2023, Chapter 17
Runner’s High
Writing these end-of-the-year chapters is one of the most fun things I do. Yes, I know, I’m weird. But I really like looking back on what we’ve done, and talking about where we’re going. In its own odd way, you can think of this as our “corporate roadmap.”
In some years, these codas have been triumphant. And some years have been cathartic. Like last year, when I started with “slam the door!” As in,
let’s get this year behind us. The future must be better. Because 2022 was rough.
So was this year better?
Yes, in many ways.
It was also a ton of work. Serious hard work. Both on products and on the company. But the feeling this year isn’t exhaustion; it’s the glow of a job well done, it’s the satisfaction of doing something difficult and well…even as the work continues. Hence, “Runner’s High.” I’m working hard, but feeling great while doing it.
“Well, yeah, dood, you introduced like 500 new products this year, of course it was a ton of work,” someone says.
Yeah, but actually we’ve been on a heavy product development path for a while now. Last year we introduced 10-13 new products, depending on how you counted. The year before that, 8. This year, 10 again. 11 if you count the card.
So the numbers aren’t out of line. It’s the
new new products!
As in,
7 new names this year!
SEVEN!
Let’s compare. Last year it was 2. The year before that, 1. Everything else was a 2 or a + or a ++ (argh)…in other words, a revision. Maybe a very big revision, as in the case of Lyr+, but still a revision. Still based on a template that came before.
In fact, there’s never been a year that we’ve introduced so many entirely new products, with entirely new names. So getting seven in a year is a big, big thing.
Sooooo… let’s start by talking about that a bit.
The Magnifiscent Seven?
At the end of last year, I promised the following products:
- Redacted 1 of Crazy Idea 1
- Redacted 2 of Crazy Idea 1
- Redacted 3 of Craziest Idea Ever 2
- Urd (yes, finally)
- Mjolnir 3 (think Folkvangr)
The first ones that appeared were Rekkr and Gjallarhorn, the Redacted 1 and 2 of the Crazy Idea 1. Then Syn, Redacted 3. Then Urd. And then there were three other new names!
Probably best to go through them…
- Rekkr. The tiniest speaker amp in our line. Fits in the palm of your hand. A pair win the award for Cutest Monoblocks Ever. Provides plenty of power for most desktop listening. Surprisingly capable with efficient speakers. Heck, 5 of them overpowered the crowd at Nueces Brewing, a massive 5000-square-foot space, when we were demoing the Syn with some midrange Klipsch speakers.
- Gjallarhorn. The desktop speaker amp that people have been asking for for years. Together with Rekkr, it dramatically expanded our line of power amps. Capable of being run in stereo or mono, this little amp has proven surprisingly popular.
Why are these two products given their own section? Because they were launched under a common theme:
Less Power, More Better.
And the theme worked. It helped counter the clamor for moar and moar from Class D amps, which yeah, are powerful and cheap as chips…about as interesting. The reality is most applications don’t need a ton of power, and a great Class AB or Continuity amp is a solid alternative.
But that wasn’t the end of themed introductions.
Syn, my Craziest Idea Ever, was launched under the cheeky banner of
Surround Set Free. Reflecting my own less-than-ideal experience with surround products over the years, Syn was my way of breaking off from the lock-step death-march of home theater “progress.”
- Syn. Surround that works with everything and is never obsolete. A dangerous idea in itself. But oh yeah, it’s alsoa headphone amp, a DAC, gaming microphone input, remote control, and a discrete main gain stage. Syn was the only product that ever had me worrying that I was tilting at the wrong windmills. It’s a niche product, yeah, but the people that have them looooooove them. And I love mine. So it stays. Neener.
Then we get to the holy-crap-you-ever-gonna-finish-that-product product, and another new product category, the Unison USB hub: Urd.
- Urd. A CD player with USB out? Yeah. But it also has USB inputs!! Holy heck, what is this thing? This thing is one of Mike’s greatest ideas ever, a CD player that also is a Unison USB hub. That means it can integrate your whole system—streamer, computer, CDs, etc—and also allow easy comparison of media, new and old. Bonus: it’s tiny, and fits the form factor of Kara and Freya perfectly.
Then all the unexpected names:
- Kara. An all-out Nexus balanced preamp with a headphone amp? Yes. You may have expected a Freya S2 at some point, but instead you got a Kara. This exceptional preamp deserved its own name. With a boost to much higher rails for much higher performance, all-new single-ended gain stages, and a super-advanced headphone amplifier, Kara is the solid-state preamp Freya+ wishes it could be. Of course, Kara also wishes it could be Freya+, but that’s another story.
