jason stoddard said:

   
Digital Audio Player. Think Pono, Fiio X5, Astell & Kern, etc.

Click to expand...

Thank you.  Just an acronym I didn't know. 
 
Pono is coming and I'm looking forward to it.  It'll either be a great sounding if content-limited device, or an expensive conversation piece, or both...

jason stoddard said:

  I won't say "never," but highly, highly, highly unlikely.

Click to expand...

Yea. Blah-blah. So what you're saying is there's a chance.   

jason stoddard said:

Two things I doubt you'll ever see from us, and why:

1. Stat amp. Limited market, safety concerns (mainly internal), politics.

2. DAP. High integration, software-UI intensive, potential Apple entry into high-res, may be made irrelevant by ongoing changes in Apple or Android phones.

I won't say "never," but highly, highly, highly unlikely.

Click to expand...

I always love seeing the "why" someone does what they do

It's also a big contrast from working with a company where the answer to "why" is usually "that's the way we've always done it"...

With more and more orthos creeping up the stats' alley of resolution, while bringing superior bass response to all electrostats, I think perhaps ES headphones are slowly entering a stage of obsolescence, especially when considering their price and the prices of (high quality) ES amps.

tigon_ridge said:

  With more and more orthos creeping up the stats' alley of resolution, while bringing superior bass response to all electrostats, I think perhaps ES headphones are slowly entering a stage of obsolescence, especially when considering their price and the prices of (high quality) ES amps.

Click to expand...

I agree but the STAX 009 still sound better than the Abyss and are more refined in terms of build and design, IMHO. Even if some think the Abyss sounds better it costs the same as the 009 so not really creeping up in terms of price. Sure the LCD's are great sounding headphones but nowhere near the 009, and most would agree with this (including Jude who reviewed the 009's as the best headphones ever made). Sure the orthos have better bass thaan most Electrostats, but the 009 is in a league of its own and has great bass. In fact it's better than all the other Stax I have heard. I have yet to hear another pair of cans (besides the 009's) where I can almost feel the singer breathing down my neck, it's just spine tingling. Whenever I listen to the 009's I forget that I have cans on, sometimes I will hear a sound I never heard and actually turn my head, only to realise with a delayed reaction that it was something I had never heard in the music and not something in the room. Again, the Abyss comes close, but at the same price, the 009's win hands down in comfort, design and sound and there's more than enough bass quality and quantity.

jason stoddard said:

2. DAP. High integration, software-UI intensive, potential Apple entry into high-res, may be made irrelevant by ongoing changes in Apple or Android phones.

Click to expand...

Not to mention the increase in service required for a DAP. It would at least double your emails and calls for help.

jason stoddard said:

  Two things I doubt you'll ever see from us, and why:
 
1. Stat amp. Limited market, safety concerns (mainly internal), politics.
 
2. DAP. High integration, software-UI intensive, potential Apple entry into high-res, may be made irrelevant by ongoing changes in Apple or Android phones.
 
I won't say "never," but highly, highly, highly unlikely.

Click to expand...

 
you know what Schiit fans, I love this statement.
 
it translates to:  "we know what we do well, we will continue doing such.  we won't step out on a limb 'just because' fanboys are barking for a new toy."
Bravo, keep it up Jason.

jason stoddard said:

  Two things I doubt you'll ever see from us, and why:
 
1. Stat amp. Limited market, safety concerns (mainly internal), politics.

Click to expand...

 
Thanks Jason. I'm curious about the politics part. What's that about? 

Criticisms from the peanut gallery. See Cavalli Liquid Lightning and Ray Samuels A-10.

andreyew said:

jason stoddard said:

 
Two things I doubt you'll ever see from us, and why:

1. Stat amp. Limited market, safety concerns (mainly internal), politics.

Click to expand...

Thanks Jason. I'm curious about the politics part. What's that about? 

Click to expand...

Most likely that the impressions of a select few will have very big influence on whether the product will sell or not

Yea, but considering regardless of those 'select few' negative impressions.. products still move. I don't think that's the issue. The safety concerns are legit, plus, yea, the market is limited comparatively. 
 
Funny enough, those 'few' who don't think very highly of the LL or the A-10 tend to like Schiit's way of doing things. Lastly, these politics are just as rampant in the dynamic world.

muziq said:

  BTW that's been around a little while...

