1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Schiit Happened: The Story of the World's Most Improbable Start-Up

Discussion in 'Jason Stoddard' started by jason stoddard, Jan 23, 2014.
  1. Miguel Barone

    I posted something about it on the thread ... impedance and phase curves are more important in general.

    But ... people here Is something special. Cats and "funny" posts are quoted hundred times, but technical info is passed by.
    Nothing wrong or right ... It's like it's. Then you see talking about decibels again and again.
    Last edited: Mar 12, 2019
  2. jimmers

    I wasn't familiar with @valiant66 Totem Mani 2 so I looked up the Stereophile review measurements, by their measurements sensitivity was 80.7dB @ 1M and 2.83V (nominal 1W, into 8 Ohms) so very low sensitivity
    also > 40 degrees phase angle at 4 Ohms impedance - pretty heavy EPDR
    https://www.stereophile.com/reference/707heavy/index.html :nerd:
    Last edited: Mar 12, 2019
  3. valiant66
    Yep. Brutal to drive. Recommended power is nominal 40-200W, Atkinson recommends at least 50w, and I’m I’m driving each speaker with two 75w channels (so effectively 150w) and I still run out of oomph with loud passages loudly played, or sometimes thermal shutdown when watching action movies. But it does what I value: nails the human voice, and soundstage, as well as my Rogers LS3/5As while providing deep, accurate bass. These days I can’t hear a damn thing above 10kHz, but I remember they did upper registers well, back in the day. And as standmounts they’re easy(ish) to move around and don’t take up a lot of visual mass in the room.

    I see from the reviews that the impedance rose with later models, but those isobaric woofers still need power.
    33na3rd and Derrick Swart like this.
  4. barondla
    When you are building the amps you always have a manufacturer's warranty :).
  5. audio philestine
    Schave the budget for more money for Schiit? :wink:
  6. Miguel Barone
    Well ... you really add information :thumbsup:
  7. wink
    US Blues likes this.
  8. Derrick Swart
    hey a dog!
    wink and AudioGal like this.
  9. Jason Stoddard
    2019 Chapter 5:
    Buzzword Bingo

    Or, A Semi-Complete Guide To Schiit’s Trademarked Tech

    “Oh gawd, they have gone full marketing-evil, naming and claiming everything!” you might be thinking.

    I mean, after all, if we have to have a glossary of made-up bullschiit, there’s gotta be something wrong, right?

    And yeah, don’t think I haven’t had some misgivings, as we name the technologies we consider important, and put that special little ™ at the end of them. During my agency days, I worked with many companies that would “name and claim” pretty much anything that wasn’t nailed down. The product line name? Sure, add a ™. The technology in the product line, sure, another ™, the additive in the product that gave it a special something, yeah, you better bet you need to name and trademark that too. The way the product is packaged, how it was shipped, the way the napkins were folded in the cafeteria…a whole forest of ™s.

    Aside: I’m kidding about the napkins, but not about anything else. We once had a company ask us to launch (a) their brand, (b) a sub-brand-name for a product line, (c) 6 product line name categories that were also fancy names, and (d) 3 different special formula ingredients in the products.

    Aside to the aside: if you’re an agency or marketing guy, you know this is full boat crazy, crazy like expecting a single ad to deliver a dozen different messages, crazy like “fully doomed to failure from day one,” crazy like “this marketing VP literally knows nothing about how people think.”

    Aside to the aside to the aside: What happened? We took the job. We told them how it was impossible to do what they wanted. We argued for focusing only on their brand, and a single trademarked name. We got them down to a three-tier system. Not great, but more manageable. The company is still around, and they’re still using our branding, so something must have gone right.

    “So why bother ™ing anything?” you might ask.

    Fair question. Let me go back to a key phrase above: the technologies we consider important.

    As in, not everything.

    Not napkins.

    Not even stuff that some companies might trademark. I mean, it’s entirely possible to come up with a fancy name for everything we do—TrueLinear™ 100% linear power supplies, SolidE™ EI-Core transformers, SmartDiscrete™ discrete circuit topologies, CleverChassis™ efficient industrial design, SuperSeries™ power regulation, CapBank™ power reserves…it’s amazing how crazy we can get with this. I must have spent a bit too much time in marketing, because BS like this just comes automatically.

