Hello and welcome to the Sound On Sound podcast channel about electronic music and all things synth. I'm Rob Pericelli and in this episode I talk to Dr. Manny Fernandez, a man whose name might not instantly ring any bells, FM or otherwise. But whilst you may not have heard his name, You will have almost certainly heard his sounds that have featured heavily in almost all of Yamaha's FM synths from the DX7 MkII right up to today's montage and MoDX workstations. Tell me a little bit about Manny Fernandez before Yamaha. What was your first musical interest? What sparked the enthusiasm? We are fortunate that a close family friend of ours, um, is a, was a top flight jazz and L. A. studio guy. And he had a home studio, so I got exposed to him. An ARP Odyssey many, many, many, many moons ago, um, and you know, the, his kids and me and my brother, we would jam in his studio, uh, when he was out, uh, out working and that was sort of my first, uh, exposure to electronics was, uh, an ARP Odyssey way back in, I want to say about 77, 78. And, and, and then where, where did it progress from, from there? So what was your first, maybe, what was the first synthesizer you owned? Uh, well, thankfully, I didn't have to buy my own for a while because I had access to borrowing a lot of gear from our friend's dad, so, you know, I got to play with, uh, JX 3Ps, Prophet 5s, uh, Rhodes Chroma, uh, Prophet T8, all that kind of fun stuff, and moving his gear around and getting it set up for him. Um, so, I didn't really have to buy my own gear until I went off to college, and that's when I was able to, uh, buy a Pro 1. So I bought that brand new, and, uh, shortly thereafter acquired the ARP Odyssey from our friend's dad, and that was the first two synths I owned. You mentioned that you were doing programming for this friend's father, and was that how you got the bug into sound design for synthesizers? Was that where it all started? Yeah, I mean, basically that's where it, uh, it started, being able to just, you know, have a synth that you could just twiddle around and figure out what it does. Um, because at that point, you know, it wasn't anything that was learned. I was in high school and just twiddled around there when we got together and, you know, played with, uh, me and my brother and his sons in his studio, just doodling around. These were synthesizers that were, um, didn't have any patch memory. So, you know, you were creating stuff and then it was gone. And then start, yeah, start all over. Um, which I'll talk about a little bit later, but that's, that's a big philosophy that I'm actually a big fan of. Things that, uh, disappear. Before we get to your time with Yamaha and the FM synthesis of the DX7, how was your progression to that point? Did you get into sampling at all? Did you have any exposure to other digital technology? Were you aware of what was going on with Dr John Chowning? Right. So, um, so when I was in college, what was interesting is that, um, after I was there for two years, I found out that we had a really nifty electronic music studio at our, uh, at our, at our college. Um, I went to a, uh, a college called UC Davis, which is outside of Sacramento here in California. Um, and it's nearest neighbor is UC Berkeley. And so the gentleman in our electronic music department was out of Berkeley and was, uh, very close to Don Buchla. So I had no idea we had a full 200 modular system, a little electronic music easel in the suitcase, sitting here in a, uh, studio for anybody who signed up for the class to have unlimited access to. So, um, that was like, oh, hello, because. I was in college from 79 to 83. That's when all the synth bands hit, you know, big internationally. Um, and so that's what was my initial interest is, you know, that kind of stuff from a, from a pop, uh, rock kind of standpoint. But I was also into the, um, oh, let's call it the, uh, uh, ambient. Orchestral mishmashes. So tangerine dream, those kinds of things. So I listened to all the stuff. Um, so that's what drove me to learn analog synthesis. And I started taking the university classes to learn all the classic techniques. We literally started with tape tape loops, music, concrete. Editing your sounds as tones on cutting blocks, your envelopes of the shape, you cut the tape and going through the whole history of everything and learning all the, uh, science behind subtractive synthesis and about this time, you know, digital synthesis was also coming online. Um, and I picked up a, uh, Um, album, uh, one of the early, uh, Larry Fass Synergy albums. Oh, yes. And, uh, it had a section on one of his tunes that was generated on the, um, uh, digital system that Max Matthews put together back on the East Coast, which eventually became the GDS. And it was, uh, very enlightening to hear this rapid agitato solo string line that the harmonic structure on it was just so crisp and precise in the middle of his washes of analog stuff. I'm going, what the hell is that? Um, and that's when, um, I started trying to figure out how You know what's available for that. So right behind right at the time I graduated university is when This first Saint Clavier came out which came out of Stanford, which again is very local to where I where I was and That was just like a level of control and specificity of control that just really interested me, um, because the more I got into synthesis, the more I wanted to create exactly what structures I wanted to with the control I wanted of them. So, um, when, uh, the year I graduated, which was 83, is the year the DX7, you know, came out. Make a little side trip, um. I would drive into San Francisco regularly to check out stuff at the music stores and the original GS1 and GS2 had already made their debut. Um, and one of the original programmers for Yamaha, Gary Lewenberger, had a very nice store, um, in San Francisco. And I would go there on the weekends and hang out for hours playing gear I could never afford. So, um, I would just clomp around on that and, again, had some experience on playing those instruments. Back in those days, nothing was programmable. Um, but that's what really got my ear into, you know, these machines are just really cool. So, then when the DX7 came out, um, I just like, had to get one. Um, was able to get one probably four or five months after it actually was released. Um, thankfully my friend's dad, you know, had some um, Yeah, I got my, I got my DX7 probably, I want to say, maybe June, July, August at the latest of 83 went right to the, um, University of Southern California, academic library and picked out all the Chinese research and started reading it and started playing around and started deconstructing all the sounds. And that's how I started. Transition from analog to digital and that's how I got into digital. So you came at it from obviously from a Sinclair via route because if I'm