Chapter 38:
Wyrd Schiit
“Hey Dave, is there any value in reclocking a USB datastream, like you guys used to do with SPDIF on Theta’s Time Linque Conditioner?”
Dave looked confused. “Wellllll…it’s not really the same…with USB, it’s more the hub chip repeating the output…but the hub chip uses, uh, a crystal oscillator, so it’s…well, maybe.”
That was pretty much the entire design brief on Wyrd.
We’d been talking about USB as an audio interface, and the increasing impact of USB port power management on USB-powered DACs and USB interfaces, and I’d just had a random thought about a little box to both power and re-clock the USB datastream. I mentioned that idea, pretty much in the same sentence as described above, to Dave as we were on our way out of the Schiitbox.
“So we might be able to do something like a TLC for USB?” I asked.
Mike groaned. “We have enough to do.”
“I know, but, like, for when we get bored (or frustrated, as in the above chapter.)”
“It’s USB. It’s like turd-polishing.” Mike said.
“Mike, in case you haven’t noticed, I’ve been using USB input at shows lately. Our USB Gen 2 input is really, really good.”
Mike grumbled something under his breath. “We still have too much to do.”
I sighed. I knew. Ragnarok still wasn’t running right, even after several firmware revisions. I had the melted-into-the-fiberglass-insulator MOSFETs to prove it. And a USB reclocker/clean power-er wasn’t something I could do (or should do.) I had plenty of analog projects to work on. But what’s the harm in talking about silly ideas, I thought.
And that, I figured, was that. We’d drop the idea, and pick it up again later, if we had time.
Except Dave.
“Just Listen To It”
About a week after our short conversation about a USB widget, Dave showed up at my place for a barbecue. Mike was coming as well, but he hadn’t arrived yet. Dave was carrying a small box of stuff. I didn’t think much about it, because Mike and Dave are always swapping various weird digital things that I don’t know anything about.
But when we were inside, Dave said, “About that USB thing you were talking about…”
“Yeah?” Maybe we would get somewhere on this, I thought.
“I made one,” Dave said, pulling a small green board out of the box.
I stared at it, then blinked a few times. It wasn’t possible he’d already designed and built the one-line “what if” product I’d just been talking about, I thought.
“What do you mean?” I asked.
“I built one. A low-noise USB power supply and hub.” Dave pushed the board at me.
I took it and looked at it. The board had Dave’s signature hacks that he did as he tested and optimized a new design—a ground lug soldered on, a few capacitors tacked here and there. Trailing off of it was a hacked old-skool gray USB cable, with a couple of capacitors and inductors hanging off of it.
“Does this work?” I asked Dave.
Dave nodded, then shook his head. “Yeah. Mostly. I mean, I tried it on a couple of machines, but I don’t know if it’s really meeting USB power spec, or if it works on everything--”
I waved a hand. With Dave, nothing ever “just works.” There are always caveats. If it wasn’t for Dave, we’d have a lot more undiscovered problems. As Mike says, he’s the “trees” guy. We’re the “forest.” But, as a consequence, few things are ever “done” for Dave.
“So I can plug it in and try it?” I asked. “Right now?”
Dave nodded. “Sure.”
“Does it sound better?” I asked.
Dave broke into a big grin. “Just listen to it.”*
*Okay, big aside here, before I am eviscerated by the objectivists: I know there is no sane, rational way in which a USB repeater/clean power device can scientifically affect the sound of a system. USB, like all digital interfaces, is, in the end, digital. The packets either get there or they don’t. If they don’t, they are resent, if possible. If they can’t be received in time for a streaming application like audio, you get glitches and disconnections. That should be it. Period, no exceptions. So clean power and shapely waveforms should, theoretically, make no difference.
*And another big aside, for the subjectivists: Yes, I also know that USB, like any other digital transmission system, is essentially using imperfect analog waveforms to transmit very high-rate data (in the case of USB 2.0, theoretically 480Mbps, or about 300x what you need to transmit stereo 16/44 content, though overhead usually makes the actual rate much lower). But the fact is, it should not matter in the least how well-shaped or noise-free the waveform is—the packet is either 100% recovered, or it isn’t. If it wasn’t, you couldn’t simply copy several hundred gigabytes of data to an external drive reliably.