Click to expand...

 
Please note the lack of copyright in my post
 
Timing is everything. Even for well-used humor.

jexby said:

  you know what Schiit fans, I love this statement.
 
it translates to:  "we know what we do well, we will continue doing such.  we won't step out on a limb 'just because' fanboys are barking for a new toy."
Bravo, keep it up Jason.

Click to expand...

 
+1. Absolutely agree. Just because you could make a widget, doesn't mean you should. There's the whole principle of return on investment to consider.
 
I don't personally know Jason or Mike, but I would lay money on their ability to make pretty much any amp or DAC they set their minds to. I don't need any more proof than my Magni and Modi. But you need to choose your products carefully. That point has been made repeatedly in this thread.
 
I could spend six months inventing a software program that a handful of people like. They might even pay a few dollars for it. But if only a handful of people will ever want it, why should I spend my time? This has been my personal experience. If you can't make a profit, there is no point in spending the time or effort on creating a product. Even if you know you can. Even if it will make a small number of people really happy. You can't turn Happiness for a few people into groceries for your family.

Chapter 13:
“Isn’t the Symbol for USB the Long Flat Rectangle?”
 
Strange title for a chapter, right? It’ll become much more clear—and much funnier—later on.
 
This chapter is really about three things:
 

• Transitioning from a “headphone amplifier company” to a “DAC/amp company,” and, eventually, into an “Audio Products” company.
• The difference between a hardware company and a software company, and some of the decisions you have to make if you’re going to be both. Plus a little primer on firmware/drivers/software.
• Products that really, really, really, really don’t want to ship on time. Or so it seems. This ties into the “development time is proportional to the square of complexity.” Or incompetence. Or both.

 
I’ll skip the now-usual summary, because we still have the same three amps, and we’re the same basic company: a fast-growing headphone amp manufacturer which produces (relatively) low-cost, high-value products in the USA. 
 
I’d announced the Bifrost DAC on June 30, 2011, with a delivery date projected to be in August. Because we’d been running the prototypes for a while, I was really confident we could ship the Bifrost early, and was looking forward to a July launch. I wanted to make sure we shipped everything early from now on, like Lyr.
 
Now, at this point in time, we were still taking pre-orders. So, starting June 30, 2011, prospective customers were able to order a Bifrost. This immediately caused several problems, as exemplified by these questions:
 

1. Hey, I just bought a Bifrost and an Asgard, do you hold them and ship them together, or can I get the Asgard early? Problem: this broke our shipping quote system, which was still in the archaic realm of “guess a range, and hope you don’t lose too much money on it, or overcharge so much that nobody buys.

 

1. Hey, if I get that Asgard earlier than my Bifrost, when does the 15-day money-back start? Problem: Oops. Kinda broke that model, didn’t we? We didn’t have a great answer for this, and ended up making all sorts of extended accommodation deals. Which gets fun if you’re trying to refund money, say, after 60 days.

 

1. Hey, do you have an exact date when you’re going to ship, I need to plan…? Problem: No, of course not. We didn’t even have metal. So what did we do? We said we’d probably  ship early, in July. Very dumb. Do not do this.

 

1. Hey, do you charge my card before you ship? Problem: No, we never charge before we ship, but the problem is—the authorization on the card expires. Go back and try to capture transactions that were authorized a couple of months ago. Guess what? Sometimes they don’t re-authorize (and, if you’re using PayPal, you can’t re-authorize them after 29 days at all, game over, contact the customer, send them an invoice.) Customers don’t like these complexities.

 

1. Does Bifrost work with the GigaComplex Server X-2000 running a modified Amiga operating system with Rubycon Black Gate capacitors and opto-coupled USB 2.2 outputs? Problem: Yeah, I’m having a little fun with this, but you get the drift—lots of people asking questions about Linux, phones, etc. We’d only tested on Macs and PCs using the three most popular PC OSes at the time (XP, Vista, 7). And that was a hell of a lot of testing. So we really couldn’t say. It wasn’t until later, when Tony arrived, that we had a chance to start looking at Linux and phones and Chromebooks and stuff.

 
Which is a great segue into the difference between hardware companies, software companies, and what happens when you have to do both. So let’s talk about that for a bit.
 