not mistaken Sinclair via licensed sub licensed the technology from Yamaha who would license it from Stanford. So you kind of came at it through that correct route, so to speak. Yes. And because, uh, you know, doing the digital had the first. You know, implementation of it, you know, out there, you know, because the Sinclair one predated the GS one NGS two, if I'm not mistaken. Yeah, yeah. So then, then you got to the DX seven, which, um. To, for want of a better word, simplified FM synthesis compared to what had gone before. Um, now you had this, you know, programmable system, but it was very, you know, menu driven, not, not at all analog like. Then you said you took Dr. John Chowning's research and started getting into the weeds and, and breaking stuff down, so you kind of reverse engineered stuff. Well, yeah, um, to explain how that sort of goes is that it's like I had it on order, I was waiting for it. So it's like, well, let's start figuring out what it does. Um, and so what's interesting is in the generation of the commercially available FM synthesizers compared to the Original Sinclavier implementation, the GS1, the GS2 and the DX7. DX7 is actually the most complex implementation of it. Um, it just so happens that, um, you know, like anything computer or chip based, the first things are really expensive and all the trickle downs get cheaper as the power goes up. So, uh, the Sinclavier was, uh, um, single. Operator pairs, four of them, uh, the GS1 and GS2 were similarly, um, these types of pairs. They had some interesting, you know, interactions on a couple of them, but they did not have the massive algorithm choices that you had in the DX7. But getting back to the research, it's like, okay, this is something completely different. I know nothing about it. Let's start figuring out how it works. So. That's why I pulled the research up, because I knew we were building things from nothing, starting with sine waves, and I wanted to get some direction on, you know, what I was going to expect when I started changing ratios and modulation index and all that. And so, um, one of the things that struck me most about looking at his papers were all the graphs that he had of the index. Harmonic, the phase inversions, the Bessel function math and all that stuff to realize that they called it linear FM, which is the process of which it modulates frequency, but in reality, the responses that you got were anything but. So that came in really handy when I started deconstructing the sounds that came in the presets and started doing my own sounds already kind of knowing it was not going to respond any way that I was familiar with from an analog synth. Um, and I think that's what got me over the hump that so many guys didn't get over is that it was so different than. I mean, forget the single slider in the menus. It was just like, it didn't do what you thought it was going to do. Yeah. I have a friend who's a big analog synthesis fan. And we have lots of discussions about the merits and the shortcomings of both analog and digital. And the one thing he says that he doesn't like about FM... Is the unpredictability the fact that you don't necessarily know exactly what's going to happen when you change a particular setting, whereas with an analog paradigm, whether it's an R policy, a pro one, a C s 80. If you grab hold of the filter cutoff, you, you know what you, you know what to expect. Would you kind of go along with that? Yes, um, and I love this question because it's in a lot of, uh, conversations I've had, um, with the guys at Yamaha and, uh, other people regarding, you know, how, how I approach and why I look at that. There's a tiny disconnect that a lot of people miss, which is... Ear training and experience. It's only unpredictable because it's not similar to what you already have learned. And for me, in knowing the background and then pulling apart and deconstructing the presets, I was using all that time to train. My let's just call it here. In other words, I'm getting a sonic memory of what happens when you if x if I do x y happens, so this gets down to something that I've been told. John Chowning says as well. You don't need to know anything about how FM works. You just need to do it. Here at what happens when you do something, and if you'd like it, do more, and if you don't do something else, which honestly is how you approach analog synthesis. The difference is, is that one comes from building something from nothing, so there's a lot more work to do as opposed to having something and taking it out, which subtractive is mostly what happens in analog, so you can't screw up analog per se. Okay, but you can build something very unusual in FM. But at the end, you know, people have asked me this. It's like After, you know, how do you program now? It's like I do it all by ear, but it's because I have all the experience of I've tweaked so many interactions for, you know, in so many variations. It's like I kind of have a sonic memory of what's going to happen. You know, is it going to be exactly? No, but that's what you tweak. You tweak by ear. So over the years in talking about how to approach X, Y, Z or whatever, you know, people say, but what's the algorithm you need to do for that kind of sound? I go any of them. What do you mean? Any of them doesn't matter. There's nothing special about an algorithm. When people say, well, how do you know if you need a 1 to 3 or a 2 to 3 or a 1 to 4? It's like, well, okay, the math tells me I'm going to kind of get a ballpark, but at the end of the day, you had to just pick one and say, is that sound like how I want it to hear? You know? So, you know, it's like, um, the best way to explain it is like, you know, everybody knows how to color with some crayons. But, you know, not everybody can take a pallet of oils and make a fantastic painting, okay, without a lot of practice. Okay, yes, more practice is needed in FM, because it builds from nothing in ways that, um, have a much wider tamper range, frankly, than, uh, subtractive synthesis does. So you have more work to do. Um, and you just have to put the time in. And if you have good ears, you take the time. It's actually not that hard. Because that is the, the perennial, uh, comeback with FM. Is that it's just so darn difficult to get something out of it. That's what people always say. It's what I always come up against when I talk about it is, Oh, it's just so hard to program. And you have to go through menu diving. There's not enough knobs per function. And, you know, I, I get that. And I see recent efforts, um, in terms of third party controllers and software things that, uh, that supposedly make that that process easier. Do you when you're programming with FM? Do you use the controls on the synthesizer? Do you use software editors? Have you got third party controllers? What's your process? That is a very