    “As opposed to the BS you want to feed us now?” someone will snark.

    Well, yes.

    Again: technologies we consider important.

    All that stuff up there, that’s just what we do. It’s accurate, but overkill. Some companies may want to name and claim everything, but that makes everything meaningless. People simply don’t care that much.

    Aside: again, back in the agency days, we once competed for a big website redesign for a large Japanese technology company. This was in the dot-com boom, and the money to be had was huge. This company came to us with a laundry list of what they wanted, which included an insane, multi-tiered brand name structure that easily included a dozen sub-brands below their main brand. We threw out their brief and created comps that were structured around how people shopped for their products, and did not include the brand names at all—and we did this because we knew that the results would be focus grouped. The long and short of it was that our “heretical” comps won all the focus groups by such a huge margin that the company killed the project entirely. As in, they never awarded any agency the job. They kept it in-house, so they could maintain their rigid brand structure. Ah well, such is life.

    So yeah, as we start naming some of the technologies we consider important, I keep things like that in mind. We don’t want to go crazy about naming everything. Nor do we want to end up with a bunch of product line brand names that nobody understands or cares about.

    However, we’ve been coming up with some cool technologies lately, and it’s time to name a few of them.

    The Past: Dynamically Adaptive ComboBurrito

    In the past, we had an…ahem…ad-hoc way of naming things. In short, Mike or I would make something up, and then we would use it. Sometimes we’d slap a ™ on it, and sometimes not. Sometimes we’d forget about it, like Adapticlock, heck, I forgot I even made that one up. I never liked the name, which is probably why I forgot it.

    So, if you’re interested, here’s a historical list of names we’ve used. Please note these are not names we’re probably going to be doing anything with; there’ll be more discussion of what to do about digital tech naming at the end.

    • Adapticlock. Refers to the VCXO-VCO adaptive reclocking used in Gungnir and Yggdrasil. A microprocessor analyzes the incoming digital signal (for center frequency and/or jitter) and, depending on the quality, it is routed either to a VCXO (best) or VCO (good) for clock regeneration. Yeah. Kinda on the nose. Kinda, well, not a great name. Oh well. Everyone has an off day. And nobody thought of anything better.
    • Crossfet. Refers to the circlotron-style output stage used in Mjolnir, Mjolnir 2, and Ragnarok. I was going for something that encompassed both the cross-coupled nature of the stage and MOSFETs, but got something that sounds like the MOSFETs are angry. Which kinda makes sense, if you’ve ever worked with them. And yeah, I know, I know, again on the nose. Another pretty barfy name.
    • Dynamically Adaptive Output Stage. Refers to the Lyr and Lyr 2’s sliding bias system that allowed it to deliver more power than Asgard without melting down. It’s a weird stage, with a sense resistor and active amplification wrapped around what is essentially a single MOSFET. Kinda neat for the time, but Continuity is the present…and future. More on that later. Also a crappy name, but just because it’s wayyyyyy tooooooo loooooonngg.
    • MegaComboBurrito Filter. Refers to the time- and frequency-domain optimized digital filter that we use in Yggdrasil, Gungnir, Bifrost, and Modi Multibit. This one took a big chunk of math wizardry to work around a divide-by-zero problem in the original 1917 equations, and ate a bunch of person-years in development. Arguably the most important piece of digital technology we have, and probably the most ™able. Blame the current name on Mike. He started referring to it as the “megacomboburrito filter,” I believe, as a reference to its dual time- and frequency-domain optimized nature.
    • “Multibit.” This refers to the 16-to-20-bit true actual not-delta-sigma multibit converters we use in Yggdrasil, Gungnir, Bifrost, and Modi Multibit. The true and actual and not-delta-sigma need to be emphasized, because now it seems like every company that’s using a 2-bit to 5-bit delta-sigma converter wants to call it “multibit” these days. Again, blame this one on Mike, and on me for ignoring the confusion. Nothing we can trademark. By itself anyway.
    • Pivot Point. This refers to the balanced current-feedback topplogy used in Jotunheim. It’s a neat topology, one with a lot of benefits: great performance, simple implementation, stable and well-mannered. It’s not the “holy grail” Nexus stage, but Nexus is a whole ‘nother ball of wax, one that needs matched parts to perform its insane balancing act. Think of it as Nexus lite. Comment on the name: maybe technically accurate, but again, not great. Two words, clunky.