- Skoll. A high-end, balanced, no-feedback, Class A, high rail, passive-RIAA phono preamp…with a remote control? That’s also the least expensive balanced phono preamp in the world? That also has world-class performance? For $400? Yeah, nobody saw this one coming. Not even me. After being asked for years by Alex, Tyler, and Rina, I finally caved and did this. Because this is what we do. It wasn’t launched with a logline like Syn and the tiny power amps, but if it had, it would be Pho-no? Pho-yes!
- Midgard. Desperation is the mother of invention. Or so it seemed during the development of Midgard, the first salvo in my quest to eliminate op-amps from our analog product. Getting spectacular performance was easy, but figuring out a real reason for the XLR output proved tough. In the end, we got something that delivers an insane amount of value and performance from a discrete, low-feedback design, and introduces Halo™, a new mixed-mode motion feedback topology that may increase performance at the driver.
But the seven wasn’t the end. Not at all.
Big Changes to Old Friends
“Wait a sec!” someone is saying. “You didn’t even talk about some of the biggest stuff you did! What’s up with that?”
What’s up with that is they aren’t new names. They’re revisions of existing products. And yeah, some really really push the definition of revision. Mjolnir 3 and Magni Unity particularly. So let’s cover the rest of the new gear:
- True Multibit Unison USB DAC Card. We tried to introduce this one in 2022, but a shortage of DSPs meant that wasn’t gonna happen. It finally debuted in 2023, which was a very nice update to the internal DAC card that fits Asgard, Jotunheim, Lyr3, and Ragnarok 2. And it remained the same price! Unlike all the other stuff that people decided should be 2.5x the price because of “inflation.” More on that later.
- Mjolnir 3. Take Mjolnir 1, take Mjolnir 2…and forget about both of those. How about a real, Class A, choke-input, dual-mono, no-compromise headphone amp? Yeah, it’s not the most powerful thing out there, yeah it’s not the best-measuring thing out there, but holy heck does it sound good. This one went from a limited-production product to something we want to keep around, because there’s a whole lot to life beyond a number and a couple of graphs.
- Hel+. Seems a lot like the old one, but it’s nothing alike—now with 3 gain levels, “invincible” protection, and a three-year warranty. I wasn’t so sure about putting in the time on this one, thinking gamers were probably going for the latest ooh-shiny Bluetoof headphones, until Evan (one of our gamers) mentioned, “Oh yeah, no serious gamer ever won a championship with Bluetooth—they’re all wired.” So if you’re serious, if your gaming is mission-critical, Hel+ is for you.
- Magni Unity. We brought the stack into the world. Now we’re taking it out. Want to keep up with inflation? It’s time for a one-box, no-compromise Magni with a modular DAC card. Benefits? Less obsolescence, less cost, higher performance…and it includes an all-new, high-linearity, low-loop-gain discrete current-feedback stage. It’s the end of the Magni Bifurcation. And the end of the stack (providing you need only USB input, of course!)
- Yggdrasil MIB. The least popular Yggdrasil ever, Yggdrasil More is Less, was redeemed this year by the introduction of new, high-measuring TI DAC11001B chips. Turns out that high measuring and good-sounding can go together, just as we proved with Kara, Midgard, and Magni Unity this year.
Whew. 10-12 products (depending on if you count cards), 10 total boxes, 7 all-new names.
That’s a big year.
“Well, that’s a crazy pace,” someone says. “Can you keep it up? Are you getting burnt out?”
LOL. Actually, development-wise, this year wasn’t a ton of work. And even if it was, it wouldn’t matter. Because I love doing this.
No. Let me expand:
I looooooooove doing this.
I can think of nothing better than coming up with new crazy audio stuff. I wake up thinking about audio. I go to sleep thinking about audio. I get wacky ideas in the middle of talking to someone and jot them down for later. I build 3 or 4 prototypes that go nowhere, or get put on hold, for every product we make.
And between me and Dave and Mike and Ivana, we have years-long list of things that may or may not happen, because it may or may not be the right time, or it may never be the right time, or suddenly it’s the 100% perfect time.
Am I weird? Absolutely.
But anyone who’s worked in a creative industry and truly loves what they do completely understands.
So don’t worry about me getting burnt out. And don’t worry about us running out of ideas. And, most of all, don’t worry about a product development cycle that includes 10 or more new products a year.
Because that’s what next year looks like: pretty much the same pace. And a ton of new ideas. Some very, very wacky indeed.