interesting question that you're going to get a totally bizarre answer. Um, a factor of How long I've done this and how many weird breadboards on a box I've had to use. In general, I'm a front panel guy, in general. However... An appropriately designed computer editor is very, uh, useful. Um, but I have some very strange idea on what that entails. And that is, for me, the ideal editor, uh, for an FM synth is essentially what they have up on Yamaha Sound Mondo for their Reface DX. Oh, yes. Okay, or how some of the old Commodore 64 editors were. Um, which is, I just want a spreadsheet with numbers. That's really all I want, you know, but because I know what all those numbers mean at this point and, um, so when I'm using, uh, an editor, um, what I like is I just like to put numbers in by 10 key. I like to tab shift tab and use cursor controls to navigate to my box and I can literally spreadsheet up a sound in no time. So I do use them when they are that type. Okay. Um, you know, I do. I've done a lot of work for Yamaha recently on montage, um, and they have a reasonable development editor, um, but in a weird way, since the navigations don't work the way that are comfortable for me from a speed standpoint, I Basically program on the front panel So It's the mid 1980s. You've got your dx7. You are learning it inside out from research notes and your own experience What happens that gets you to the, the DX seven mark two and programming, uh, those sounds that I guess for, for FM nerds like me are some of the best FM sounds out there. What, what, what was that, that process from owning a DX seven to getting to work on them? Right. So, so basically. At that time, I was just doing it for my own fun, you know, again, I was doing work for my friend's, uh, uh, dad. So because he also would use a DX in the studio and playing live with my own, our own band, I got the DX seven to, um, right when it came out, just because it was like, well, off. A lot of people love the 12 bit grunge, the DX7 1, okay? For me, it was, it was more annoying than it was pleasurable. And so, um, and being able to play with a DX5, and seeing the fact that you could, you know, duel up these things. Um, I thought to myself, Hey, well, that's really, really cool. So when that came out, the fact that it had the floppy drive and it had the dual engine, it's like, okay, I'm all over that. So I, I bought that, um, you know, again, very early on. So in those three or four year period of time, I had struck up a very good friendship with a, uh, owner of a local mom and pop music store. And when he had the DX7 2, I used to do custom libraries for him. Um, because at the time, you know, when you would buy from Guitar Center or Sam Ash or whomever, you know, and you were buying any Yamaha gear, everybody wanted to know, what library do I get at it? Because no one was going to program the thing. So, I started doing just stuff for him, so he had his own, you know, custom library. Um, and then... The first inflection point was you went from analog to digital and then you went from digital rev one or gen one to digital rev two because, you know, you always, you always had fair light choice, had emulators and stuff like that and they were trying to, you know, mix all these technologies. So, you know, the DX seven and it's sound, it was different. It was great. You know, where could you get 16 notes of aftertouch velocity sensitivity for two grand Jupiter 8 was 4 grand, right? Yeah. You know, kind of thing, so. Um, obviously for better or worse, it took over the market for synthesizers. And then you had phase two and phase two was the D 50 and the D 50 introduced sample snippets, but more importantly, onboard effects. So now there's a whole nother different sound, you know, that came in that was different. So, the fact I was doing custom work for my friend's store, when, you know, D 50 came in, that was where everybody was gravitating to. Because now it was the next shiny, shiny light bulb to gather everybody to. So, he said, it's just so sad, Manny. I said, I know you love the FM stuff, but it's kind of like dead. Because, you know, we're just not moving this. Everybody wants the D 50s. It's got like the new sound. I go, show me what it is. And everybody comes in and goes, wow. So, he, you know, played me a bunch of sounds. And I said, okay. So about, eh, two and a half weeks later, I came back with a disk. Hey, George, check this out. Load it up. And that's where I started doing what became sort of the sound sets that you talked about, which was utilizing FM to emulate, quote, other synthesis types. Or more importantly, just... Showing again the breadth of FM that, uh, you know, is possible, um, using techniques that I don't want, I don't want to say aren't readily, uh, evident, but again, you know, uh, mother of necessity, as it were, and that is, how do you make something that sounds chorused, how do you add some of these, you know, breathy noise bursts, how do you, uh, add things that are like pseudo reverb tails to give it, let's just call it a non FM sound. And so I made a disc of, um, you know, copying what he told me was everybody's favorite, you know, D 50 patches. Um, a lot of them were just dual voices appropriately done. There was a couple of them that were unison poly because that was one of the key features. Unison poly and the multi l f O mode that you could turn off the synchronization in terms of creating things that were very, you know, fair lighty, synclavier, you know, uh, D 50 ish. Um, and just doing that. So, um, we, you know, named them all the same, uh, voice names. So if somebody came in and just said they want this, that, but then they still want the dx what it, you know, it could do, which the D 50, you know, couldn't, in terms of a wider tambo range. Then he started selling DX sevens again. You know, and abnormally so, you know, his rep was like, seriously, this store is just one guy, you know, his mom worked in the back book in the music lessons. He had, you know, two part time college kids helping out and, you know, he was selling on a monthly basis. You know, as many more DX7s than Guitar Center was in L. A. And so, um, his rep said, What's going on? He says, Well, I have this disc and you need to talk to this guy. He's programming for me. Maybe he should do some work for you. So, that's sort of the long story. So, eventually my name filtered into Yamaha. And I got called to do a rescue project. And what the rescue project was, was to officially Do voicing of the DX7 2 to have sounds that were all in the timbre space that was popular for everything that was going on right at that time. So we did a series of a number of titles and the guys that were also involved, uh, doing that, um, from market research and other standpoints, uh, were, um, the core that ended up forming SoundSource Unlimited. Back in the day, so