    The Present: Coherence, Continuity, Nexus, Unison USB

    Of late, you may have noticed a change in how we name our technologies. Names like Coherence™, Continuity™, Nexus™, and Unison USB™ are now getting out there. And this change isn’t by chance. It’s something I started purposefully last year, as I realized that some of the analog technologies I was developing were actually (a) gonna work, and (b) be pretty cool. That’s when I realized that we should claim some of these technologies…

    Aside: again, not ALL of them.

    …that we should claim them, because they were important, they were foundational, they would end up in a lot of different Schiit products, and they were differentiators—they made us different than, say, a IC maker’s reference design stuffed in a fancy box.

    Aside: nothing wrong with that, it’s just not us.

    That’s when I started actually thinking about names the way that I used to for clients of Centric. What’s a good name? What is short, meaningful, and memorable? What could we use for a long, long time, and apply to a bunch of products?

    And that, in short, is how we ended up with Coherence and Continuity on Lyr 3.

    Aside: these names were interesting enough that someone chose to use them in the official Lyr 3 thread here at Head-Fi. We didn’t start it.

    And that’s how we were able to deploy Continuity on Aegir, and get the attention of John Broskie, who branded it the future of amplifiers. (I believe he is actually wrong on that, it’s still not efficient enough to be used everywhere. It will never show up in Vidar, for example. Well, not without changing it so much that it really wouldn’t be Continuity anymore—it would be an insanely complex output stage with amplified current sense. Whooooo boy, that would be nuts. Do-able, sure, but at what cost. But I ramble.)

    And so, here we go: our current names of key technologies that we consider important.

    Coherence™. This is our combobulated tube-BJT hybrid topology that delivers on the promise of hybrid designs the most, at least in our experience. Unlike other hybrid topologies we’ve used, Coherence allows us to run a high voltage at the tube stage (better for tubes), use the tubes for voltage gain (better application of a tube’s strengths), combine the tube and transistor stages without input, output, or interstage coupling capacitors, and integrate it with a wide variety of output stages, including Continuity. This topology you’ll see in Lyr 3, and (to a lesser extent), in Vali 2, though Vali still uses output coupling capacitors to simplify the power supply requirements and keep cost down.
    • Importance: like I said, it’s the best hybrid topology we’ve found to date, and (as far as I know), we’re the only ones using it. Hence it’s worth naming as a differentiator to all the other hybrid topologies out there.
    • Found in: Lyr 3, Vali 2
    Continuity™. This is our constant-transconductance output stage that I’ve already written a couple of chapters about (Lyr 3 and Aegir chapters). It’s a big deal because it addresses a problem baked into literally every Class AB amplifier out there: the problem of transconductance droop as the amplifer falls out of its Class A biasing region. It also addresses the problem of N and P device mismatch, another gotcha that is baked into most Class A and Class AB amplifiers out there (except for a few using Sziklai output stages or something like that). It is a relatively simple topology that extends the benefits of Class A bias by adding additional output transistors as the stage exits Class A. Is it as good as weighs-a-billion-pounds-and-can-double-as-a-grille Class A? Maybe, maybe not. You tell us. Is it a panacea that fixes every Class AB stage? No, it still runs plenty hot and isn’t really applicable for low-bias or higher-power applications without heroic heatsinking, or a complete re-engineering of how it works (as mentioned above).
    • Importance: this is a unique take on the Class AB transconductance droop problem as enumerated by Bob Cordell and John Broskie, and is present in every Class A amp, and (as far as I know), we’re the only ones using it. Again, something worth claiming as a differentiator.
    • Where it’s found: Lyr 3, Aegir
    Nexus™. This one may be the most difficult to describe, because I haven’t yet written a chapter on it. Nexus refers to what I’ve called the “holy grail” circuit, one that allows us to run SE in to balanced out and balanced in to SE out in a single discrete gain stage, while preserving high-impedance signal inputs and offering an N/N gain relationship on the inverting and non-inverting inputs. That whole bunch of technoese almost describes the Pivot Point topology of Jotunheim, with the exception of the N/N gain relationship. So, it may not sound very exciting, especially when Pivot Point works so well as it is, and does not require the insane amount of matched parts that Nexus needs. But it’s very exciting to me, and it’s definitely the best all-purpose topology we’ve developed. Expect a chapter comparing this to Pivot Point and Crossfet, as well as, say, Supersymmetry, in depth in the near future.
    • Importance: as far as I know, this is the only discrete toplogy that succeeds in delivering SE or balanced in to SE or balanced out while offering high impedance signal inputs and maintaining an N/N gain relationship.
    • Where it’s found: Ragnarok 2
    Aside: Supersymmetry is a great name. Nelson Pass is a very smart guy. Supersymmetry’s patents were also licenced by TI. Coincidence? Maybe. But names are important. Hence this chapter