*And a final big aside, for everyone: If insanity is defined by the ability to hold two completely contradictory ideas in your head and accept both of them, then me, Mike, and Dave are all completely bonkers. Because we run into this same objective/subjective thing all the time.
“You’re not telling me it sounds better,” I asked Dave.
“Just listen to it.”
“But, there’s nothing that should…I mean, how would it change anything?”
Dave looked sheepish and shook his head. “Just listen to it.”
Okay. Fine. I hooked it up between my Macbook and the Gungnir/Mjolnir stack I use as a main reference system. But, before I fired it up, I dangled the weird hacked gray cable in front of Dave. “What’s this?”
Dave laughed. “Oh, that. That’s just an idea. I saw that some people are separating power and data in their cables. I wondered about filtering the power instead. Maybe try that first. No USB board.”
“Does it sound better?”
Dave was deadpan. “Just listen to it.”
I groaned. “Now I know you gotta be screwing with me.”
“Just listen.”
Fine. Might as well give it a true test. I unplugged the USB board and went back to the full-stock system (no reclocker/power supply, no hacked cable) for a quick listen. It sounded, well, like I expected it to.
Next, I used the hacked cable. I went through several tracks I use to get a subjective gauge on the sound, trying to hear a difference. If there was any, it was really too slight to notice.
“I don’t hear it,” I told Dave. “Sounds about the same.”
Dave nodded. “That’s what I thought. But the level of filtering you can get with that approach is really low. If you have a couple hundred millivolts of noise on your USB port power, this might cut it down by half. Maybe. You can’t put too much resistance in the line.”
“But the board is better?”
That big grin again. “Just listen to it.”
Fine. Now, I put the board in-line between my computer and the Gungnir and sparked it up again. If there was any difference in sound, I figured it would be in the textbook, “Sharper, more resolution” direction. But more likely, there wouldn’t be any difference at all. At least not in theory.
I clicked on the first track.
And sat back, a little shocked. Because I could swear I heard a difference—and not in the expected “sharper, brighter” direction. If anything, it sounded a bit smoother and more natural. At the same time, the noise floor seemed lower, like I could hear deeper into the track.
I looked up at Dave, my mouth open. Dave immediately doubled over, laughing.
Nah. Not possible. It was a trick. I was fooling myself. Or Dave had somehow put real-time processing into the USB stream (ha, ha.)
Still, as I went through the tracks, the difference was consistent—and consistently positive. Dave kept laughing as I shook my head again and again.
I had to be fooling myself.
I unplugged the USB board and went back to just a straight old USB cable. And suddenly it sounded a bit flatter, more one-dimensional. Dave kept laughing. I switched back, and it got better again.
“How does it do that?” I asked Dave. Thinking, at the same time, Nobody’s gonna believe this.
Dave shrugged and shook his head. The same thing he does whenever confronted with something that goes beyond logic. This is a guy who is much more analytical, and much more structured, than Mike and I combined. This is a guy who has designed uber-high-performance systems that deliver floor-of-the-analyzer performance, together with ultra-high complexity, for various systems far more ambitious than anything Schiit has produced to date. Of course, this is also a guy who has worked in an environment where they used RF to seal thick plastic, with water-cooled tube outputs and stray field that would light fluorescent lights in open air several yards away from the machine. So maybe that explains some of it.
“But…why?” I persisted.
Dave shrugged. “Lots of reasons. The power supply is super-low-noise. It’s using LM723 regulators, which are rated 4.17uV of noise, plus organic polymer capacitors, plus I’m using a tight crystal for the USB repeater. Lots of little things.”
“But it still shouldn’t matter,” I said.
Dave nodded. “Right.”
“But it does something.”
“Right.”
I sighed. Okay. Run with it. “So this is, like, done?”
Dave shook his head. “It won’t output the full 500mA USB spec, because you have a 16VAC wall-wart and it’s only rated for 500mA. We need something more like 6VAC, maybe 1A or more.”
“We could do 1500mA in that same package at 6VAC,” I told Dave.
“That would work.”
A sudden thought hit me. What if someone plugged a 16VAC wall-wart into this product? That might be a bad day. I shrugged. We could make the pin bigger, so you couldn’t use the 16VAC transformer on this product. The 6VAC version could plug into the 16VAC product, but that wouldn’t hurt—they simply wouldn’t work.
And then I realized…I’m thinking about this like a product. Like we were already going to sell it.