we did a bunch of accessories, voice sets. And, uh, it culminated into doing the FDE project. Uh, the gray matter board had come out, uh, for the Mark II, which increased a lot of capabilities for it. And, um, that was a really cool thing, that before other workstation type things were out there, so even before the M1, Yamaha had the ability to make their DX72 a workstation. Uh, they decided to make it an official product, because people couldn't get the boards installed necessarily, unless they were... You know, in the right metropolitan area, so we went into the warehouse and we opened them up, we installed them all, we soldered the boards, we loaded up the sounds, and Yamaha USA had a special SKU number that was the FTE that you could buy from a store right at Iraq. And so that bought it some more time again until the M1 came around and, you know, basically that was the transition of the last curve of that level of FM for Yamaha. So, not only the dawn of a new form of synthesis. And kind of mass production, but also the dawn of a whole new industry that you became, you know, heavily involved with with presets before then, you know, there weren't many programmable synthesizers with memory patches. Obviously, the profit five had those, but getting that data. Between machines was virtually impossible. Along comes the DX7, which has MIDI, of course, which you can transmit data between, but it also has the all important cartridge slot as well as the tape interface. And so now the distribution of preset sounds was. Was just easy to do so very fortuitous time for you to to fall into that really I suppose well yeah, and yeah, you sort of hit the nail on the head there, which was in DX7 was one of the original MIDI devices and you know at that point you were starting to see the original patch editors and librarians and such. And because it was awkward to program, you know, that did definitely start the third party business, you know, here, here in the States, um, you know, one of the first companies to really start that for the DX was a company called KeyClick, uh, which was started by Bo Toblin, who was a L. A. Session guy who had a lot of involvement with, uh, Yamaha, uh, at that time as well. And, you know, back in those days, Um, just with the MIDI interface, you'd, it was still on the five and a quarter soft floppies, you know, that was the format. Um, of course there are also people, um, and they, they did as well, sold, uh, the cartridges, you know, but the, the price of the cartridges was, you know, tough for real big mass market because, um, a blank cartridge from Yamaha at the time was like 90, you know, blank and unless you were going to make ROMs, you know, everybody was just overwriting RAM carts. And the, uh, up front cost to do that was high, so. Uh, but some companies did actually do some custom ROMs, especially in Japan, a lot of stuff came out. But yeah, the DX7 was, uh, really well supported with third party sounds, either through, uh, computer loading disks or, uh, special ROM or RAM cartridges for a number of period of time. Um, but it was still sort of niche just because the, um, you know, the costs were relatively expensive, but the Mark two, since it had the disk drive, that really was a game changer because you could cut the cost significantly for the user because floppy disks were not expensive and they were not expensive to duplicate. And so it really took off. And of course, everything else. Around it started coming out. Remember you and Sonic Mirage was starting to hit around this time. Um, and so you had the libraries for sample discs, you know, Yamaha had the TX, uh, 16 W, you know, and samplers and we're starting to come out. You know, they were probably the 3rd or 4th generation emulator. The product, the emulator to and people that bought that, you know, you were just buying those discs. So, um, As more of the computer based, uh, digital synths, you know, started coming to market, there was more of that. And so, yeah, it really expanded, uh, as other guys, you know, came on the line. You know, you got your K2000s, you know, Kawai, everybody's diving into the market, and they all needed sounds, and so... Um, yeah, it expanded really quickly. So the DX7, um, Mark II has had its time in, in, in the, in the spotlight. It's fading away. New boys are coming on the block. Um, Yamaha move on to the SY and the TG range, which brings in, uh, samples, you know, in a similar way to, to, to Roland and the way they brought samples in. But then Yamaha and they're mixing FM with. With sample technology and how different was that for you to, to now, you know, embrace this world of digital sampling. So, um, it's kind of interesting. I'm going to give you a little historical lineage on that transition. Um, because when you look at the 80s and you look at like, what were the kings of the market? You know, you had your DX7, then you had your D50, then you had your M1. And the M1 was, you know, D50 on steroids. Quote, full samples, workstation capabilities, better effects, more sophisticated, you know, mixing on that. So that's what set sort of the template of the modern synthesizer, for better or worse, that we still see today in a lot of implementations. And so Yamaha had nothing in that space, and a DX7 with... He was a very, you know, in the hands of somebody who would take the time. It was definitely usable, but it was, you know, just a bunch of hacks and, uh, you know, things to get it set up and going. Um, so they embarked on developing their next status of technology. So what was interesting is, The development cycle of a whole new thing from scratch that was outside of their thinking was going to take a certain period of time. So they were looking for a way to leverage a workstation, full on workstation products using their existing technology. So they brought in the gray matter response guys because they knew how to implement multi temporality on their FM chips, the sequencer functions and all that stuff. And so, um, they were trying to develop, um, FM based. Systems that we actually got to Proto's and showed at NAMM called the Yamaha V80. And of course, the V50 was the little brother of that. And it's kind of interesting on those two products because the V80, um, was quite an interesting beast because it actually had as part of its spec, the expansion boards. That they'd later had with all their PLG stuff, you know, in the CS6X, the S80s and all that stuff and that it was going to ship with the full blown sequencer and just an FM engine, but you had slots where you could put your sample playback card, you could put your drum machine card, you could put all these cards in and we showed all that at NAMM, um, in, um, private suite and it, it got some, you know, pretty good responses because it integrated the effects and, and all that. And, um, I don't want to say at the