We first interviewed Eric Valentine in Tape Op #45 and discussed his production career and techniques. Eric is well known as a producer and engineer for working with Third Eye Blind, Smash Mouth, Good Charlotte, Queens of the Stone Age, and more recently Grace Potter. But in 2007 he took the unusual step of starting a pro audio gear company called UnderToneAudio. With the fairly glutted pro audio marketplace, why would a successful producer go down the black hole of manufacturing audio gear? We sat down with Eric to find out.
How did you go from producing records to starting UnderTone Audio?
It really started when I was at a crossroads, just dealing with having a large-format recording console that would do what I needed it to do. I've always been a Neve guy. I was lucky enough early on to have bought an 8038 Neve that had 32 1081s in it from Ocean Way back in 1990. Just because everybody said, "Neves are great. That's the warm sound." So I got one, but there was a learning curve with that thing. It did not sound amazing right out of the gate. But once I learned it, it did sound amazing. All of those early records, the Third Eye Blind record and Smash Mouth's first couple records, were all tracked on that console. Then there's a certain point where I hit the limitations of mixing on that console. I just couldn't mix on it anymore; I'd outgrown it. I upgraded to a Neve 8128. I know some people would probably consider that a downgrade, but I loved that console. The 8128 was incredible. That console actually moved to L.A. with me and was set up at Barefoot Recording. I did Queens of the Stone Age and the first Good Charlotte record with that console. But then that console was really at the end of its lifespan. There was just no amount of contact cleaner that was going to make me trust that thing ever again. It was over. So, I decided, "What the hell. I'm going to get the latest and greatest Neve console." So I got the Neve 88R. It's an amazing console but they had travelled too far away from what I need a console to do. There's so much stuff on there for film stages and all these other things that I didn't need, and the thing's gigantic. The channel strips are three feet long. The nearfield monitors weren't nearfield monitors anymore. There's this giant sea of flat metal surface between me and the [Yamaha] NS-10s. The speakers sounded weird, and I could never really get comfortable with it. That console just didn't work for me at all. It drove me nuts. During the same time, I met Larry Jasper. He was repairing and modifying all of my vintage gear. He's just so good; he's a total savant with this stuff. He memorized all of the schematics for all of the equipment in my studio. He already knew them all. Every single item that I mentioned, like my [Teletronix] LA-2A seems to be weird, and he'd say, "Yeah, they got it right with this, and it was really good they did this thing, but it'd be a lot better if you bypass the output transformers." We'd try the stuff that he suggested, and he was always right. The changes that he'd make were amazing. So we modified my LA-2As and did a bunch of work on my [Urie] 1176s, tape machines, and everything. Everything came back incredible, super reliable, and sounding amazing. At this point I got kind of cocky and thought maybe we could just put together a console ourselves! Originally, I was going to build it around some module that I liked; the original 1084 Neve module, sort of a smaller version of the 1081 that was still Class A. I was going to find a bunch of those and then build a console for the bussing and routing. Then I got more ambitious and thought I'd really like the EQ to be more flexible. Maybe we can modify them. That looked like it was going to be too complicated, so finally I just said, "Fuck it. Let's go from scratch and see what happens." With every single thing, Larry would say, "This aspect of the design was good, but it could have been better if they did this, that, and the other." I said, "Let's do everything the way that you think would be the most amazing, and then let's just see what happens!" I was lucky enough to be in a place and have the resources to do an experiment like that. I paid him and we both worked on it together to build two consoles; a 60-channel console and a 48-channel console. One for Studio A and one for Studio B.
What year was that?
That started in 2007. It took us two years to finish the first one in 2009 and got installed in Studio A. We genuinely built the best-sounding analog console I've ever used in my life. On top of that, it was a new psychological context to work in, where I'd just spent two years designing what is conceptually the greatest analog console ever built. "This console is going to make everything I do sound better from now on." I'd start every project with that mindset. "This is going to be better than everything I've done before." It's a very inspiring context to work in. That's what I find the most useful about new equipment. There are always points along the way where I feel like I'm using stuff that's worked in the past, and it's probably fine and I can probably use it for the rest of my life, but I want to feel that sensation of I'm going to change things so that things will unavoidably be better and different in the future. It just adds this new, fresh inspiration to everything that you do. There's an expectation that things are going to be better than they were. That was the most powerful version I've ever had of that feeling in my life, because I was so emotionally invested in having built that console. That aspect of it was really inspiring and really wonderful. When we were done with the whole thing, it was very apparent that the equalizer we had designed was very unique. Extremely powerful, very musical. There was simply nothing like it that had ever been done before. Other people started expressing interest in it, so we decided to build that as a product. That was really the beginning of UnderToneAudio.