More Than Products–Key Technologies
I know a ton of you are waiting for the classic What We Did Right/Wrong/Etc sections, but sorry, I’m gonna nerd out a bit. So bear with me.
Because it’s one thing to make a list of new products, even a long list of really new stuff, and it’s a whole ‘nother thing to talk about the tech behind them. Because that’s where you get a much clearer idea of where we’re going, what’s driving us, and why we introduce the things we do.
And oh boy, did we see a lot of new tech this year:
- Unison USB Host. With the introduction of the Unison USB host, we have the only product that is not a general-purpose computer that is a USB host. What does this mean? Lots of stuff. The main thing is that, instead of using Android or Linux for our USB output, and taking the concomitant support, updates, security patches (and the nailbiting that ensues when the patches stop coming), we have a 100%, totally secure, totally non-connected USB host that is dedicated to a single function: the best possible audio. It makes a real difference—and it will certainly enhance your peace of mind relative to having yet another computer to deal with, and yet another connected device on your network.
- (The Return of) VFA. VFA stands for Voltage Feedback Amplifier. This is what we started with in the original iteration of Magni, and further refined in Magni 2 and 2 Uber. It’s the most popular discrete topology for power amps everywhere. And, in the case of Syn, it was the only option that provided the performance we wanted, together with the fine-grained control we needed for the Width and Presence adjustments. So we took what we learned with the first 2 Magni generations much, much further—to provide the highest performance voltage feedback stage we’ve ever done, while retaining a fully discrete topology.
- Equipoise. Equipoise is the internal name we use to describe our super-simple, complementary-pair, error-cancelling differential topology we first used in Freya+’s buffer stages. This year, we took Equipoise much farther, using it as the basis of Mjolnir 3, and as the super-high-gain, super-low-noise stages in Skoll. It’s a powerful, versatile, and simple topology that provides good performance—and sounds fantastic.
- High Precision, High Linearity CFA. This unwieldy moniker is what we’re using to describe a refined, much more linear current-feedback (CFA = Current Feedback Amplifier) topology that debuted in Midgard and was quickly extended to Magni Unity. The result? Better measurements without higher feedback, by linearizing the open-loop topology.
- “Invincible” Protection System. Excuse the scare quotes, because calling something “invincible” without caveats gets you, well, a date with Murphy’s law. But in this case, we needed something industrial-strength to protect Kara’s high-voltage headphone output stage from common faults, like plugging or unplugging headphones with the volume turned up. So we went all-out with a microprocessor-based sensing and relay-based protection system that ensures nothing bad can happen to Kara…or the headphones connected to it.
- Halo™. Our new controversial output topology—a mixed-mode motion feedback system that attempts to bring the driver into the feedback loop, for possible improved acoustic distortion measurements. Yeah, that’s a whole lot of weasel-words, even more than the scare quotes above. Because, while we do have a Ph.D like some audio companies, and while we could spend years getting some research results, we prefer to work on more interesting stuff. But hey, if you want to do acoustic measurements on a Halo product, you get one for free! This tech appeared first on Midgard, then on Aegir 2.
Sounds like a lot? Oh hell guys, wait till you see 2024. If only half of what we have planned works out, it’s a crazy new year.
And yeah, I know I say that every year, but holy moly…I mean, Singularity is only a relatively small part of it!
More on this later. For now, let’s talk turkey. And not in the Thanksgiving sense.
More Than Key Technologies—Solid State Advancement
Magni, Midgard, Pietus, Mjolnir 3…it wasn’t intentional, but it turned out that we really moved the needle in solid state, discrete designs—at all price points! Heck, you can even count Kara in on that.
But this wasn’t something planned, heck, it wasn’t something marketed—it was something we just did, and it took a colleague mentioning, “Holy crap you guys pretty much single-handedly changed the landscape in discrete design this year.”
In retrospect, this makes sense when people are kinda scared of tubes. And it’s a great thing that tube scarcity brought real advancements in solid state. Of course, people’ll do the usual thing and get bored with one thing and move on…maybe going go back to tubes, maybe starting next year.
The Cold Equations
This year was the first time we really had to grow up a bit, and acknowledge the reality of, well, money and stuff.
I mean, you know we started in a garage, right? You know we’ve never taken VC money or any other outside investment, right? You know that other than the mortgage on our Corpus Christi facility, we have no debt, right?
Because of this, we’ve been able to operate with some impunity, beholden to no deep-pockets dood or faceless evil Board of Robed Higher-Ups or profit-squeezed bankers, just kinda doing our thing. And, to be clear, that’s how we’ll be able to continue to operate.
It’s just we have to
plan a bit more.