last minute, but the engineers were looking at that, looking at what they were going to be able to do to actually get the support. In other words, when you're going to get the drum machine card is really what it came down to, because until they got that, I mean, I love FM drums, and I'm pretty proud of all the FM drum work that I've done, um, that are fine to do stuff with, but you really needed to have the sample drums to play. So they killed the product. Because the engineer said we've got next gen SY coming along and we can move it up and get it out. In X timeframe. So rightly so they didn't want to spend the money to tool up a literally a temporary one year product. But what they ended up doing was fast tracking and getting the V 50 going. So Yamaha's first workstation was the V 50. And it's actually a really cool box in all honesty, you know, because it's got the effects chips. It's got drum samples. It's got a four op engine. It's got the same engine is the TX 81 Z. Um, and it's a really cool little box. So they, Okay, bought their little bit of time with the V 50 until the S. Y. Series came out. So the S. Y. Was originally conceived to basically be their version of an M one plus. And the plus was they're going to have multiple synthesis systems because they obviously had All this FM technology, so they developed their own sample playback system, which was a W M, which is literally what an M one is multi samples, filters and a sequencer. But they developed a FM. Which was their next generation FM and they put these two things together and, um, in the box it was really kind of cool. And I've spoken about this in some other venues in that. AFM and AWM can be used interactively in something that Yamaha calls real time convolution modulation, or RCM, which is using a sample as a waveform as a modulator of FM. Um, so as much as the DX7 was just sine waves, uh, the 81Z and the DX100 had, I think, their Six or eight waveforms that they had. The AFM had 16 waveform choices, but now you could go really complex with a sample. The, I don't want to say the intent, but one of the things that they had hoped to accomplish with their synthesis engine is utilizing That interaction as a re synthesizer. So there's gonna be a little technical for a bit, but Yamaha FM is not frequency modulation. It's phase modulation. Yeah. Okay. Um, and there is a lot of very important reasons why they do that. Um, uh, the number one reason for it is it's much more, uh, simple to implement in their chips, but it also does something very important without. Getting out of the frequency and going to phase, um, you can't do feedback. Let's just say you can't do feedback the way you need to. Because if you implement feedback in true FM, the pitch center changes as feedback changes. So in real time, you can't change a modulation index of a feedback operator. Otherwise, it's not what you want. But when you do it as phase modulation, you can. And so the Yamaha expansion of the original Chowning FM algorithm was Implementing feedback now. Why is feedback important? Um, it's important for a couple of reasons, but one of the main reasons why it's important is something you talked about how guys aren't used to how FM works. Feedback tames or maybe a better way to say it is feedback analog analogs up harmonic response with index. So it makes it behave more in tune to how a filter does. So it tames all the peaks and valleys in those Bessel functions so that you can get a sound brighter without changing the core balance of the inner harmonics. So that's how it came to be. So looping back around. Why is this important? AFM in the SY series, besides the waveforms, Actually also gave you phase control at every operator, so you have initial phase control. And what that does is that allows for, um, basically frequency shifting of whatever modulators coming into it. So. The early stages of the S Y algorithm, when you dive into the pages, they had something that would show up in the little screen of all the little patch points as O P H, and O P H was how they set up the F M section to be interacted from the sample section instead of using Channing's equation to use their modification of what's called a Hilbert transform, um, I'm going to butcher this. So anybody who knows all this specifically, don't put the comment section, but basically you can analyze audio signals, extract a bunch of phase information from them and then use less complex signals to reconstruct the original one by reconstructing the phase. So RCM, the original intent was to get dynamic samples playing back by utilizing. The phase reconstructing phase information in real time from all these FML operators. So, by having different phase shifts, different waveforms, and different velocity sensitivities and keyboard scaling and that kind of stuff, you could now take a static sample and get a harmonic playability that was more in tune to what a multi sample would sound like, but just using one sample. That was their goal of it. For a myriad of reasons it did not work as planned. So it was removed and They left the patch point there Where you can feed your sample in but it's just regular straight FM math And so a lot of people over the years is like wow, we got this and they go But what do we do with it? So here is the issue that it's not my issue But the people say it doesn't seem very useful And if you understand the math of FM, you'll understand why, is that FM does not need complex harmonic structures because the interaction generates tons of harmonics. So when you start using complex waveforms as modulators, you have a very limited range of modulation index before you just get I don't want to say unusable noise, but it just gets very, very harsh, very, very quickly. And one of the other quirks is, is that when you use a sample, all the harmonics are recorded from the original instrument and those aren't also perfectly in tune, which means you have all these out of tune things interacting with in tune things because the first five harmonics are in tune and all the upper harmonics are all skewed sharp and then they all come down. So that's why it does that. But what is interesting is Yamaha included a bunch of stuff in digital waves, which are what I call reduced harmonic complex waves. Those are the waves you want to do all your RCM stuff with, okay? Because now you have timbre spaces that you have that are unique, but they don't get out of control. But, you know, taking the piano sample and trying to throw it through the FM thing to make a super piano, okay? Basically doesn't work. Okay, or let's just say it doesn't work well. Do I have stuff that does it to give more playability? Yeah, but, you know, it's like a 10 percent gain. And that's just because you don't want to use a ton of harmonic content as a modulator, because it just gets out of control really