That was the EQ standalone, or the EQ with the mic pre?
The EQ with the mic pre. It was a channel strip, the MPEQ-1. We built 100 of those ourselves with a small group of people who were all part of UnderTone. We didn't use a contract manufacturer or anything. We set up our own manufacturing installation upstairs at Barefoot. I had three or four people up there who were stuffing boards and soldering. It was all through-hole components, and we did it all ourselves. The small boards have surface-mount components, so those we did contract out. The submodules with all the active circuitry. But all the larger boards were hand soldered. We built the first 100 MPEQ modules. Then people wanted consoles, so we started building consoles for other people. Greg Wells [Tape Op #123] bought the first UnderTone console. Then we built one for Fraser T. Smith and a guy in Alabama named Jason Elgin, and then there's one in Switzerland. We built one for John Kuker who had a studio in North Hollywood called Seedy Underbelly. Really, really cool guy. He ordered a big console. He passed away literally in the last week of installing his console.
Yeah. It was really crazy. I was in Switzerland finishing this recording school installation of the console we built for them and got the email that John had passed away. That console got tied up in a bunch of probate stuff for a long time and ultimately found a home in Brazil. Anyway, there's a total of eight UnderToneAudio custom consoles on the planet.
Are you still making them for people?
We don't make consoles now. Really the main issue is that the consoles require a lot of my time. There's a lot of mechanical design work that I did on those consoles and configuring a custom version of it is something that I have to do, and I just don't have the time to do it. We're just not offering those right now. We do have requests from time to time from people who are interested. But building a custom Class A console in California is insanely expensive. A lot of times people think, "Oh, if we just build a really small one it'll be super affordable." No, that's actually the most expensive way to do it, for what you get. When they compare it to mass-manufactured API consoles, they're like, "Hey, I can get the API for this." "Well, buy it then." We cannot hand-build something like that in the United States right now.
Have you thought about just doing a stock size console? Like here's the 16-channel console, and there are no options.
I have considered that. I just think it's been more important to focus on getting all the rackmount products up and running and have them going smoothly before I got back into that. That's still a lot of design work for me. There's a learning curve to becoming an audio manufacturing company. After building the first 100 MPEQ-1s, my guys were like, "This sucks. It's super repetitious, super boring. It's mind-numbing. We're not into this. We'll build custom consoles, but we're not cranking out repetitious stuff. It's too boring." I said, "Okay, let's get a contract manufacturer to do that." We signed up with a contract manufacturer down in Orange County. They started off great and then sort of fell apart. There was a certain point where we had to make a change. They made some really fatal mistakes that were difficult to undo and fix. I couldn't have stuff go out that was unreliable and not put together properly. All of it had to get fixed. We switched to another contract manufacturer almost two years ago. There's a whole process of getting up to speed with a new manufacturer, getting all the bills of materials, ordering parts, understanding how to build them and test them. It's this whole crazy, complicated process where you have to teach somebody how to be an expert on your products. We went through that whole thing with them, and they're great. That company's JMP Electronics. They're amazing. I think things are finally smoothing out. We just placed another purchase order for another hundred UnFairchilds, which is kind of sobering to initiate. Those things are expensive to build, so it's a huge commitment of hundreds of thousands of dollars when you say, "Okay, we're going to build 100 of these." You're sorting through all these expensive parts, transformers and tubes. I know that we're going to need hundreds of thousands of dollars to pay for all of these parts and pay for people to build these things. We don't have hundreds of thousands of dollars right now, so we've gotta sell stuff! It's been an interesting transition. I was subsidizing UnderTone with my producer income for a long time. Then in 2018 I took the training wheels off. So starting in 2018, UnderTone's gotta pay for itself. It's tight, but it's working.
You had mentioned you went through a whole process of blind listening tests when you were initially developing the gear?