 
Hardware, Software, and Restaurants
 
Restaurants? Yes, restaurants. As in, places you go where they make food for you. Although I’m a foodie, I have exactly zero desire to ever open a restaurant (even though I’ve had some neat ideas. Why? Because restaurants combine the problems of manufacturing with the problems of service with an extra problem of the stock actually goes bad. No thanks. So, a tip of the hat to anyone who can make it as a restaurateur. That’s a helluva business.
 
What does that have to do with hardware and software? Well, the problems with hardware are largely a matter of manufacturing (buying/holding stock, assembly, margins), while the problems of software are largely a matter of service (finding and keeping a team happy building great code, aftersales questions, dealing with a constantly changing OS and software regime.) Of course, I’m oversimplifying, but you get the general picture.
 
Hardware Business Problems. When you’re making hardware, like analog headphone amplifiers, your problem really comes down to this: can you make enough margin on what you’re making to run the company and keep it growing.
 
Because, when you’re talking hardware, you’re talking stock. You have to buy chassis, transformers, boards, transistors, ICs, and a myriad of other parts before you can ship anything. That means there’s a lot of money up front in hard bits. Then you have to assemble it (with either in-house or contract assemblers), test every product, ship it, and support it. If you’re doing it right, your repair support will be minimal—your failure rate should be much less than 0.5% (ideally 10x lower than this.) And, if the products are simple, your tech support should also be minimal.
 
So, for parts, you could be looking at 33-50% of the cost of your product. So if you’re making a run of 100 $1K amps, you’ll be out $33-50K at the start. Then you have the added costs of personnel, facilities, support, and shipping on top of that.
 
Note: never discount the importance of shipping—this is not a trivial task, and it can get very complex. Use “free shipping”—actually, “cost-inclusive shipping,” to be more accurate—at your own risk.
 
So, in hardware, before you’ve shipped anything, you’re out a sizable investment. Moreso when you find out the parts you spec’d aren’t available, or you ordered the wrong parts, or something’s wrong with them, etc. But once it’s shipped, you should be able to make some money, if you’re doing it right. And you can keep doing it, as long as there is market demand—but of course every product you make has a hard cost.
 
Software Business Problems. Software’s a bit different. Your “hard costs” are actually salaries for the programming staff, your facilities costs, your admin costs, etc. These aren’t tied to the amount of software you sell. If you spend $50K developing software, it doesn’t matter if you sell 100 copies or 1,000,000 copies at $99 each, because you’re going to be distributing it via download or via inexpensive media.
 
But—and here’s the big but—you have to support it. Unlike simple hardware, you have to assume that it won’t install on some systems, it will conflict with other software, many customers may need some hand-holding in using it. And, as an added bonus, the OSes and other software changes from year to year. This means that what worked yesterday may not work today, and vice-versa.
 
What this all adds up to is that software is an ongoing business—you need to keep programmers constantly employed and engaged, you need to test your product against changes in OSes and software, you need to issue updates to deal with incompatibilities, and you need to have a significant service staff to provide technical support. So, with software, even if your costs are not directly related to how much you sell, you have to sell enough to cover your costs and make a profit.
 
Note: of course, there are different models (open-source, SaaS, etc) that I’m not covering here, and admittedly I’ve never had a software company, but the principles are somewhat similar on the agency side (heavy service, overhead unrelated to amount of revenue.)
 
So, back to resturants. They have hard costs (ingredients), service costs (chefs, servers, etc)…and as added bonuses, their stock goes bad over time and they have additional regulation (liquor licenses, inspections) and accidents (sick people) to deal with. So, no restaurants in our future. But I’m certainly hoping others are up to the task.
 
 
What the Heck Does This Have to Do With Anything?
 
Yeah, I figured you’d ask that. What it has to do with Schiit is that, with the launch of Bifrost, we became a software company by default. A very, very lightweight software company, yes, but we had new complexities of firmware and drivers to deal with.
 
“Oh, big deal, you get the Giant Baby of the Year Award,” say the hardcore software developers now.
 
Yeah, and that may be true. But by using firmware to run the Bifrost, and planning to release drivers for Windows, we were crossing another invisible line in business: from a pure hardware company, to a hardware/software company.
 