quick. So, SYTG moves onwards and I can't remember if there was anything specific between those and because I think Yamaha just decided that FM really wasn't. Where it was at anymore until the nineties, and we saw the emergence of the FS1R module, which again was horribly, um, confusing to most people because it's this 1U rack mount, but it's got probably, certainly to my ears, because I don't own one, unfortunately, but to my ears, it's one of the most beautiful sounding FM engines, and it also builds in formant shaping, which, if my research is, is correct, um, Dr. Chowning said was the next logical step for that, that form of synthesis was to introduce formant shaping. And this, this module flopped. It was only around for about a year or so. Uh, and yet now it's one of the most sought after objects. Of many synthesists. I mean, I've been looking for one for years, and they, whenever they come up, they're ridiculously priced. Um, but you worked on that one. So how did you find that implementation of FM? Yeah, what's really interesting is, um, I hope you won't find this story just superfluous, but it's, Yalaha has always had an interesting manner in which they implement things in their synthesizers. And it goes back to their CS series. Um, in terms of like they do destination routing as opposed to, you know, source routing or some things just kind of like, you know, are quirky on how it puts together and it's not good or bad, but they just have a weird way of doing things and one of the things that for me and the S Y I it's got really, really good sounding digital filters, but their implementation is Exquisitely quirky and to get them to do what you would normally do on an m1 or Whatever derivative is actually kind of a pain in the butt. You can do it It's just that's so much more work than it needs to be And then the way that you do things like add velocity sensitivities to it and modulation depths. It's just weird. So the fs Comes around and there's two things that are really cool about the s the fs1r. One is it's the first Synth that they let me take that back. They kind of fix this all in the ex5 in the ex7 They finally did filters like the rest of the world Meaning on how you apply your modulations all that stuff plus it was multi mode and you know One of the things that happened between FS and DX7 and SY is the one you see behind me, which people don't realize is important for another reason, which is the VL. The VL, besides being a physical model, also did much more sophisticated filter modeling and effects modeling. And so that was carried through the EX5. So Yamaha is really getting much better at doing their modeling of everything. So now you get the FS1R and you have multi mode filters that sound really good. And the structure of it is basically carried through to what's in montage. I mean, they've basically kept just, you know, improving clock speeds and stuff of those implementations. Um, So that's cool. And then, you know, the velocity works right. The, uh, modulations to attack rates and all that stuff. Everything works the way you anticipated to work. So they fixed it up and that's really cool. And then the second half is that they have this new engine that has the noise operators and the format behavior of the regular operators. So it's very interesting when we were doing that project because You know, sometimes we get things that are very well specified, and sometimes you get things dropped in our lap that fall under that we don't know what it does. Um, the VLs was obviously one of those projects, you know, like we know what we designed it to, but we really don't know what it does in a musically useful way. And so the FS was kind of like that as well. Um, the mindset of the development of it is. What is FM good for? And that is dynamic harmonic response that is very quote acoustic or real instrument sounding, you know, but what is it not so good at? Well, it's not so good at Noises. And if you really want to do acoustic instruments, there's lots of, um, harmonic content. There's a lot of noise content, and there's a lot of format behaviors. So that's where they said, Okay, well, let's just put this together. So we get this plopped down and we start to play around with it. Now we gotta figure out what the hell we're gonna do with it. And we found a couple of really interesting things that you can do with this box. One is it's a very low phi sampler. Which some guys have really, you know, thought is really, uh, uh, cool use of it. Um, it's also an exquisitely versatile vocoding type of device. Um, and then it is now a enhanced FM engine. Uh, an FM synthesizer as well. So you have all these capabilities in the box that you can't see a dang thing from the front panel with four knobs and menu diving. It's just like, uh, now, thankfully, I never programmed it from the front panel always always had a dev system that, you know, eventually a version of it was what was released with the editor that came in the box because you have to, you have to bring them in the box. But, um, what was interesting in what we We're doing with this is something else that they did which was it read original DX7 SysX So what was cool about that was, you know enough time had passed It was kind of like a little bit of an FM, you know renaissance Um, but more importantly, hundreds of thousands of sounds that you could fly into this box and so you can use them just as is. But more importantly, put great sounding filters on a great sounding effects. So a lot of stuff that people wanted to always do with their DX7. Um, you know, you could do right off the bat. So, um. What we started to do was to enhance those voices for better acoustic control as it were. So we would take these format operators to build in formats in the guitar sounds, in the road sounds, use the noise operators to do more thunks and clicks and clacks and pick noises and put all these things on the knobs on the front of it so that you could actually control some of this stuff in real time. This is just speculation on my part, you know. Because it literally, you couldn't use it without a computer. Okay. And if you didn't like what was unique about it, which was that weird lo fi sampler vocoded type of sound. Okay. It, why would you, it was sort of like lost in the market as it were. And, um, so yeah, it was a failure to the degree that, you know, they were. You know, I don't say they couldn't give them away, but you know, they were selling them ridiculously cheap to dealers. You know, they were offering ridiculously cheap to employees. So, you know, regardless of what it goes for today, you know, I I'm smiling because you say you've been looking for one. I have one new in box unopened. Sitting on my shelf. Um, that I bought for 350 back when they gave him away and it's still in the box. Um, but what's interesting about it is like anything