บทถอดความฉบับเต็มของ Wavetable Synthesis - Single-Cycle Waveforms

practical terms, is there anything you'd highlight for what that approach gives you as a player? 00:06:23,424 --> 00:06:24,224 [Anthony Marinelli] Sure. Well, for me, I dedicate my life to practical terms because I just wanna create music. So to me, it's... I call it additive synthesis, and it's a great name because you're adding harmonics. And that is kind of like what the digital world of synthesizers likes to do. It can do other stuff, right? You add harmonics. And then analogue synthesis is subtractive synthesis, and you basically remove harmonics. So when I look at the 2, I go, "Well, do I wanna add harmonics as I go musically, or do I wanna subtract harmonics as I go?" And basically, I'm just shaping tones the way you would on any instrument. On a clarinet, I'm getting brighter. Well, what's happening? I'm adding harmonics. So you're just choosing different ways of doing those things. And then as Bob got into more nuances, there's a lot of detail in additive synthesis that you can control, whereas a filter is when you're removing harmonics, has more of an overall shape to it, which is very musical. And you can do that with additive synthesis, but you can also with additive synthesis, which is what we're talking about here today, you can pinpoint certain harmonic. Well, what's a harmonic? Harmonic is a musical thing. So it starts with the fundamental. If you just go in the basic overtone series, then the octave, then the 5th, then the octave again. So now you're at 2 octaves. Then basically you're into, like, triads, and now you added the 7th chord into the 3rd octave of the overtone series. You have the fundamental, the 3rd, the 5th, the 7th. Then you get into the 4th octave of the overtone series, and you have the 9th, the 11th, and the 13th. Then you get into more of the dissonant stuff. So it's all musical. So I'm just looking at it that way, if that answers your question. And you combine them. 00:08:27,654 --> 00:08:31,804 [Bob Coover] And what's interesting about what Anthony said, going into the history of this, whereas the PPG was the 1st commercially available wavetable synthesizer, and those were just recordings that, as Anthony just said, usually had very similar harmonic structure as you walk through it, The Fairlight, which was also actually a wavetable synthesizer, but they never said it, that actually was literally additive. You would take and you would add harmonics to a given what they called a frame, and instead of having 64 single cycle waves, you would have 128 frames and you could morph between them, like only set up 2 of them, or you could actually walk through them with each one defined, just like a wavetable. So essentially, Fairlight and PPG were both doing very similar work at the same time, probably not knowing that they were doing very similar work at the same time. But that was their additive, literally additive way of dividing up a wavetable. You tell it what harmonics you want, and you walk to the next frame, as they called it. 00:09:46,664 --> 00:09:53,964 [Anthony Marinelli] I'm gonna ask you Bob, too, but I look at it like a wavetable is the static version. It's a two-dimensional thing. 00:09:53,964 --> 00:09:54,444 [Bob Coover] Exactly. 00:09:54,444 --> 00:10:57,324 [Anthony Marinelli] It doesn't live. And so what Bob is saying, it's like when you're watching a movie and there's persistence of vision happening. You're just looking at light flashing, but it's flashing fast enough, more than 16 frames a second, so that your eyes perceives it as continuous motion. And so with the wavetable, you have the static thing. Well, that's fine. And that's just like a sawtooth wave or a pulse wave or anything. It just sits there, and then you can filter it or whatever. But what he's talking about is stringing them together like frames or sprockets on a reel of film, and then you get them to either crossfade or just, if they're quick, if they're close enough, you don't even have to crossfade very much, but you get it to move. So when you just say wavetable, it kind of doesn't give it the full justice of its musical value. I think its musical value comes a lot from getting it to move, 'cause music is just not static wavetables. It's motion, and that's why I like the additive synthesis term, because you're adding harmonics, and that's what makes it live. 00:10:57,324 --> 00:11:05,204 [Bob Coover] And you're absolutely right. Yeah, if you move through that wavetable, you can create just about anything you want. 00:11:05,204 --> 00:11:08,144 [Anthony Marinelli] Right. Like, better than a sample. 00:11:08,144 --> 00:11:27,244 [Julian Rodgers] Am I doing the idea justice if I kind of say that with a wavetable, one of the really principal advantages of it is access to alternative harmonic content? I mean, we put so much work in subtractive synthesis into trying to introduce movement into other essentially static sounds. It's right there. You can just right here, there, jump to the next one, jump to the previous one or further. Is that a fair statement? 