Yes, there was a part of this process of designing the circuitry for the first consoles that was a real revelation for me. We were going through this process where Larry was designing his version of the little amplifier modules that are in all the classic circuitry. Like in the Neve stuff, there were smaller ones, like the BA183. It's a little amplifier module that has this 2N3055 transistor in it. That same little card plugs into the mic preamp. It'll plug into 1272 modules, which are used for summing and stuff like that. It's this all-purpose amplifier. Larry was designing his version of that. He came up with this unbelievably brilliant design that has all of the sonic characteristics of a Class A circuit, but it's still push-pull, like a Class AB circuit. There's a way to do that where you still get all even-order harmonics, like a single-ended Class A circuit. In that process, we were listening to all these different versions of the amplifiers. He would mock up a version of one of these amplifiers, and then we'd have to listen to it. It's a differential amp that creates an active balanced input for a circuit; a line amp. He would try different ways of doing feedback, different capacitors and stuff, and I was listening to everything to try and see which versions we liked better. At this level, when you're just talking about a line amp and swapping out a 15 picofarad capacitor for a 20 picofarad capacitor in the feedback loop, and trying different dielectrics in the caps and all this stuff, the differences are insanely minute. Some of the differences are beyond the scope of human hearing. You might be able to test it and find a difference, but the differences could be way up in the 80 kHz range. When we started listening to this stuff, I'd have both of them set up. I'd sit down, put some audio through both, and he would always say, "Okay, these polypropylene caps are going to be smoother on the high end and blah blah blah." I'd listen to it, and sure as shit every time, I was like, "Oh yeah, totally. These new caps sound so much better on the high end." At a certain point I started to get suspicious, and I thought, "Okay, I want to prove there's a difference here." I started doing blind listening tests. This crazy thing happened where I would listen to the two things un-blind and clearly hear a difference. "Oh yeah, this is the one with the new cap. It's way smoother-sounding than the previous version. Let me do a blind listening test. I'm sure I can pick it out." Then I'd do a blind listening test and I couldn't pick it out. The differences I was hearing would literally vanish before my ears. It was so dramatic and so crazy that it's like, "What the fuck is going on with this? This is too crazy. Is everything I've been listening to in the control room just a total joke? Is it not real? Am I just kidding myself?" So I went down this whole rabbit hole of investigating the psychological influence on sensory perception, this whole thing about expectation bias. If you have an expectation, your brain will manufacture the result that you're expecting to hear. Your sensory input is put through filters that will modulate what you're hearing to actually create the experience that you're expecting to have. It's very powerful. Way more powerful than the differences I was listening for in these little amplifiers. It was crazy. I went through all of these scientific papers, these abstracts that people had written about these subjects, not as much in audio, but more in taste. The realm people were very curious about this was in wine tasting. There are all these studies that have been done in wine tasting. They do this crazy stuff, like put MRI scanners on peoples' brains, and then let them drink different types of wine. They tell them different things as they're drinking the wine, like, "This is a very highly-rated, expensive wine." They have them drink it and then express the experience they had and measure the brain pattern that correlated to that experience. The crazy thing was that when they'd trick people by giving them cheap wine and telling them, "This is a very highly-regarded, great-tasting wine," they'd drink the wine and very predictably go, "Oh, this is a great-tasting wine." But it wasn't them just saying that. The brain response, the brain measurements they took from the fMRI scanner, matched the experience they actually had when they drank the expensive wine. So you are actually really having that experience when you expect to have it. Your brain creates that experience for you. It was a total revelation for me in everything that I did, everything in the studio, everything in designing the equipment and how to make meaningful decisions about the things that really matter and the things that really don't. Investing my energy, money, and purpose on things that will really manifest as things that will actually sound better and you can hear immediately, and not doing things that just don't matter. One of the things that was revealed to me was that this whole thing with an eight million dollar clock was a complete crock. It doesn't matter. Nobody has been able to hear the difference between one clock and another clock in a blind listening test. I've tested so many different people with it, and every time they fail. Even the person who works for the company who builds the clock failed the listening test. That's something that I'll never spend my money on. I have a large Pro Tools rig in the main studio in my place, so loop sync won't work. It's too long to cover all the converters that I have, so I do have a clock that distributes signal because it's better to do it that way. It's a healthier signal for everything. But it's not an expensive clock. It's $1,500. The system sounds great. So anyway, it's a lot of stuff like that. We were able to go through that process. When you're really dealing with the actual differences that are perceivable, you can start to tune into very minute differences when you're doing blind listening tests. You can train yourself to do that, and get much better at actually identifying real differences in things that I can pick out every time in a blind listening test where before I really couldn't. I didn't know how to focus the right way to be able to actually listen and really pick up on that stuff. That was a huge, huge part of the process that I found really interesting. I actually gave a presentation on doing comparative listening at the AES convention in Northern California and in San Francisco, I think in 2010. That was right when it was all really fresh for me.