when things are really cheap, a lot of people get access to it that wouldn't otherwise, and they find some unique way to use it. So, you know, Some people think this is like totally off the wall, but it is absolutely completely accurate as a, as an analogy. So the FS1R has become legendary for the exact same reason that the Roland 808 drum machine and the TB 303 did, which was they were flops. They were dirt cheap. A bunch of people got them and then found some new thing to do with it. And that's what happened with the FS1R. Is that when you had the chance to dive into it, it's like, wow, this does a lot and more so than other Yamaha products, you know, it's suffered from, you know, the disease that all of us call shipping deadlines, right? How much can you accomplish in the sound set on the time frame that you have available? So, uh, the thing that it just became really known for is these. You know weird use of the format sequences and all these weird form behaviors throwing them on Controllers and it was a texture and a sound that if you're into certain, you know types of you know Musical genres is it's the only thing that can do it and it sounds fantastic Doing it and so a number of very creative individuals started turning out content with it and people go what the hell is that? You know and um, yeah, so it's it's it's a really cool box the approach that I had from Programming it is very similar to what I do with everything I get from yamaha, which is they give it to us There's a spec There's an outline of what it's supposed to do, and I throw that right out, and I start poking in all the places to try to literally do everything it's not supposed to. In other words, the first thing I do with any of their systems is I try to break them. So, I investigate the extremes of every parameter. Um, I, whatever if, you know, nobody in a million years would want to patch this to that. Well, I'm going to do that first. Because again, I want to find out what quirks it has, because if I want to make cool... interesting stuff. In my experience, those are always in the cracks of the synthesis system, so you want to find those right away. So, um, they gave me the original analysis software, um, to generate the format sequences and stuff. Um, and so I really played in a really wide spectrum of totally ridiculously avant garde stuff to very standard DX stuff to try to find what the sweet spot of it was. And, um, for me, You know what? I wrote this in a report, and I said, I think how you're going to catch most people's attention with this is what it does in a pseudo vocoder sense, because that is usable without having to be editable, whereas using as a lo fi sampler, you can't unless you have all the software and all that kind of stuff. So I suggested a bunch of stuff. I gave him a bunch of stuff. Examples of things that I thought, you know, you should have for four man sequences and send it off to them. Next thing I know, things that I suggested to them to do professionally. They ended up generating off the cassettes I sent them. I literally sent them. You know, cassette, because a digital file, you know, back in those days to transfer over the internet was like, you know, ridiculous, so I gave him a chromium cassette of a bunch of things that I thought would make for good format sequences of me doing them vocally, you know, some of them were, you know, whatever sound, some of them were just, you know, I don't want to have to build, you know, by scratch, it's like, analyze those things. So I was just doing things like that to pop stuff in. So anyway, I just thought it was hilarious that Oh, really? You used what I sent you? But the first demo in it, you know, utilizes that aspect of it. Which is the, um, what I call the chordal vocoding, you know, it's a monophonic format sequence that you can play over, uh, chordal voicings that gives you a really cool talking synthesizer, um, thing, but it also has format sequences that you do that you get those timbre shifts over, over moving, um, harmonic structures that you can't generate any other way. And yeah, it's just, it's a really cool box that, Nobody appreciated in its time. So the FS one are brought in this. Um, well, it wasn't the first obvious because there was an eight operator FM expansion module, I believe, for one of the electron organs that's quite rare. But, um, it kind of popularized eight operator FM to say the masses, but probably not so much because of the numbers that it sold. However, uh, In recent years, Yamaha has kind of come back with, uh, with its FM, with the montage, and of course the MoDX, which takes that whole FM engine, just puts it into a, a single unit, and of course we've got the Reface DX as well. Now, I understand you've, you've worked with all three of those machines. Yes, um, It's, it's a funny story, uh, actually, um, after trying to think after FS1R, you left one out. You left out the, uh, DX200. Oh, I did. Yes, that's true. Um, and the only reason why I mentioned that is because the last, so from. Basically, DX7 through AN and DX200. You know, I basically worked on almost everything that came out of their ProSynth division, but pretty much everything. But those were the last two official products I worked on when I went on sabbatical. Basically getting on with the rest of my life, having a family and all that kind of stuff. Um, and so, what was interesting is, Um, I still go to NAMM regularly and I'm still very friendly with all the Yamaha guys and over the time when I was in doing this is, um, you know, I did stuff for Korg, I did stuff for Kurzweil, I did stuff for, uh, Ensoniq, I did some stuff for Emu, um, and so, you know, I know all the guys and that stuff, so I always like to go to NAMM shows just for, you know, meeting people and, you know, seeing gear. Um, and, yeah. I remember, um, hearing about, you know, montage and the reface, uh, DX, you know, coming out. And so I went to the NAMM show where they had those, uh, out and was, you know, playing around with them and I ran into, uh, my good friend, Nate, uh, Cheddar. And I said to, I said to him, first thing I said, Hey, Nate, how's it going? Yeah, well, that's good. So I just looked at him straight in the eye and said, Dude, you guys released FM products? And didn't call me, you know, my number hasn't changed in 15 years and he's he smiled. He goes, I have a project for you now. Truth be told, partially part of that was my fault because, um, well, yeah, my montage and reface were still in development. They had this big 40th anniversary celebration here in L. A. And I was at that and Nate at that time said, I have some stuff I think you might be interested in and I don't remember what happened, but I never followed through to call him. So it's not really all his fault, but I I neglected to follow up for whatever reason. But yeah, so they introduced those two