00:11:36,864 --> 00:12:14,328 [Bob Coover] Right. Yeah, yeah. Because what you're doing and again, it's arbitrary how long the wavetable is. PPG and what we're doing with the 3rd Wave, and a lot of people tend to run with 64 single-cycle wave forms, right? You can make it longer, you can make it as short as you want, even though I would call a Prophet VS not a wave table synthesizer. It's only got 4 wave forms, but you're still moving between them. So any time that you're creating movement between single-cycle waves, you've got something that could be at least loosely defined as a wave table. 00:12:14,328 --> 00:12:21,048 [Julian Rodgers] So morphing between the different, the Prophet VS had a, like an XY controller. I mean, I don't know what similar there is with things like, I don't know, the SYs and things like that, that also had similar things. But being able to access... 00:12:27,728 -->

00:12:28,308 [Bob Coover] Very similar. 00:12:28,308 --> 00:12:35,368 [Julian Rodgers] Okay. There's probably a story there. But being able to quickly access alternatives is a theme here. 00:12:35,368 --> 00:13:25,478 [Bob Coover] Right, right, because you've got vector synthesis as the Prophet VS, you know, that was 4 wave forms, so it was much, much more abbreviated than a full wave table. And you've even got things like currently UDO, they have their Super Gemini, Super 6. Those aren't wave tables, but they're still moving, especially the Super Gemini. It allows you to pick 2 wave forms and move between them. So you're still getting that morphing movement, if you will. The thing with the wave table, a more fully blown wave table like what we're talking about, is that you actually get to pick a lot of different wave forms, and you can actually make it talk to you. You can make it sound like, I don't know, whatever you want basically within that framework. 00:13:25,478 --> 00:13:36,548 [Anthony Marinelli] And there's one thing, like, I just wanna add one thing, or 2 things actually, but they're quick. You can get movement, moving wave tables with analogue synthesis. It's called pulse-width modulation. And you're changing the wave form. So, you know, if you don't have this, you can just hear it, what it sounds like to morph, basically, a wave form. And secondly, it's also called sometimes re-synthesis, where you're synthesizing a sample using wave tables connected together the way Bob was describing, or sometimes frame synthesis. Synclavier was doing this. Cameron Jones came up with this in the early 1980s, and he was doing re-synthesis the same way with the Synclavier, using these wave tables and cross-fading them like a movie. And it was a little more primitive than what Bob's doing now, but it sounded great. And to your point, Julian, you could get into the middle of the sound, and you could elongate it. So if Bob said a word like Bob, I could have it go Boooob. So now I made the ah, the O, like a sample, you can't do that. So it's more like a real instrument, like a voice or a acoustic instrument where the player could get into the middle of the sound and express it. And that's the beauty of this. 00:14:42,528 --> 00:14:46,847 [Julian Rodgers] We've kind of wandered unintentionally. Well, so semi-intentionally... 00:14:46,848 --> 00:14:47,838 [Anthony Marinelli] We wandered into music. 00:14:47,838 --> 00:14:50,188 [Julian Rodgers] Into, into the next point, well, into the next point which is we've been discussing kind of like historical roots. We've mentioned the PPG a couple of times. We've mentioned Synclavier. So I mean, a potted history, this could go long, so we should probably try to limit ourselves. But a potted history if you like of milestone moments in development. PPG was 1st? Is that fair? Or is it... 00:15:13,268 --> 00:16:07,418 [Bob Coover] PPG was 1st as far as commercially available. Now, again, Fairlight never

called it a wave table, but they were around the same time. We're talking late '70s here. And then I think the PPG 2.0, I gotta have my history right so don't hold me to this, I think it was 1980. Then we got to the PPG 2.2, which was 2 oscillators, and the 2.3. By then, Fairlight had moved on to sampling. Synclavier, which of course Anthony's got, they had also entered the sampling game and then had started the re-synthesis, which is what he was talking about, where again, you're taking frames, just like what Fairlight was doing, and you're stringing them together to re-synthesize a sound. That's a wave table. So you've got it any way you wanna cook it. Those 3 were kind of, they were super expensive, do not get me wrong, but they were also, doing roughly the same thing, or at least circling around the same area of sound synthesis at the same time. 00:16:19,268 --> 00:16:39,448 [Julian Rodgers] I mean, 1st generation, I mean, PPG probably is associated with this because, I mean, correct me if you don't agree, but I'd say that the Synclavier and the Fairlight were kind of overshadowed by sampling capability, and that's the thing that people hooked onto and possibly glossed over some of the other stuff that they did. 00:16:39,448 --> 00:16:45,408 [Bob Coover] Yeah, exactly. And PPG historically were the 1st ones to call it a wave table. 