And, at the same time,
not plan even more than that.
Here’s the thing: coming out of 2022, I said, “we need to align production with demand.” This super corporatey-sounding pronouncement means simply, “stop building crap until we need more.”
Sounds simple, right?
Yes. Right.
Except…we had agreements in place with some suppliers to keep producing at a certain rate.
Cue the ominous music. You know where this is going.
These agreements were mainly hangovers from 2021/2022, when we needed to do what we needed to do to ensure supply of some critical parts. But, as 2022 ended on a slow note, and 2023 showed no giant rebound, those agreements became problematic.
Of course we talked to them. Of course we renegotiated. And, in the end, to use the passive voice,
understandings were reached, even if those understandings didn’t make anyone happy. Complicating things were some overbuying stupidity on both sides, also hangovers from the pre-ERP days. And complicating the complications was some really dumb crap like buying a reel of 3000 parts, using 1000 of them, and then shelving the rest. And no I’m not talking resistors. I’m talking DACs. And complicating all those other complications was the apparently psychotic buying spree of one employee that resulted in us having thousands and thousands of transformers too many—which the transformer manufacturer happily made us take in early 2023, and is now less than happy after we found another manufacturer.
More on this schadenfreude later.
But, to oversimplify, we went into 2023 on a single note: suppliers ruled. They had wayyy too many orders. They didn’t have time for your nonsense. Your quote was your quote. Take it or leave it. Yes, even if it was 2.5x the quote you got 4 months ago. The excuse: always inflation.
I was pissed. Inflation is not 2.5x quotes. Inflation is 10-20%.
Maybe. More on that later too.
But nobody listened. The price was the price. No negotiation. GFY if you don’t like it.
It got so bad that we looked seriously into bringing in our own SMT line. And, who knows, we may still do so. Though whenever you start looking at going vertical, you have to consider the cost of leasing, staffing, maintenance, downtime, and supplies. Many companies only look at the lease and staff cost and call it good. Many companies don’t get anywhere near the utilization they plan. Going vertical is a big decision, and it shouldn’t be taken lightly.
Luckily, we have good staff and enough suppliers and good enough relationships with critical suppliers that we were able to get through the first half of the year. Though it got rough. Rough enough to look outside the USA for suppliers for the new Magni chassis. We never went through with it, though.
But then, in late spring, things started to change.
Just as I expected.
Now, this isn’t to say I have any magical predictive ability, or some crazy supernatural foresight. It’s just that when things get too expensive too fast, everyone looks for alteratives. Or stops buying. Which kills demand. Which…makes things change.
As I said to my first boss out of college, when he was crowing about how young engineers could no longer afford houses, and he owned 9 of them…”and that doesn’t worry you?” The blank look on his face preceeded a LA housing market crisis that didn’t end for 6+ years, and saw homes go from $250K to 80K in a couple of years.
So, to the nuts who tried to squeeze us:
it’s always like this.
You will always get yours.
And so, yeah, in summer, suddenly everyone was like, “Well, what about that quote?” “When is that order gonna come in?” “What does your schedule look like?”
And my favorite: “Now is a good time to get started.”
Again, hint to the buttnuggets: everyone in business knows this phrase means, “We are hurting, please give us some money.”
Yeah. Screw you.
Because we are managed extremely conservatively and have no debt, we can sit this out. So we did.
The result? Now some quotes are 1/3 of what they were. Prices have cratered on assembly and are coming down on metal. New suppliers for transformers have put a crimp on our old supplier’s price increases.
Yeah. Inflation is real. 2.5x quotes are not.
And we’re now getting back to reality.
But the reality is still we have a ton of transformers and we’re a bit overweight on some products. Hello, Magni Heretic. Ah well. Nothing is perfect. The trend is in the right direction. And things are much better.
But you’re still going to see Brown Wednesday pricing on stuff that is truly closeout. And even that small concession to the consumerist side of the end of the year pains me. At least it’s not a real sale, a discount on ongoing product. So we’ve made progress relative to last year.
What I’d really like to see is no Brown Wednesday at all, or, as the ultimate middle finger to the buy buy buy mentality, a price increase on Black Friday. I’m not sure if we’ll ever do that.
In fact, it’s more likely that we’ll continue some kind of closeout deal at the end of the year.
Why?
Blame accrual accounting. Or, as I call it, “The Cold Equations.”
I hate accrual accounting. Only in accrual accounting can you have big profit and no cash. But you’re forced into accrual accounting once you reach a certain business size. I don’t know the exact reason, but I’m sure it’s because accrual accounting allows for more bookkeeping ****ery than cash accounting, which is super simple and wayyyyyyyyy closer to the way actual normal businesses run by actual normal humans operate.