And yes, a driver isn’t the biggest deal in the world. But with this tiny piece of software, two things should be noted:
 

• 80% of our tech support is for Windows driver issues.
• The have been updated no less than 6 times since we (finally) launched Bifrost in October 2011. The biggest change, of course, came when Windows 8 was released.

 
So, perhaps not a big deal in software-land, but a big deal for a hardware company. Luckily, we knew what we were getting into, and had long discussions about leaving Bifrost 24/96 only, which would eliminate the need for Windows drivers. The discussions went something like this, usually in a summer-hot garage:
 
“Mike, we need to offer 24/192 support. There’s 24/192 music available on HDTracks,” I told him.
 
“What, seven tracks of it?” Mike sneered.
 
“It’s limited, yes, but people are asking for 24/192 support. And it would be a good differentiator for Bifrost, now that we’ve nixed the balanced outputs.”
 
“Balanced should only be hardware-balanced,” Mike pontificated. “Two DACs. Summed single-ended. That’s what we did at Theta.”
 
“I know, but don’t change the subject. 24/192.”
 
Mike doesn’t change course fast, though. “If we’re going to do balanced, we’re going to do it right. Hell, Bifrost would sound better if we used two DACs per channel.”
 
“Yeah, and it would cost $700.”
 
“But it would sound good!” Mike insisted.
 
“Yeah, and you can do that one later. For now, Bifrost. 24/192.”
 
“You don’t have any balanced inputs on our amps anyway,” Mike said.
 
I frowned. He had a point there. Balanced inputs—done right—required a 4-gang potentiometer, which we didn’t have back then. We had grand plans, sure, but no balanced inputs yet. “Mike. 24/192. Must have it.”
 
“24/192 uses a half-rate master clock to the DAC,” Mike said. “24/96 probably sounds better.”
 
“Is this a limitation of the AKM?” I asked.
 
“No, it’s a limitation of most delta-sigma DACs. The master clock can only be so fast.”
 
“I still want 24/192. Period.”
 
Mike sighed. “Who’s going to do the tech support?”
 
“Me, for now.”
 
“You’re going to want to shoot yourself,” Mike predicted.
 
“Maybe. But we need 24/192. We’ve tested 24/192. It works. Let’s support it.”
 
“Ohhh…kay,” Mike said.
 
And that’s how Bifrost got 24/192 support. It seems funny today, with 32/384 or even higher sampling rates. Not that there’s any PCM music available there, but hey, it’s like megapixels. Meaningless numbers to use in marketing. Buzzword compliance.
 
Note to self: we should do a 32/384 DAC that has a switch for “easy mode,” supporting 24/96 without drivers, and “expert mode,” where you’ll need Windows drivers. Except unlike everyone else, we’ll tell everyone why 32/384 is meaningless. See a couple of chapters back.
 
 
Hardware, Firmware, Software
 
With Bifrost, though, we didn’t just have software. We had hardware (the DAC—chassis, motherboard, USB daughterboard, DAC/analog daughterboard), firmware (for the motherboard and USB daughterboard), and software (Windows drivers.) This was a level of complexity higher than anything we’d ever done before. Still not insanely complex, but complexity was one of the reasons we were late.
 
Yes, I’m getting back to the late part. But first, some of you are probably saying, “Software? Firmware? What the hell is the difference, anyway, and what do they do?”
 
Firmware is embedded code you don’t expect to change very much, if at all. Examples of this include the code that runs the LCD display on your refrigerator, or makes the buttons on your sprinkler controller work. Firmware runs behind the scenes, usually never needing updating.
 
Software is installed code that can change quite a bit. This is everything from your nVidia driver to your copy of Microsoft Office. You update them from time to time, or even switch to different versions.
 
In the case of Bifrost, we had to develop firmware to run the motherboard, and modify firmware to run the USB input.
 
The motherboard firmware runs the front panel button, the bitperfect clock management system, and the hard mute protection system. Not very exciting—but necessary if you are going to have clock management like Bifrost. It’s programmed via an Ethernet jack inside the Bifrost (intentionally inconveniently placed, because we don’t expect to ever update it…unless there are major changes to the DAC/analog card or something like that.)
 
The USB firmware runs the CM6631A USB receiver, generating the specific clocks and formats that we need for our D/A implementation. It’s programmed via the USB port.
 