00:16:45,408 --> 00:16:45,598 [Julian Rodgers] Okay. 00:16:45,598 --> 00:16:47,488 [Bob Coover] That's probably the biggest thing. 00:16:47,488 --> 00:16:59,648 [Julian Rodgers] Look at this bit here, yeah. And I suppose, so was there like a next generation, a maturing? I mean, Anthony, so any notable products along that timeline that you'd like to think deserve highlighting? 00:16:59,648 --> 00:17:01,928 [Anthony Marinelli] You mean for doing like wavetable-type stuff? 00:17:01,928 --> 00:17:02,588 [Julian Rodgers] Yeah, yeah, yeah. 00:17:02,588 --> 00:17:15,468 [Anthony Marinelli] One thing that the Synclavier didn't get looked at as much was, it started out as an FM synthesizer with wave tables. So you could create a wave table and then apply FM to it. And that's kind of the basis of the Synclavier. And then sampling happened later. And then you were able to mix and match. The Synclavier didn't have a filter, but it allowed you to very easily send out control voltage, which I did all the time, and I used Roland filters. I used the Roland 100M so I could do what the PPG was doing, which is a beautiful sound to be able to still filter, you know, wave tables. But I don't think back then I could think of anything else. There might have been. Well, there was like the thing, what was Wendy Carlos using, that was a wave table. 00:17:51,828 --> 00:18:22,968 [Bob Coover] Kind of. That was additive. Yeah, that was a Synergy 2+, as they called it, or the general development device from Bell Labs. But yeah, that was, and that would also, believe it or not Anthony, I didn't know this until, you know, the last

couple of years, that was supposed to be additive. It actually did FM as well, and it would allow you to do FM with triangle waves, not just sine waves. So that was interesting as well. 00:18:22,968 --> 00:18:25,328 [Anthony Marinelli] Circling the same wagons, though, like you're saying, Julian. 00:18:25,328 --> 00:18:27,568 [Julian Rodgers] Absolutely. 00:18:27,568 --> 00:18:56,847 [Bob Coover] And the interesting thing, I'm gonna make sure we add this in that Anthony just said and that hits the nail on the head, that Synclavier didn't have a filter. So you were literally animating the sound by moving through harmonic structures that changed, whether it was a wave table, whether you were running a FM envelope on it and creating side bands, you were moving not by doing it with... 00:18:56,847 --> 00:19:04,788 [Julian Rodgers] But it was bottom-up synthesis rather than top-down, which hence, I suppose, your distinction Anthony, between subtractive and additive. 00:19:04,788 --> 00:19:05,948 [Anthony Marinelli] Yeah, you could look at it that way. 00:19:05,948 --> 00:20:15,628 [Julian Rodgers] Excellent, so I mean, we should, we've been discussing oh yes, it's this, it's that. But I mean, some people listening to this might not be familiar with the timbres we're talking about. I remember trying to, I used to teach and talking to a student and saying do you like oscillator sync? If you like oscillator sync, you're gonna love this. And that was kind of my in. That was the kind of music that people were making at the time, lots of kind of like gnarly, dubstep-y things and definitely all software-based. But we can do better than that because we have, we should err on the side of audio for this. But I know we can kind of play some examples and possibly show some stuff. What does it sound, I mean, what's distinctive about the characteristic sound of wavetables? Not what it can do, but what it's used for is probably what we're talking about. The defining feature is harmonics, as pointed out by Anthony. So what exactly is it that would catch either of your ears and go ah, wavetables? 00:20:15,628 --> 00:21:00,248 [Bob Coover] You usually get this, so you can make it sound natural, as we were trying to say, including, you know, trying to re-synthesize a sound. But you usually get this very distinctive kind of industrial if you will, jumps and things. I'm gonna go back in history again, think of Depeche Mode, "The Policy of Truth". There's a very distinct sound in that chorus and it even ends the song with just that wavetable walking through. It doesn't sound natural. It sounds like oh my God, what is that sound? How did they create that? Because you can walk through very similar waveforms and make it, you know, do something somewhat... 00:21:00,248 --> 00:21:02,098 [Julian Rodgers] Behave like sound behaves in nature. Yeah, yeah. 00:21:02,098 --> 00:21:22,468 [Bob Coover] Yeah, but you can also make it jump and do really odd things and you