products that have FM implementations and they're really both very interesting for a couple of reasons. Reface is Very interesting because of how they implemented it. It's a lot more versatile than people think. And, um, one of the things that Nate talked to me about at that NAMM show was he said, you know, Yamaha's got all this great history, but we don't have any like official content on anything on how to program it or approach it. Would you be interested in doing something for us? And he talked to me about Thank you. What was going to become the sound mondo platform and the Yamaha synth, uh, you know, page on it. And that came into me starting to do my tutorial series. So as part of the Yamaha synth website and with associated libraries and examples through sound mondo, um, you know, we put together, you know, a quote official, you know, Yamaha series of FM programming techniques and stuff. And so, um, I started that, uh, doing that on reface. And, um, it was very well received, we did a lot of, uh, uh, patches for it, and it's, you know, it covers a ridiculous, you know, breadth of, uh, tambours that you wouldn't anticipate from a 4 op engine, but it has to do with the fact that they're, uh, that, um, square wave, saw wave feedback, and a couple of unique algorithms, and this is the thing that some people, you know, don't understand. Just having an onboard effects chip, even if it doesn't do a lot, does a lot, you know, um, so, um, we did, we did that full project and then, um, did the same thing for montage explaining FMX because FMX is the strict FM side of an FS1R. Then. That got me officially back into develop doing development voices for all the various firmwares that came up as they added new features so starting with the, um, 2. 5 was it 2. 5. Anyway, the last two firmware updates I did country contributing voices for again and and such so yeah so I'm sort of back in the fold and it seems FM is making. Another comeback because it it never goes away. It just kind of fades away. And then something new happens. You know, we, for example, you know, we get the S. Y. And the T. G. S. With the addition of a W. M. And A. F. M. And then you have the F. S. One R. Which is extra operators and formant shaping. Now we have Montage, um, mod DX with the, with the F M X engine, refacing, that nice little compact unit. And there are whispers and rumors that other companies are working on FM synthesizers, Corg have got the, the ulca, the, the little tabletop thing that that even reads. Um, DX seven csx. As somebody who is in at that kind of, that level that you are, where do you think FM can go next? Or does it have a future? Oh, I think it has a future. Um, I think the reason why it's still around is because You know, I believe it is of all the things out there. It's the sweet spot of versatility versus complexity. And I say that because the context of what I do as a sound designer, anybody who just wants to do whatever. Okay. There's a famous, you know, uh, Wendy Carlos quote, which is the, this is a paraphrase, but it's like, you know, the best thing about additive synthesis is that you can create anything. The bad news is you have to create everything. So if you think about the context of I can control an infinite number of harmonics with an infinite number of staged envelopes to make whatever movement I want with any controller. Yeah, you can make any sound in the world. You know, it's one of the reasons why I wish I had a G D SS system or a synergy with a programmer. Um, but the, the time is ridiculous. Okay. So FM is in inhabits a really nice. Space for me, which is, I can get so much of that core level of control without having to do that much work. Now, it's more work than an analog synth, but an analog synth or subtractive synthesis or any of these other types of synthesis don't give you the timbre space they have in FM. So for me, it's always going to be around because unbeknownst to many people is the techniques that it uses. Can sound like whatever the hell you want it to. One of the things that's quite interesting is that, you know, I've been programming my S Y for what? 27 years. Um, I can pretty much make it sound like whatever the hell I want to. And if you've ever gotten up to the, uh, S Y group on Facebook or seen some of the postings on YouTube, um, FM can be virtual analog, it can be gen one, shut up, don't play electronic piano one preset again, electric piano sound, um, it can mimic physical modeling, it can do all these things, um, and it does so in a manner that is so computationally efficient that it has a lot of interest, um, I do a lot of things with montage that are, um, Um, the biggest strength about montage is less the FM engine because there's some significant limitations. It has some significant things that the S Y doesn't, but what it has is the real time controllability of so many important parameters, which none of the other systems have. So, you know, I have stuff on the montage that it's gonna sound really weird. Yeah. It sounds more, I can make it sound more like a VP1 than a VP1 does, and it takes a lot less effort for me to do it. Okay. Now, obviously, I have an experience and background with all those things to be able to do that. But the point is, is that it's capable of it. And so when you see things like FM7 and FM8 that are available as SoftSense and, um, you know. All the dext and all these other kinds of things. It's just so easy to implement and it's got a huge range and if you take the time and don't fall into the stereotype I mean you get a level of control. You don't get an end to other other synthesis systems with so much less effort Okay, and create anything from I can I can make you the biggest fattest silkiest, you know Analog sample hybrid string patch all in FM. Okay. And I can also, you know, make a high index sound that's going to take your head off, you know, and then you can go and do real subtle shifting stuff that, you know, did in, you know, uh, music for airports. It's, you know, and there's not that many, uh, synthesis systems that can literally go from zero to a million and back with a. Moderate amount of input, Manny. It's been an absolute pleasure to hear your insights into this, to your experience and your history. So thank you ever so much. Thank you very much for talking to us today. Yeah, no worries. And here, let me, let me, let me play you out with something. Thank you for listening and be sure to check out the show notes page for this episode where you'll find further information along with web links and details of all the other episodes. Before you go, make sure you visit the Sound On Sound podcast page at soundonsound. com forward slash podcasts, where you can explore all the other great content playing across the other channels. I'm Rod Pericelli, and this has been a failed Muso production for Sound On Sound.