    Unison USB™. This refers to our new USB interface that Mike and team have been prototyping for the last year, and in beta on for the last three months or so. Sounds pretty boring in a world where you can talk to the air and have Alexa play your favorite house mix for you? Yeah, well, enjoy your compressed Bluetooth sound over a walkie-talkie speaker, and the paranoia of wondering if Jeff Bezos is listening to you take a dump. Instead, we’ll concentrate on solving the problem of getting great audio over a general-purpose interface using a general-purpose microprocessor and our own single-minded code. Which actually turned out to be quite a challenge. Which also ended up working so well that Mike now prefers it over SPDIF. Yes. Mike. No, no alien pods have been found decomposing near his house. Maybe the reason it works so well is that it’s designed to do one thing and one thing only: transfer PCM digital audio from a source and convert it to I2S. That’s it. No 500x DSD, no MQA, no Brutoof, no shamalamadingdong, no 32/768 or 64/1536, no unicorn formats period.
    • Why it’s important: it provides exceptional performance, frees us from commercial USB interfaces, and allows us to interface USB audio data with a wider variety of (ahem) nontraditional DACs. It is also our own technology, not XMOS, not CMedia, not Tenor, and we are actual USB-IF members with our own VID, which is something not many other high-end companies can claim.
    • Where it’s found: nowhere yet. It’s still in beta. Not sure where you’ll see it first. Bug Mike. He loves you.
    And that’s it. Four names. Four technologies. Two of which aren’t even in any products that have shipped yet, with one still in beta. I personally don’t think this is overkill, and I also personally think it’s important that we have some differentiators.

    But is that it? What about that ad-hoc “megacomboburrito filter” and the whole “everybody claims “multibit” problem.

    Good question. Let’s talk about that.

    The Future: What We (Probably) Should Name

    When I made the decision to start getting serious about technology names, I quickly realized that we had never named our most important digital assets: namely, our use of true, actual multibit D/A converters, and Mike’s unique time- and frequency-domain optimized digital filter.

    This, guys, is pretty dumb.

    Hey, I never claimed we were all that bright. Again, I’ve said that you may not ever buy anything from us ever again, once you get a close look at how much we mess up. This is just another example. I mean, Mike never thought it was that important, and I never thought it was that important, so maybe, in the end, it isn’t really that big of a deal.

    But maybe it is. “Multibit” is in danger of becoming a generic term thanks to overuse. The phrase “time- and frequency-domain optimized digital filter” doesn’t exactly roll off the tongue. So maybe we should name these technologies.

    So what do we call them? Let’s discuss that a little bit.