know, Trent Reznor, Nine Inch Nails stuff, and you know, just really wild stuff by changing, you know, what you're jumping from and to. And I'll play you an example here in a little bit that is oh my God. 00:21:24,008 --> 00:21:24,348 [Julian Rodgers] Okay. Cool, cool. 00:21:24,348 --> 00:21:45,797 [Anthony Marinelli] Well, we tend to define the sound of wave table synthesis by the primitive version of it, because when it first came out, it was like 8-bit, and it was just trying to make any sound, like in a cheap way. That was kind of what some, like video games and a lot of that stuff was wavetable synthesis. But where Bob's taking it, it can be, like we need to rethink it because how it was and some of the iconic things maybe you're gonna play or think of, those are more like... and it's endearing for some reason. Like, people love the past. But its primitive version of wavetable synthesis. It can sound more organic, really, than a sample because it could be as accurate as a sample now because of the resolution capabilities, and then move. So I wanna make it clear, I don't know that we're really able to define it with examples because the technology is just catching up to itself. 00:22:16,518 --> 00:22:40,968 [Julian Rodgers] A good point and you, I mean you mentioned sampling and I mean, early sampling grabbed the ear because it sounded like something real. But of course now, sampling just is something real, the distinction's fallen away. So I suppose that's possibly what you mean, but obviously with the flexibility and real-time control and all of the things you can do when you get away from sampling, so yeah. Interesting. 00:22:40,968 --> 00:23:13,428 [Bob Coover] So I'll play you real quick, so let's go back in time. This is the PPG and I'll follow it up by the same thing on the 3rd Wave, but then I'll play you one of those modern, like really, like it's almost even not musical. So here's a PPG [music plays], and the same thing [music plays] on the 3rd Wave. But now let me play you, if I can get over to bank 5, and if you guys have a PP or a 3rd Wave out there, this is something called Wavy Gravy, and it is [music plays]. Hear all that crazy movement that you're hearing? It's almost not even musical. And that's what I mean by you can do very nice-sounding things, but you can also do very jarring things as you move through it. And that's the beauty of a wavetable. Make it sound as nice as you want, and there are some very nice examples out there, but also make it sound as crazy as you want by just moving through very unrelated waveforms. 00:23:46,348 --> 00:24:03,888 [Anthony Marinelli] Well, because everything's a wavetable. Like sampling is a wavetable. FM synthesis can be described in series of wavetables. Same thing as sampling. So essentially, it's almost like it doesn't even have an identity because it can do almost anything, right? Because some of what I just heard there sounded like FM modulation. 00:24:04,560 --> 00:24:20,840 [Bob Coover] And you hit the nail on the head. So I didn't say that. That wavetable that I was just playing at the end there is also getting modulated by a sine wave. So as it would spread out, you're getting a little bit of FM. But that's the beauty though, of a wavetable, is you can make the jumps as close as you want and make that movement very subtle, or you can make the jump super wild and make it sound just, you know, like it's from another planet. 00:24:41,320 --> 00:24:48,600 [Julian Rodgers] We're talking about historical landmark instruments, but it's gone more mainstream since. I mean, I had a question that I wasn't gonna ask, but actually I think it's probably quite a good time. I'll start because you will definitely be able to beat mine, but I first became aware of wavetable synthesis, I think with, they were called Wavestation, a band I was very fond of in the sort of, that would have been the early 90s, I suppose. Used to use it for evolving kind of complex sounds. But hands-on, I don't think I came across a wavetable synthesizer until Propellerheads introduced Thor into Reason, and that was extremely popular, and then kind of especially with, I was teaching around the time dubstep was getting huge and everyone, all the students were using NI Massive and things like that. 00:25:34,660 --> 00:25:35,200 [Bob Coover] I was gonna say Massive. Yeah, Massive was another one. And then the one that a lot of people use now, and I think it's, you know, become ubiquitous at least in the software land, is Serum, Xfer Serum. So Serum, you can actually see the wavetables moving in front of you, and it makes it make more sense to people. We actually allow you to import Serum into the 3rd Wave, but those are the yeah, as you said, the Reason, but Massive. Massive was probably the first one that really started to get used a lot, at least that I know of, and then Serum. And there's other ones since then as well, but Serum still seems to be the elephant in the room as far as software wavetable synthesis. 