed farther. Maybe a lot farther.
- The performance of our tube hybrid stages has really evolved over time—Lyr 3 is almost 100x better in terms of THD+N when compared to the original Lyr. Saga and Saga+ measure almost like IC designs. Both have very low noise floors. And, at the same time, both are noted as sounding quite nice by those who prefer a subjective approach. Who says we can’t take this even farther?
- Although we haven’t done a heroic-approach multibit DAC, the possibility for high performance is there. A quick look at the performance of the golden-age PCM63 (see Stereophile’s review of Theta’s Gen 5 processor) confirms there’s a lot more potential there. And Mike, Dave, and Ivana are always working on something crazy. So who knows? If we come up with something that we believe sounds good, and doesn’t cost like a car, we’ll see what you think of it…and we’ll be able to do it on our upgradable platforms, so you don’t have to buy a new DAC.
- Op-amps and other new-production, high-integration parts give us a whole new toolset to work with. This means much, much higher performance is available for insanely low prices. The performance of Bifrost 2 is in no small part due to a great balanced op-amp, and the new TI switching supply parts give us huge power output capability in Hel, and new feedforward op-amps from TI help us deliver class-leading performance from Magni Heresy. But there’s a lot more to play with here, including compound and hybridized designs that could push performance even higher.
- And, although we haven’t done any completely insane, heroic discrete designs, this is where we’ll go to smoke the performance of the best integrated stages for an all-out phono preamp. Getting near -100dB THD+N with 60dB of gain…yes, this is doable. And yes, I know, this is for an antique sound reproduction method that might seem completely stone-age when compared to the convenience of streaming 24/96 from Qobuz, but hey, a lot of people really like it…and there’s a lot of spinners out there.
- At the same time, you’ve probably seen a lot of evolution in the way we do discrete designs over the last year or so. Much higher integration with multiple transistors in a single part, and much tighter matching thanks to die-matched parts in those same multiple packages. Thank Nexus and its need for matching, but the benefits are moving all the way across the line. And the devices continue to evolve. There are some amazing new matched N and P-channel MOSFETs that I need to take a look at. (Yes, MOSFETs…they have changed quite a bit, so it’s time to look at what we might be able to do with them.)
- Aaaaaaaanndd…there are the entirely new vistas. Is anyone else playing with stuff like The Gadget, with dynamic re-tuning and other functionality to bring entirely new controls to audio? We’re also going fast into some really crazy analog processing functionality. And another trojan-horse-type device that you’ll most likely see this year. We’re actively looking at ways to give you meaningful-but-transparent ways to control the listening experience, regardless of whether you’re a “hard-core tube guy,” or just want “the measurements, please.”
Also, no matter what we’re doing from now on is run through three stringent test regimes—one on the Audio Precision APx555, one amongst our trusted listeners for early feedback, and one in our Blind Listening panels. Because, despite the recent closing of the Schiitr (for a while, we’ll be back when the virus is under control), blind listening continues apace. It’s becoming an important part of product development.
In fact, as I probably mentioned, one upcoming product had its direction decided by blind listening. We built two prototypes, one with an approach similar to what we’ve used in other products, and one with a new approach. I took both of them home over the weekend for sighted listening, and thought I preferred one, but it was only a moderate preference. I gave them to our internal trusted listeners, and they also thought they preferred one, but with milder preferences. We didn’t share our picks until we reconvened…and they all liked the one I wasn’t as thrilled with.
And yes, I know, this isn’t blind listening. That’s where we started the blind listening. We set up a level-matched, double-blind test internally, and we all went through the same process. Surprise surprise, we actually picked the ones we liked best during sighted listening!
Yes…the blind results were congruent with sighted.
And still, our other listeners preferred the one I didn’t like as much. So we took the whole setup to the Schiitr and let the public have a listen. This was at the last Schiitrmeet event (hopefully we can get back to it soon). The public consensus was close to 50/50, with a slight preference to the one I didn’t like as much.
So here’s what we did: we decided to make the one that most people liked. Not the one I liked.
(Again, the differences are very small—see Lighted by the Blind—and I’m happy with the way we went, because I am not the High and Mighty Arbiter of All That Is Right And Good In Audio.)
Summing up? It’s simple: we'll keep exploring, we'll keep having fun, and we'll keep making things we hope you have fun with.
That’s how we’ll continue to do things around here—exploring all avenues of audio reproduction, staying away from car-priced stuff, running it through the triple gauntlet of tests, panels, and blind listening, and bringing the fun, interesting stuff to you.
(And maybe Thunderdoming some of it, when we can’t decide ourselves.)
We hope you enjoy!