“Aha!” some of you are saying. “I can update the USB firmware on the Bifrost via the USB port? What kinds of cool firmware do you have?”
 
Not so fast. Yes, you could, but there’s really only been two versions of firmware—one for the original USB input, one for the Gen 2 USB input. There are a ton of other complexities as well. It’s easy to brick a USB input by playing around with the firmware. So, just don’t.
 
Note: We’ll know if you change the firmware, because the programmer reports the current version of firmware on the card.
 
And firmware is where we get into the genesis of the title of this chapter. But let’s talk about problems some more, first…
 
 
Products That Don’t Want To Ship, Like Bifrost
 
Okay, back to Bifrost problems. Announced in June, stated to be shipping in August. When did it ship? Late October. Doesn’t sound like much, but when you’re already ordered into backorder before shipping, and you have hundreds of angry customers yelling at you—some of which waited almost 4 months to get theirs—it’s a very big deal indeed.
 
Aside: Haven’t I said something before about opening your mouth and pre-ordering? Well, it took us past Mjolnir and Gungnir to finally learn that lesson. Mainly because, “Nothing could be as bad as Bifrost.” Yeah, right.
 
So, what went wrong? In this case, mainly mechanical problems, but there were electronic ones as well.
 
First, it was the DAC/Analog board. My first attempt at a hardware summer (from the DAC’s differential output) was simple, elegant, good-sounding—and completely useless, because it would have fixed Bifrost at 1V out, rather than the industry-standard 2V RMS. So I had to go back to the drawing board, for a completely different topology.
 
Then, the way we were going to attach the daughterboards didn’t work out. We’d expected to use press-in connectors, like I did back in the Sumo days. Those didn’t work so well, because the Bifrost daughterboards stood a lot taller than the Sumo ones (they had to clear parts underneath.) We played around with various plastic options before saying, “the hell with it,” and using metal standoffs and screws instead, to ensure the boards would stay in place during shipping.
 
Next, the board interconnections themselves. Now, there are plenty of header options out there, but not at the length we wanted to use. Which meant Bifrost headers were custom. Which meant an 8 week lead time. Which we hadn’t counted on.
 
Note to other guys who want to start hardware companies: check the lead times. Then add a few weeks. Some parts are 14-16 weeks. And sometimes they don’t come on time.
 
Then the metal came. Unlike Valhalla, it was beautiful…but it didn’t fit. I’d gotten used to getting the metal “done in one,” and we hadn’t done a first article for fit. So, now we were sitting on hundreds of chassis that didn’t fit…and we had boards already out for assembly.
 
This was by far the biggest problem. Because when you’re in a situation like this, you either have to redesign the board (and throw away the assembled ones) or redesign the chassis (and scrap the chassis).
 
In the end, we did a modification to the inner chassis that allowed it to work, then made a longer-term change—the only running change we’ve ever done to metal.
 
Another note to other guys: don’t change the metal if you can help it. Having two (or more) versions of the same chassis plays hell with production—as in, “Hey, we found a box of chassis, but the boards don’t fit.”
 
In-between the non-fitting metal and the stuff that worked, we wasted a ton of time trying to figure out how to make it work. Which is dumb. Metal, especially cosmetic parts, is fixed. You can’t just “oval out” a hole in a cosmetic part, or cover up things with plates, or any other of a dozen stupid ideas we came up with. Throw it away and start over—or fix the board to fit the metal. Period.
 
And, when the modified metal came, we found that we’d inverted the left and right designators on the output jacks. Argh! Back for rescreening.
 
In short, Bifrost really didn’t want to ship. It was a product from hell. And it’s what we get for trying to take shortcuts, like no first articles, combined with the most complex product we’d ever made.
 
But, by RMAF 2011, we had a working Bifrost, in a cosmetically perfect chassis. So, what did we do? We took it to the show, of course!
 
This time, we weren’t exhibiting with Sennheiser. We’d grown up enough to have our own booth! But we hadn’t grown up enough to do anything other than take the products there. We had no signs, no literature, nothing. Lisa clipped a couple of 8.5 x 11” printouts of the logo to the curtains behind us, but other than that, there was no indication of who we were. Great marketing.
 
That was the last show we did like that. I knew we looked bad, and I vowed to change it for the next time.
 