00:26:23,520 --> 00:26:43,820 [Julian Rodgers] And of course it's these things are probably a really good illustration of the fact that this isn't an either / or, you know, binary choice between are you doing wavetable synthesis or something else. As Anthony was saying, when you were using those early hardware synthesizers, you were combining them with a filter that you were lacking. 00:26:43,820 --> 00:26:44,040 [Anthony Marinelli] Yes. 00:26:44,040 --> 00:26:45,500 [Julian Rodgers] Yeah. Combining techniques. 00:26:46,580 --> 00:27:08,060 [Julian Rodgers] So I mean, in terms of gaining a foothold in public consciousness, I mean Anthony, any thoughts on kind of like, have you seen this, are we experiencing a renaissance of it? Did it not go away anyway and I'm imagining that? Are we, you know, where are we with wavetable and the synthesis landscape? Any thoughts on that? 00:27:08,060 --> 00:27:27,600 [Anthony Marinelli] I

think the creators work with developers, and developers work with creatives, with creative people at its best case. So I think, like Bob just mentioned Serum, I think that's bringing it to the forefront because there's a lot of Serum users. And then, it was given for free, the next version of Serum, to the users. So they are getting wavetable ability, like Bob's saying, like you could look at it better, and plus FM. So now the creators are gonna be trying it because it's there. You know, there's a lot of people that play by ear, compose, you know, great composers. Same thing with sound synthesis. People just find things, and so I think that's how it happens. It just happens organically, not like I'm gonna use FM synthesis or wavetable synthesizers, or I'm, you know, only analogue. I don't think it really works that way. It's like well, what instrument do you have, what's in front of you, what's inspiring you? And then the music happens. And then if you're not inspired, you just move on, basically, because it's human nature, to where the inspiration is. You know, like, where's the fire because I just wanna get warm. And I think that's how I see it working. 00:28:29,620 --> 00:28:50,960 [Bob Coover] And we've got to mention real quick, because as Anthony said, that is the software. Then you've now got hardware again, right? So you've got our 3rd Wave. You've also got ASM, Hydrasynth. You've got, we cannot go through this without mentioning Waldorf. Waldorf has got the Microwave, and now they've got the Microwave M, which does a very good job of just the PPG part of this. But wavetables exist, you know, again, Waldorf, ASM, us, Groove Synthesis, to some degree Novation Peak. That's an abbreviated version. 00:29:15,640 --> 00:30:05,820 [Anthony Marinelli] And you mentioned UDO. UDO you mentioned earlier. But this is where the developers, you know, the instrument makers, let's call them that, just like always in music, Stradivarius would inspire you, or great piano makers, and the musicians would say Oh, let me check out this instrument. And then it's like they don't know necessarily what it even is, but it's there. And if it makes the inspiring kind of sounds and I like what Bob's saying, that, you know, he's really focusing on re-integrating hardware into the thing and making it really affordable, because it does so much when you think about it. Like, maybe seems like a high price tag, but not for what you're getting, for a level of musical inspiration you're getting. And I think that's a big part of the experience too, being a human is just being able to turn the knobs and feel it and touch it. 00:30:05,820 --> 00:30:10,056 [Julian Rodgers] And there's, there's that accessibility, I suppose, is the word I'm coming back to with being able to control a sound by grabbing something and it making a satisfying difference to the sound, is what we're all after. 00:30:25,646 --> 00:30:28,556 [Anthony Marinelli] How do you like your Hammond, Julian?

00:30:28,556 --> 00:30:29,996 [Julian Rodgers] Oh well, I mean, absolutely. I mean, I didn't know how much I would use physical drawbars until I got some. I mean, that's just how it is. And learning to play an organ as opposed to just playing an organ patch. 00:30:42,046 --> 00:30:44,876 [Anthony Marinelli] Well, by the way, that's additive synthesis. 00:30:44,876 --> 00:30:45,996 [Julian Rodgers] Oh, absolutely. 00:30:45,996 --> 00:30:46,676 [Bob Coover] Exactly. 00:30:46,676 --> 00:30:47,996 [Julian Rodgers] Not much of it, but yeah. 00:30:47,996 --> 00:30:57,176 [Bob Coover] And if you could make it into discrete little slices, that's wavetable synthesis. So to bring this whole thing back around, you got a Hammond organ, you pull the drawbars to add or remove harmonics, and if you recorded it up and put it in a table, you got wavetable synthesis. 