    Janus Filter, Impossible Filter, Synchronicity Filter, Unity Filter.I hope you didn’t throw up when you got to Synchronicity filter. Sorry about that. But what do we call a time- and frequency-domain optimized digital filter? Do we stick with MegaComboBurrito? Or do we try to play on its dual nature (Janus—yes, I know, wrong mythology)? Or do we play on the “impossible” nature of the math and getting around the divide-by-zero problem? Or do we go all Sting and Schiit and syncronicity ourselves away back to the 80’s? Or do we all feel good about optimizing both time and frequency that we have unity? Or something else entirely. Feel free to weigh in, we have operators standing by…er, wait, not really.
    • Why it’s important: this is a key digital technology used in literally every one of our multibit DACs.
    • Where you’ll find it: Modi Multibit, Bifrost Multibit, Gungnir, Yggdrasil, Universal Multibit card
    Deep Multibit, Integrated Multibit, Pure Multibit, Real Multibit, True Multibit, Multibit Neat, No-BS Multibit.Yeah, so what do we tack on to Multibit to differentiate it from everyone using a “multibit” delta-sigma D/A converter? Do we go deep (as in bit depth, 16-20 bits)? Do we play up the integrated aspect (as in, not a discrete ladder), do we talk about purity? Or do we just say what it is and let others defend: Real Multibit. What’s yours? True Multibit—kind of a nice ring to that one. Or do we go Multibit Neat, like scotch. Nothing added. Or play off the MegaComboBurrito and go with No-BS Multibit?
    • Why it’s important: there are really three ways to do D/A conversion these days: use an off-the-shelf delta-sigma chip, do a discrete ladder, or go with an industrial/medical integrated high-bit-depth multi-bit chip. No wait, there’s four: you can also convert to DSD internally. The point is, we pretty much brought the integrated high-bit-depth DAC back from the grave and extended it to price points (low price points) never imagined in the high end. We should probably call it something to differentiate it, right?
    • Where you’ll find it: Modi Multibit, Bifrost Multibit, Gungnir, Yggdrasil, Universal Multibit card
    And there you have it. Four names we’ve defined. Two we should probably define better. That’s the universe. We’re not going to start naming and claiming everything. Because, let’s face it, that’s really too much buzzword bingo going on.

    Thanks for reading my thoughts on this, as we try to figure out how to best explain what we do. Let me know what you think!
    Last edited: Mar 13, 2019
    Schiit Audio Stay updated on Schiit Audio at their sponsor page on Head-Fi.
    https://www.facebook.com/Schiit/ http://www.schiit.com/
    mikerp, kehorton, Currawong and 29 others like this.
  10. Ableza
    Very nice. I used to work in the IP area and know how important names, trademarks and copyright claims can be when others knock-off your inventions. On the marketing side, though, I've not done a lot, although I can claim the names for two products now being sold in parts of the semiconductor OEM industry: The LHT controller (stands for low-hanging testicle) and the BR-549 sensor (an homage to Junior Samples from the TV show Hee Haw.) I tried to name a custom implementation of a DeviceNet communications protocol the HISBA interface (HISBA = help, I'm surrounded by arseholes) but that was rejected by the Sales team. :)
  11. Ableza
    By the way I like Janus Filter and True Multibit.

    And now I've got the song Synchronicity stuck in my head, damn you. :wink:
    Last edited: Mar 13, 2019
  12. Jason Stoddard
    Oh that's beautiful.

    Similar to why we append an -FUA to our Amazon part numbers.
    Schiit Audio Stay updated on Schiit Audio at their sponsor page on Head-Fi.
    https://www.facebook.com/Schiit/ http://www.schiit.com/
    Currawong, Khalaris, motberg and 15 others like this.
  13. Brian D
    Yeah, I think "True Multibit" is a winner. I'd prefer "Burrito Filter" but Janus is ok too.

    That got a laugh out of me, thanks!
    Last edited: Mar 13, 2019
    ScubaMan2017, Mike-WI and Ableza like this.
  14. ScubaMan2017
    This was a long, overdue chapter. I can stop cringing whenever I use mega combo burrito to describe my Modi Multibit. And re-appreciate what the hell’s so special about my Vali (my favourite toaster). Hopefully, some of this chapter will get into the snarky FAQ on the website (or add a link to it).
    Ableza likes this.
  15. Mr Rick
    I'm old enough to remember the Infinity POS 1 speakers. You can imagine what the guys in the shop called that one. :L3000:

Share This Page