It was also the last show that we took all the products in a single backpack. Literally. Asgard, Valhalla, Lyr, Bifrost, a pair of LCD-2s and a couple other cans—all in Rina’s backpack.
 
That was her idea. It would save on shipping, and there was no chance they’d get lost, she said.
 
“So what happens when Homeland Security takes one look at the X-Ray machine and stops you for having 40 pounds of electronics in your bag?”
 
“Then I smile nicely at them and play dumb.”
 
I sighed. “Then I’ll tell them I don’t know you.”
 
“It’ll be fine,” she insisted.
 
And—frighteningly enough—she was right. They didn’t even blink at LAX. Didn’t stop us. Didn’t ask about what all the electronics were. Not a single question.
 
Nor did they stop her, or ask anything, at Denver International, when we were coming back. Makes you feel really good about flying. 3.5 ounces of shampoo? Bad boy, go to the little room. 40 pounds of aluminum, steel, copper, transformers, and wiring? No problem, move along.
 
 
Isn’t the Symbol for USB…
 
Now, since we had a cosmetically perfect Bifrost, it meant we could ship other cosmetically perfect Bifrosts, too. That day we put together the first ten Bifrosts, I felt about 1000 lbs lighter—and about 10 years younger.
 
At least until the next day, when we had probably 20 more.
 
Rina was listening to the Bifrosts, prior to cleaning, bagging and packing. She knew what an ordeal it had been getting the product to market. She knew all of our little (and big) frustrations. She knew what a sore point it was.
 
So, it was perhaps with a bit of fear and trepidation that she said, “You’re always going on about how Mike did the first DAC, right?”
 
“Yeah, yeah,” I said. “So?”
 
“But this is his first DAC in a while, right?” Rina continued.
 
“Yeah,” I said, wondering what she was going on about. “So what?”
 
Lisa sighed. “Isn’t the symbol for USB the long flat rectangle?” She pointed at the icon on the front of the Bifrost to clarify.
 
“Of course it is!” I said, thinking, That’s a dumb question.
 
“It’s not the round circle?”
 
“No, that’s the coaxial input. Why?”
 
“Because this Bifrost is playing USB when it’s on the round circle.”
 
What? I stomped over to take a look. She was right. It was happily playing music from the USB input, with the front panel LED indicating that it was on the coaxial input.
 
My heart skipped a beat. “Are they all like this?” I asked.
 
“I think so,” she said, and grabbed another.
 
In a few minutes, we confirmed: yes, they were all like that. Including the show Bifrost. Which we’d never noticed.
 
Great, just great.
 
Now, I knew what was happening. The front panel LEDs were controlled by the microprocessor. It was simply lighting up the wrong LED. It was nothing that changing the firmware couldn’t fix.
 
But…changing the firmware took Dave. I had to alert him and get a new copy of the firmware. Which only took hours. But then we had to take apart every Bifrost so we could get to the Ethernet connector and re-program them. Again, not the end of the world. But Bifrost had to get that one last shot in, before it would happily ship.
 
And…what about all the other Bifrosts we shipped yesterday?
 
Yep. You got it. They all had the wrong firmware. We offered to replace it, but many customers didn’t want it changed…as if it was a Bifrost Special Edition or something. By my estimate, there’s 6-7 Bifrosts out there still with reversed LEDs. As well as my personal Bifrost, the original show unit. I’ve simply never bothered to change it.
 
So, if you buy a used Bifrost, and see that the LEDs are wrong, let us know…we’ll change it. Or not. You have a piece of history.

jason stoddard said:

 
And—frighteningly enough—she was right. They didn’t even blink at LAX. Didn’t stop us. Didn’t ask about what all the electronics were. Not a single question.
 
Nor did they stop her, or ask anything, at Denver International, when we were coming back. Makes you feel really good about flying. 3.5 ounces of shampoo? Bad boy, go to the little room. 40 pounds of aluminum, steel, copper, transformers, and wiring? No problem, move along.
 

Click to expand...

 
I travel for work 70%, and spend a lot of time in airport security lines. How the hell does that go by?  And yeah, I had a 3.5 ounce bottle of $80 cologne way in the beginning.  Bastards took it and trashed it. I was not a happy flyer that day. But good word, how does that not draw attention?