00:31:04,616 --> 00:31:30,756 [Julian Rodgers] Absolutely, absolutely. Well I suppose, I mean there are, you know, unkind memes on the internet about people just kind of, you know, tweaking the cutoff of a filter rather than programming a synthesizer and you know, people being slightly unkind about that. But it's true that there are an awful lot of controls on a synthesizer that, to the novice, don't really achieve a great deal. But that cutoff knob always does something. I think that's something of the appeal to get people to break out of, you know, preset cruising, et cetera. If you've got a wavetable, if you turn that control, it does something. 00:31:41,836 --> 00:31:58,656 [Bob Coover] Yeah, we've got a wave surfer knob on there that walks through the wavetable. Or even if you don't walk through the wavetable, if all you do is turn the wave table selector and change the wavetable, oh my God, do you get a large amount of movement. 00:31:58,656 --> 00:32:02,416 [Julian Rodgers] Yeah and then you find the sweet spot and you can interact with it. And this changes from being choosing something that someone else has made to making something. And yeah, absolutely. 00:32:08,046 --> 00:32:08,946 [Anthony Marinelli] You're so right about that, though. Like, having just the magic knob doesn't come around very often. Like a filter knob, right? That's like if you go to any trade show on synthesizers, you know, what's the over-under on how many, like, filter knob turns are gonna happen within the 1st few seconds? That's the go-to for anybody. 00:32:26,816 --> 00:32:29,856 [Julian Rodgers] That's the one I'd reach for, you know? 00:32:29,856 --> 00:32:40,476 [Bob Coover] And hopefully now with some of these other ones out here, the 3rd Wave in-particular, now you've got the wave surfer knob. So that's your equivalent of changing... 00:32:40,476 --> 00:32:49,136 [Anthony Marinelli] Yeah, but I'm glad you mentioned it, Julian, because like, you're approaching that. You need that though, in music. You need that where you can go and get. 00:32:49,136 --> 00:33:22,076 [Bob Coover] Yeah, one knob. And Dave Smith, I would be remiss if I didn't do some kind of podcast and didn't mention my hero who I used to work for, Dave. He would always say, "If I can't grab a knob and create something interesting, there's no point in having the knob on the synthesizer". So the filter cutoff, as you said, the wave surfer to change the wave position, all of that is, as Anthony was also saying, creative, you know? You hear this and you go ah, that's grabbing me, that's what I wanna hear. 00:33:22,076 --> 00:33:25,036 [Anthony Marinelli] Like the reverb knob. 00:33:25,036 --> 00:33:26,416 [Bob Coover] Yes, yes. 00:33:26,416 --> 00:34:28,276 [Julian Rodgers] Guilty. We've all done that. I mean, many, many years ago, I mean, one of the first people I worked with had a extensive vintage synth collection, and he had a VCS3 amongst other things. And I was terribly impressed with this thing. However, I knew nothing about what I was dealing with and it wasn't a very successful interaction. I had a go on a VCS3 and achieved virtually nothing because, you know, that's not got the magic knob. Well, there's probably a spring reverb on it. But yes, anyway. So we should maybe name check. I mean Bob, you mentioned Policy of Truth which, maybe we should at least kind of reference a couple of kind of recommended listenings, should we say, for use of wavetables in well-known recordings. Anthony, anything that you'd kind of point a listener to in the wavetable canon? 00:34:28,276 --> 00:34:35,856 [Anthony Marinelli] I mean, I could talk about stuff that I did. There's the movie War Games, it's being re-released on Blu-ray. 00:34:35,856 --> 00:34:36,966 [Julian Rodgers] Good one. Yeah. 00:34:36,966 --> 00:34:50,656 [Anthony Marinelli] And so I just did the interview for the DVD, and it's gonna be a great version of it. And I did a ton of that throughout the movie, and it was like, early days. It was in 1983-ish or something. 00:34:50,656 --> 00:34:52,116 [Julian Rodgers] Was that Synclavier? I don't know, I probably should. 00:34:52,116 --> 00:34:52,326 [Anthony Marinelli] Yeah, it was all Synclavier. But it was like wavetables with FM and both filtered and unfiltered. And you know, I used the switches because Cameron put these two switches on the back. They're just like medium and high and if you just left them open, you got all the hash. And Bob knows what I'm talking about, because it was 8-bit back then. And that's sort of like the reminiscent, when we look back on the beautiful sound of, you know, wavetables and early FM. But it's loaded with that stuff because I left the switches wide open most of the time. 00:35:29,876 --> 00:35:40,006 [Julian Rodgers] And it's perfect for it because it pins it in time in a way that if you'd had a, you know, big orchestral thing going on, it wouldn't have stuck in the mind quite the same way. 00:35:40,006 --> 00:35:43,996 [Anthony Marinelli] Well, but there was a big orchestral thing that went on over the top of