But, why not? Dave had already used the form-factor of our small products—the board was laid out to fit in a Magni-sized box. All we needed was a different transformer. And it would solve the USB port power management problem.
And it would sound better.
Which made me remember my first thought: nobody will believe this.
Argh.
Truth and Marketing
If Schiit was a purely subjective company, the dilemma of having a product that made stuff sound better, without having any rational explanation as to why, wouldn’t be a dilemma at all. We’d wrap it up in nice flowery language, throw in some pseudo-meaningful charts that showed the difference in power supply noise levels, and call it a day.
If Schiit was a purely objective company, the dilemma might not be a dilemma at all. Because we might have convinced ourselves that, even though there was a difference, there really was no difference, and so why bother making something that didn’t make a difference?
But as a company that uses both objective measurement and subjective listening, it’s not so clear. We could do the pure subjective thing with the words about how you’re transported in space and time to a wonderful world where unicorns dance and crap like that. Sure. We could.
But that isn’t us.
And that isn’t honest. Because, you know what? We’re really talking about small differences here. It might not be important to a lot of people. It can be easily dismissed.
But for other listeners, it might be big enough to be significant.
So how to market it?
“It does solve real problems,” Mike said, when I was angsting to him. “I have a laptop with power so bad it’s noisy—as in, you hear the noise even through a Modi and Magni. This kills it dead.”
“And it does make those weak USB ports usable,” Dave said.
“So does a powered hub,” I countered.
Dave shrugged. “And it sounds better.”
I groaned.
I really didn’t want to get into the “hey, trust us, it sounds better,” theme. Especially if people thought we were trying to sell a $99 add-on to a $99 DAC like Modi.
Aside: that is pretty silly, when you think about it. And less silly, when you think about it some more. Think of the Wyrd as the “separate outboard power supply” for the Modi.
Another aside: it’s still pretty silly.
“And what are we going to call it?” Mike asked.
I groaned again. That was another question. What was the name? I hadn’t figured that out yet. And that was strange. We usually have the name—and the descriptive copy—done long before the product is heading towards production. And, as we were having this conversation, the USB board was on its second rev and considered production-ready.
Name. And copy. Usually not a problem. But in this case, it was.
Wyrd Schiit, and Wyrder Marketing
Inspiration struck on the name when I was searching for terms relating to the Norse concepts of “time” or “stability.” This led me to Wyrd, which is a term used for the complex interconnecting web that binds all things, living or otherwise. It is so intricate, it is much more powerful than the concept of “fate.”
In fact, the word “weird” is derived from “wyrd.” And when you pronounce it that way, everything falls into place: the weird fact that it does seem to make a difference, even when it should not. The complex interconnectivity of USB. And the connection to Yggdrasil and other Norse concepts.
Cool. So we had a name.
What about the marketing? Well, after some more thought, I decided that the best way to portray Wyrd was as a product that solved real USB problems, and stay away from any subjective claims at all.
“But it makes a difference,” Dave said.
I laughed. Dave virtually lives with an Audio Precision SYS-2722 and one of our Stanfords. To hear him say that, with no real reason why it should be so, is very funny to me.*
Aside: actually, there have now been measurements of lower jitter with Wyrd from independent sources, but we’re still talking about levels that shouldn’t matter—which is why we never published our own.
“I know. But let’s let everyone decide for themselves.”
“Ohhh…kay,” Dave said, doubtfully.
And that’s what we did. We first showed Wyrd at TheShow Newport in final production form. At set-up, we ran into a problem with a new, low-cost Windows 8 laptop—one that had not one single port that could run a Modi. Wyrd fixed that right up. A perfect application for it.
Wyrd was one of those nice, uneventful launches—the kind you want to have.
Aside: I’ll probably have to add another chapter here sometime about all the pain associated with building Ragnaroks, and the reality of having our first real production-line product in house (to date, everything else is a simple build—drop a board in a box, done. Ragnarok’s build sheet is several pages long, single-spaced, and has 6 separate productions stations involved.
A few weeks after the show, we started shipping Wyrd to little fanfare. It’s a good-selling product, though, probably because it’s still the only product we know of that both cleans up the USB power, and repeats the signal with a precision crystal oscillator—and does it for half the cost of other products that are simply power supplies.
And yeah, I still use one with the Gungnir that I have at home.
Annnnndddd…not because I have noise, glitches, or USB port power problems.