Is that out on one of their white papers?
I believe it is. I wrote a whole thing on it.
So another unique aspect of the consoles is the surface, right?
Yes. That was another issue that I needed to overcome. It really came from the 88R. That console was so deep that the issue of reflection off the face of the console was really problematic for me. You could measure the difference. If you put an NS-10 or whatever speaker you like on a free stand in the room, put a mic in front of it wherever your head would be, and then measure the frequency response. It's generally quite flat. There might be a couple notches here and there. If you set it on the console, it's pretty flat up until about 1 kHz, and then it goes like this [Eric waves his hands up and down to demonstrate waves up and down] until 5 kHz. It's a disaster. And that's right in the range that's extremely important. All the presence frequencies that you're working with are just this comb-filtered mess. I'd sold the 88R while we were still building the UnderTone consoles. I put together a little makeshift setup. It was just a bunch of faders. We just made a summing mixer for a bunch of Flying Faders so I could do some summing with faders and have them automated for some mixing and stuff. I'd just patch in EQs and things when I needed them. So there was no desk in front of me. I got very used to the way speakers sound without it. When we started building a console again, I thought we have to solve this issue. I don't know if I can ever go back. The speakers sound too fucked up sitting on a console. I'm too aware of the damage that's doing. We were designing the circuitry, and I was counting on the fact that we'd come up with some sort of a solution along the way. I said, "I'm not building a console if we don't come up with a solution. I'll never do it again." So we had one idea of trying to angle everything more like a cockpit so that none of the reflections would actually affect the throw of the speakers and all this crazy stuff. But I can't climb into a cockpit in my control room with a bunch of people trying to do work. It's not happening. Finally one day, Larry came up with an idea. It's always Larry. He's one of these super-extraordinary, out-of-the-box thinkers. His brain is uninfluenced by convention in any way. He can look at things in such a unique way. I had a session set up with this goofy old AKG mic, this D200. It has this weird sort of molded windscreen. When you look at it, it looks very solid, but it's just a bunch of tiny, tiny beads of metal. Larry looked at the mic, and he thought, "This could work! This could be the surface of our console." It's transparent enough for you to sing through, so the sound is passing through this thing, and when you feel it, this is definitely sturdy enough to be a work surface. It turns out that it's made of sintered metal. Sintering is this specific process they use for filters. They mostly use it for fuel filters and stuff. They make these little disks, and depending on the size of the bead that fuses together, it'll control whether you get a 200 micron filter or a 100 micron filter or whatever. And so I found a company in Southern California that worked with sintered metal. I had them send me a sheet of it, actually a few sheets of it, so that I could simulate a console surface with it. I did a test where I set up outside, so there were no other reflections. I had an NS-10 speaker, a place where I could have a solid metal sheet, and a microphone. I set up the solid sheet and did a graph of the frequency response. I got the same results, lots of comb filtering in the midrange. Then you replace the solid metal sheet with the sintered metal sheet, and it's just dead flat. If you take the sintered metal off, the response is exactly the same. All the sound passes right through it. It doesn't reflect anything. It's astonishing how well it works. It looks very solid. It doesn't look like much would get through it, but it works amazingly well.
Is that what you're using on the rack here too?
We only did it, really as a novelty, on the first 100 MPEQ-1s.
Okay, so now it just sort of looks that way?