that. 00:35:43,996 --> 00:35:46,196 [Julian Rodgers] Oh, yeah, yeah. But what I'm saying is that it defines the time it's referencing in a way that wouldn't have done without, if you see what I mean. 00:35:54,056 --> 00:35:54,776 [Anthony Marinelli] Right. 00:35:55,516 --> 00:36:04,396 [Bob Coover] And you can't mention wavetables of course, without mentioning Thomas Dolby. So Windpower from his first record or Mulu and the Rainforest from his second record, at least over here, those both have got just bare wavetables that you will hear moving throughout them and in some cases, as Anthony said, yeah Windpower was so early, he was using, I think, 360, if I'm remembering right. It's an even earlier version of Wolfgang Palm's stuff, and it didn't move correctly, and it kind of made a loop that made a rhythm out of the wavetable itself. So you, you've got those. You've got, as Anthony said, War Games. You've got Depeche Mode, and then there's so many examples in modern music that are using wavetables, as we said, even in the software with Serum or something like that. So yeah, there's... listen and hear weird movement in the sound, and you are likely going to be listening to a wavetable or a variant of a wavetable, especially if it's something that you can't listen to and go that's a flute, or that's a violin, or that's something. If it's moving very oddly, chances are you might have a wavetable in there, not a subtractive analogue synthesizer. 00:37:23,626 --> 00:37:28,036 [Anthony Marinelli] Well, an iconic single sound is the opening of Beat It. 00:37:29,276 --> 00:37:30,476 [Bob Coover] Yes, there you go. 00:37:30,476 --> 00:37:30,756 [Anthony Marinelli] That's basically a wavetable... 00:37:33,166 --> 00:37:33,166 [Bob Coover] FM wavetable 00:37:33,166 --> 00:37:34,966 [Anthony Marinelli] ... with a little bit of FM. 00:37:34,966 --> 00:37:34,996 [Bob Coover] Yeah. 00:37:34,996 --> 00:37:37,146 [Anthony Marinelli] But it's times 4. It's 4 parts happening, so it's like multi-timbral wavetables with some FM. But yeah, I mean like Bob said, you hear that and you go what's that? And that was the purpose of it. But good musicians can really isolate it, it's not like... because when you pile it on, for me, that type of synthesis doesn't... it's like you can add a lot of guitars, you can overdub them forever and they keep sounding great. And analogue synthesis, same thing. Wavetable synthesis in the early version, not the later version where you can create clean, organic sounds, like where Bob is going, but that early stuff is so complex that if you start piling it on top of itself, you end up with noise because there's so many harmonics, right? You end up with a lot of noise. 00:38:25,596 --> 00:38:29,176 [Julian Rodgers] Yeah and if you end up, yeah, you'll inevitably get more and more noise, like the more... 00:38:29,176 --> 00:38:30,526 [Anthony Marinelli] Yeah, so early usage I find it to be more pointillistic is the preferred,

you know, musical shape of it. 00:38:38,556 --> 00:38:43,496 [Julian Rodgers] So to make something very distinctive, it's a great way to get there. I mean Thriller's a great example, just because you know, some records have such recognisable like, motif sounds, if you like, that announce what they are and yeah, if you hear that, you know exactly what you're about to hear. 00:39:02,245 --> 00:39:07,176 [Anthony Marinelli] But I mean the pipe organs at the end of Thriller, because I was talking about the gong on Beat It, but the pipe organs at Thriller, all at the end, that's all wavetable. 00:39:12,936 --> 00:39:34,436 [Julian Rodgers] I wanted to at least brush on possible future development. We've looked at the historical timeline, we've looked at where we are now and sort of recent events. It's got cleaner, although I suppose things that have been less clean in the past, we can access those again. And access we've already identified as being kind of like, quite important. Software obviously brings newer ways of visualising what's happening and I mean Bob, you mentioned that as being a positive thing, for kind of selling the concept to new users possibly. But anything that either of you kind of think is something that either needs addressing or has been addressed or could potentially be developed or, you know, the current hardware boom that we've seen for the last few years as well and how that's kind of working out. Any thoughts on the future of wavetable synthesis? 