So, let’s be clear: with accrual accounting, we can go into December with a big profit and weak cash. Which is dangerous because the IRS only cares about profit, and wants their cut in cash. It’s pretty easy to get out over your skis and get to a point where you don’t have the cash to pay your taxes. Argh!
Also with accrual accounting, you can have a giant sale that generates a bunch of cash, but also reduces your profit, so you end December with much more cash and a much lower tax bill.
Huh?
Yeah. Again. This is dumb.
But this is how the game is played. So we play it. So it’s more likely there will be Brown Wednesdays in the future. The smaller the better, though. Definitely just close-out.
Because, after all, we’ve finally aligned production with demand. Which is great for us. But it also means maybe some out-of-stock situations for you.
Sorry about that.
Why don’t we get to what you’re waiting for?
What We Did Right
OMG, it is so tempting to say, “Pretty much everything.” Because we did an insane amount of things right this year, including:
- Significantly updating the line
- Introducing a lot of new product categories
- Introducing significant new technology
- Having a ton of flawless launches
- Renegotiating a ton of agreements that were holding us back
- Playing hardball with “inflation” claiming vendors
- Realigning production with demand
- Despite all the economic craziness, maintaining our sales
And we did a bunch of things that were less visible, like anticipating the tube slowdown and not over-investing in tube products. Heck, Kara is almost entirely a result of realizing that
hey, there are gonna be a ton of people out there too uncertain to get into the tube thing, so maybe best to do a great solid state preamp. We also postponed the introduction of a couple of tube products.
You could also say that Mjolnir 3 was a result of the Great Tube Uncertainty, but it’s also a hearkening back to the original Mjolnir as well. In any case, it’s a great-sounding amp, and we’re working on applying what we learned with Mjolnir 3 to future products.
And maybe it’s best to leave it at that…just say this year taught us a lot, and we are much stronger for it. Without having company-imploding mistakes. This is big. This is fine. This is a fantastic year.
What We Did Wrong
Not much. Maybe we waited a bit too long to get aggressive with suppliers, maybe we were a bit too nice. But that’s how we are. We aren’t hard-charging, Type A executives. We’re more go-along, get-along types.
And maybe we should have anticipated the demand for Mjolnir 3 better (but how?)
Also, it took a while to realize we needed to adjust advertising.
Yes, advertising.
Let me explain. Google Adwords started as the holy grail of the advertising industry—a completely measurable, traceable platform where you knew exactly what your ads were doing for you.
Unfortunately, over the years, it has devolved into a much squidgier, hard-or-impossible-to-quantify thing.
Why?
Because Google likes to take credit for everything. Anyone maybe possibly saw your ads in the last few months? Oh yeah that’s a sale. So let’s serve up a spectacular ROAS (return on ad spend) and conversion percentage for you.
Except that’s bull. Especially when you have a strong brand, it means Google is taking credit for sales that simply would have happened, with or without them.
Sigh. Why do humans always mess up something good?
Yeah, I know. A bigger conversation for another time. But to the point: we’re reducing spend on Google and increasing it on platforms we can actually measure. And that includes Amazon. Maybe we’ll end up selling a bit more on Amazon—but not enough to be beholden to them, oh no.
Where We’re Going
Oh gawd, what can I tell you without sounding like a cult leader or shill?
That 2024 promises to be the most insane year ever? Yeah, if everything works out. But we all know about how things have a tendency not to work out. So let’s start with what I can promise:
- Aegir 2. We already previewed this at the Schiitr, and soon it’ll be ready for prime time. A bit-more-powerful version of Aegir, now with Halo!
- Valhalla 3. Yeah, I know. It’s wayyyyyyyyyyyyy past time for The Fully Modern Tube Amp. Expect to see it by summer. Not sooner.
- Stjarna. Tube phono preamp. Limited run. Probably. Also by summer.
- Gungnir 2. Hello, Singularity Converter™. Our own high-bit-depth, high-rate delta-sigma modulator, based on our own math, our own filter, combined DSP/FPGA architecture, Unison USB—is this the first DAC ever by any manufacturer using 100% their own tech, USB to output?
And this is only the tease. If things go well, you’ll also see what we’ve been calling The Processor. Which is much, much more than The Gadget. Welcome to a whole new realm of audio control you’ve never had before, from any company, at any price.
And if things go really well, something much, much bigger…something that will send shockwaves through the industry.
As always, thanks for reading!