Yeah. There really isn't that much benefit to it. It's sitting in a rack. There is a thermal benefit, because you can vent out of the front panel. It works amazing on the consoles. We would have these sintered metal faceplates made that were a little oversized. They would laser cut the actual channel strip faceplate and all the little openings for the switches and controls and stuff. We used that for every single surface on the console, particularly the master section right wherever you're sitting. It works astonishingly well when you put speakers on the console. It's dead flat. It's exactly like having no console there at all. The console is open underneath, so the sound just passes all the way through. It's really like there's nothing there. It's really incredible. So we managed to solve that problem. That's the kind of stuff that Larry does every day. He'll wake up and eat a banana and have some coffee and go, "Hmm, I'm going to solve a decades-old issue with amplifier designs and just do a classic push-pull. Yeah, sure, sintered metal on a console." Now there's no reflection. We did the variable capacitance cable. That was something that was driving me nuts forever, the sound of different instrument cables. It was always very mysterious to me. I'd buy a really expensive instrument cable and think it would sound amazing. Then I'd plug it in and it'd be this dark, piece of shit. This cheap crappy cable I have sounds way better than this expensive one. What's the problem? I bought a big, expensive cable, but it was 20 feet long, and I was comparing it to a 10 foot cable. All coax cable, which is instrument cable, has a measurement, a rating for capacitance per foot. The more capacitance there is in the cable, the darker the sound's going to be when you're using a passive instrument. Larry finally explained that to me, he said it's the capacitance of the cable. When you add a capacitor to an inductor, which is what the pickup in your guitar is, it creates a filter. The higher the capacitance, the lower the cutoff point for the low-pass filter. It has a resonance to it. It creates a little peak and falls off. It's a huge difference when you have like a [Fender] Tele or a Strat or even humbucker pickups. You'll really hear it on a [Gibson] 335 or a good vintage guitar that's not too overwound in the pickups. Man, the difference between a lot of capacitance and very little in a cable is like night and day. It's a filter. It's filtering out everything above 1.5 or 2 kHz when you get up into the 1,000 or 1,200 picofarad range with a cable. You can never get those frequencies back. They're gone for good. There's no amount of turning up the treble on the amp that can get those frequencies back. They're filtered out. So Larry found this coax cable that had a super low capacitance rating. It's only 13 picofarads per foot. I made a 10 foot cable with that and had a cable that's only 130 picofarads. When you plug that in with a guitar, it's this amazing, crystalline, open, and huge beautiful sound. I thought, "Oh my god, this is the solution!" I was trying to use it on everything. I was torturing guitar players with these 5 or 6 foot cables so they couldn't move. These cables were incredible. I could have them longer and still get that beautiful, open sound. Then I started to realize that all of that high-end wasn't necessarily the best thing all the time. I was making the first record I did with Slash, and he had a setup that he had dialed in for himself; this modified Marshall head, a particular cable that he used, and his same Les Paul that he's used for years. He acquired it on the Appetite... record, and it's always been his studio guitar. It's actually not a Gibson; it's a copy of a Gibson, believe it or not. So his rig worked the way he had it. He has this super long Monster Cable that he used. I was like, "Oh man, wait until you try this other cable with super low capacitance. It's going to be incredible." I plugged it in, and the thing just squealed out of control. It didn't work at all. His modified Marshall has so much gain in it and so much boost in the highs and the mids that with my low capacitance cable, it would feedback uncontrollably. So he needed to have a cable that filtered off all that high end stuff. The cable he was using was 1500 picofarads. Super high-capacitance cable, specifically designed to do that. Now what do I do? There are times when you want to have more capacitance in the cable and other times when you don't, so I made a little box where you had some switches and could incrementally add more capacitance. We'd use one of the low capacitance cables to go to the box, and then I could dial it in to get to the sweet spot. Then I was talking to Larry about it one day, and I was told him, "This is a pain in the ass. It's really clumsy." He said we should just make a variable capacitance cable. He came up with this brilliant design where it uses this tiny component called a hexadecimal encoder where you can create a huge number of incremental values by combining a few capacitors. So you have 100 picofarad, 200, 400, and 800 picofarad capacitors, and this little dial on the first position uses none of them. On the second position, the 100. Third position, 200. Fourth position, it's the 100 and the 200, so you get 300, and on and on and on. So basically you can go in 100 picofarad increments from zero to 1500. That became the Vari-Cap cable. That cable is a life-changer for me. I will never record an instrument without that cable for the rest of my life. We just did a session where we had bass and two guitar players. Every person has to use that cable. I'd set it up and people would play, then I'd adjust that thing and get it to the sweet spot. It's amazing. It's just another one of those things where one morning Larry says, "Hey, we should make a variable capacitance cable." It's been this combination of me trying to figure out how to solve problems and having this extraordinarily brilliant person available to me to say, "Sure, there's an answer for that." He thinks of things in a way nobody else does. He'll come up with solutions to things in the equalizer design, the amplifier design, the physical design of the console, even something as conventional as the instrument cable that's never been done before, because he's just such a brilliant guy.
The UnFairchild's a little bit different for you guys then? That's emulating an older design.