00:40:06,156 --> 00:41:00,306 [Bob Coover] Right, right. So the two things at least on my side that come to mind are more memory. One of the reasons that wavetables got created in the first place, as we mentioned already, was lack of memory and that's one of the reasons that a PPG sounds so special is because of lack of memory for storage, and you get the imaging as I call it, the beautiful garbage down at the bottom. But now that you've got memory out there, you can not only make the waveforms longer, but you can add more waveforms. So we keep saying 64 waveforms in a wavetable. You can expand this out as far as you want and get it very close to a sample. And I'm not gonna give away something that might come out, but yes. And then the other thing is non-linear movement. That's a nerdy term but meaning that you're gonna not walk through the wavetable sequentially or walk backwards sequentially, but actually I think there's at least one software company out there, and it's escaping my memory right now, that allows you to actually pick where you wanna walk and jump and do non-linear movements throughout that, so that you get even more variety out of the same wavetable. So those are two things that I think of immediately, but I'm not gonna completely give away the farm on the rest of this. 00:41:35,096 --> 00:41:40,916 [Julian Rodgers] No, no, fair enough, fair enough. I mean, Anthony, any thoughts? Any kind of, any what ifs? Any I wish we coulds? 00:41:40,916 --> 00:42:18,116 [Anthony Marinelli] Yeah, thank-you. No, I mean for me, it's just all the stuff Bob said, but with expression. I think that's where a lot of this is heading, like putting the knobs back on and like, but better controllers. And then the routing is a lot about being able to have the controllers control more things. Like when you say, Bob, like, it could jump. Yeah, well, it's great. It jumps when it wants to, but now it's getting to the point as you're developing the 3rd Wave where it can jump when I want it to as the musician, you know? And I think that's something where we'll be able to play a whole different level of music on synthesizers. And the more and more I go to these shows, you know, like Superbooth or NAMM, I see more expressive controllers hitting the thing and on the panel. So I'm hopeful for that because it was kind of stagnant for a long time. 00:42:34,096 --> 00:42:36,536 [Bob Coover] And MPE is one of those things that's making it possible. You know, you've got 3 axes of control instead of just hitting a note. So yeah, that absolutely what Anthony just said- 00:42:47,816 --> 00:42:47,886 [Anthony Marinelli] But I think 4- 00:42:47,886 --> 00:42:50,116 [Bob Coover] ... more expressive control. 00:42:50,116 --> 00:42:58,056 [Anthony Marinelli] Oh my God. I don't know if there's an end to it but I'll show you like, this MPE. It converts MPE to control voltage and there's 4. 00:42:58,056 --> 00:42:59,996 [Bob Coover] Oh, that's great. 00:42:59,996 --> 00:43:08,096 [Anthony Marinelli] There's W, X, Y and Z. So I have 4 channels, so I take anything MPE, like the 3rd Wave, and I can control my big Moog modular. So that's insane expression. And I could mix and match, like the beauty of the 3rd Wave, but with the Moog modular or any, you know, old school thing that's beautiful. So yeah, I mean, and maybe, Bob, you can put more, right? Is it only, is it locked to 4? 00:43:29,036 --> 00:43:54,516 [Bob Coover] Right now the spec is the 4 but yeah, you could. The sky's the limit as far as how many axes you wanna do. And yeah that expression, and it doesn't have to be, you know, a keyboard instrument. When you think of what Roger Linn's doing or what some of these other, you know, Roli, Osmose, they're giving you the 3 axes, but... 00:43:54,516 --> 00:43:55,516 [Anthony Marinelli] And Continuum. 00:43:55,516 --> 00:43:58,136 [Bob Coover] Gosh, I guarantee you that there's gonna be more. 00:43:58,136 --> 00:44:00,786 [Anthony Marinelli] And Continuum, that's groundbreaking and it's been around for 20 years. 00:44:02,376 --> 00:44:24,936 [Bob Coover] Yep, yep. And you know, people are just, we forced ourself with the 3rd Wave to put MPE in there so that it could be supported by, or that it would support using those types of controllers. And that, along with those other items that I mentioned and who knows what else somebody's thinking about out there that we don't even know, that's the future of wavetable synthesis. 00:44:24,936 --> 00:44:25,395 [Anthony Marinelli] Yeah. 00:44:25,396 --> 00:44:48,615 [Julian Rodgers] Amazing. And with these new control opportunities comes a whole new regimen if you like, of players to learn to, you know, exploit them fully and you know, the amount of people who ignore the pitch bend, and it's kind of like yeah, there's pitch bend and some, so you know, that wheel goes round again. It's been a really interesting conversation. I think we're to time, so thank you so much. 00:44:48,616 --> 00:44:49,796 [Anthony Marinelli] Thank you, Julian. 00:44:49,796 --> 00:44:50,796 [Bob Coover] Thank-you very much. 00:44:50,796 --> 00:44:55,736 [Julian Rodgers] Thank-you, Bob. Thank-you, Anthony. It's been great. 00:44:58,076 --> 00:45:29,466 [Julian Rodgers] Thank-you for listening, and be sure to check out the show notes page for this episode, where you'll find further information, relevant links, and details of our other episodes. And before you go, do visit soundonsound.com/podcasts, where you can explore what's playing across the Sound on Sound podcast channels.