Yes, it is. There was a different arc to the inclusion of that in the UnderTone world. There was another guy I knew, Garen Avetisyan, who was more of a hobbyist. He just enjoyed old tube circuitry. He modified a couple of my old tube compressors. At this point, Larry was very busy with UnderTone and designing new things from scratch, so he wasn't as available to service my old gear. Garen started servicing my gear and made his own custom tube compressors. One day he showed up at the studio and he said, "I made a Fairchild!" For that one, he used the basic topology of the original Fairchild circuit, but he made changes where he needed to. The original tubes, the 6386, for a Fairchild, when he designed the first UnFairchild, were $500 a piece. Insanely expensive. JJ wasn't making those tubes yet. So he made some alterations to the circuit to have it work with the 6BC8. It's the same tube that's in the Universal Audio 175 and 176. It's in the Altec compressors. It's a great-sounding tube compressor tube. He built this thing and it sounded really cool. It didn't sound exactly like an original Fairchild, but it had a cool, aggressive quality to it. A lot of color, and very musical. It was great. I thought, "Okay, this is awesome. First of all, I want one." It seemed like this wasn't that hard to build. We built ten of those that we sold to a very small group of folks. Those were all hand wired, and they were a bear to build. It's a lot of work. My Studio A tracking room became the UnFairchild assembly floor. We had three or four people down there just doing wiring harnesses; the whole thing. We built ten of them and sold them all. After that I realized, "Okay, there's a way to build a Fairchild without it costing an insane amount of money. We need to do this. This circuit is so good and so musical. I want to have it in my life, and I'm never going to pay $50,000 for a compressor. No way. Won't do it. I can build one from scratch for way less than that." And there were important things that had happened in how compressors worked and the controls people expect to have on a compressor that they just didn't understand in the ‘50s. It's not hard to make those additions to the circuit without losing what it originally was. After the first batch of them came out and JJ started making 6386s again. I said, "Let's design this thing so it can actually be built in mass quantities so we're not hand wiring everything." We switched back to the 6386s and really went for it. "This time, I'm not going to just add some extra presets that I think are cool. I'm done with that. Let's just put attack and release controls on it and have this thing function the way people want a compressor to work." That's what we did. Larry got more involved at that point and helped design the new timing circuitry so you could have the six original presets and then go into these variable modes that would allow you extraordinary control over attack and release. You could get the attack way faster than it ever was before, or way slower than it ever was before, and all sorts of combinations between attack and release. We made it so the DC Threshold is user-controllable on the front. It was originally just a calibrated setting on the back of the machine. It's an adjustment that puts the compression in a different range of the slope of the vari-mu tube, so it essentially becomes like a ratio control. You can have a much harder-knee compression or a much softer-knee, depending on where you set the DC Threshold and the actual main threshold gain. So we added that, we added a sidechain, we cleaned up a bunch of other things in how you can calibrate and tweak the unit with fine adjust for left and right levels; all the things you really need on a compressor if you're going to be using it on your mix bus or for mastering or something. The left and right sides need to match exactly. On the original, there was no trim control for left and right levels. It was a pain in the ass the few times I used them. You have to add an extra fader just to get left and right to still be the same level with 1 dB increments on the input gain control. We fixed all that stuff and maintained the integrity, and with the second version, we got it so that it really is like the original Fairchild with 6386 tubes. We only made small changes here and there. We designed our own custom input transformers for it with Cinemag, so we could do all of the M/S and the other cool functionalities with it. It's an incredible box. Even in my teeny little Topanga studio, I have two of them in that room, which is a total gluttony of tube compression. I use it on everything. I use it for tracking all the vocals now. I used to have this one 1176 that was like my magic vocal compressor. Now the UnFairchild is what I use every time on all the vocals. Guitar, piano, drums, bass; it's just a beautiful-sounding box. People have responded very positively to the UnFairchild. Building it in mass quantities, we're going to be selling our 200th unit coming up shortly.
Wow, so this next run of 100 will be 300 then?
Yeah, that'll take us out to 300 units. It's a wild ride building stuff like that. It's big and heavy and fragile and complicated inside. Shipping that thing to Uruguay or wherever they go. They go all over the world now. It's challenging. It's difficult to really foresee the destructive force that is FedEx and UPS. It's unbelievable, when you get 50 pounds of mass in a box. We've worked so hard on the packaging for that thing. It's double-boxed, double-foamed, super, super double-thick outer box, everything we can do to make that thing as indestructible as possible. Still, they find ways to destroy it. I don't know what they do, throw it off a third-story building or whatever. But now we have to stand behind them and make sure they're all working and reliable and serviced and that everybody has a happy, healthy UnFairchild in their studio, which I'm excited to do. It's an incredible box. I'm very proud of that thing.