 Our next speaker tonight, Matt Westcott, has been writing music and demos on the ZX Spectrum for over 20 years. He's had some amazing career with that. He's won competitions at assembly, the biggest demo party there is, and tonight he's going to take all of that experience and try and condense it to you in an hour. He's going to teach you how to ride a chip tune from scratch and best of all you're going to get to choose the chip tune. So have a think what you'd like to see turned into all its 8-bit glory and please give a warm welcome to Matt Westcott also known as Gasman. Thanks very much. So given that you've come to this talk, I hope you have at least a rough idea of what chip tunes are at least at the level of it's that sort of blippy-bloppy video game music. It's like many retro things. It's kind of something that's been borrowed as a kind of aesthetic choice. So if you're making music on something like Ableton Live or Logic, then you can easily find these chip tune plugins and to make the music in this style. But it's only really when you get to doing it on the original machines where chip tunes originated from like the Spectrum, the Commodore 64, the NAS that you can really get a sense of how the whole style of it has been born out of the technical constraints and it kind of emerged as ways of getting around those constraints. So hopefully that's what I'm going to be able to communicate to you today. So as you mentioned, I'll shortly be taking a request for a pop song to cover as Spectrum Chip Tune, which I'll attempt to cover in an hour and explain the process as I go along. But while you're thinking of requests, I'll give a bit of background to the parameters that we're working in and where the chip in chip tunes comes from. So the original home computers, when they were being designed and they wanted to add audio support, they were very much in the mindset of doing the simplest thing that could possibly work, which meant wiring up, going from one of the CPU's output ports, wiring that straight up to a speaker so that if you're output to one, it would send a voltage across the speaker. If you're output to zero, it would turn that off. And by alternating between those two states at a given rate, then you would create an audio wave. And the advantage of this method is because it's all been done in software, is that there is basically no limits audio quality aside to the sort of sounds that you could produce. You could do speech synthesis of multi-channel audio. If you can program it, then you can do it. But the real disadvantage of that is that you can't really do anything else at the same time. Because it's, well, if you're going to be turning this speaker on and off at a precise rate, you can't just set the CPU to say, okay, set a timer to do this thing while I go off and do my own thing. You've got to be, even if you're just waiting, you need to count cycles to make sure everything is happening at precisely the right time. So it doesn't mean that you're very limited and what else you can do at the same time. I'll play an example of a load up Manic Minor, which is famous as one of the first ever games with background music. And you can kind of hear that it's got this very choppy sort of quality. And that's because it is literally having to divide its time, like a fraction of a second at a time, playing the music and then stopping to actually do the next bit of game and then playing the next bit of music. And people just, they wanted to do better than this. So the next generation of home computers, including the Spectrum 128K and the Commodore 64, they started shipping with built in sound chip. That's all this of unlike the CPU based approach that we've just shown. This is a fixed capabilities. So usually it would be something like you have three or four tone generators and they would just be able to play a square wave or a sawtooth wave and nothing else. You'd be able to sort of set the speed of these generators and the volume, but you're kind of stuck with the sounds that they provided. But the big upside, of course, is that these things can run under their own steam. Once you've actually told it to start playing a note at a particular frequency, it will just carry on doing that until you tell it to do something else. So in the case of the Spectrum, that sound chip, it was the AY38912, which is also used in a few of the machines like the Amstrad CBC. For the Commodore 64, there's the very famous SID chip. So that's basically the constraints that we're working within. So you don't have the flexibility that you have for doing things in software, but there's not enough flexibility to totally reinvent the sound it makes, but there's enough to be creative enough with it. So having said that, I think it's time to take requests. In fairness to people who aren't feeling like shouting stuff out, if you want to come to the microphone, if you've got an idea or shout out if you want. I see some theme from the arches. That's a really hardcore audience here. Any other suggestions? I kind of like to pick and choose what the time is. I think I'm going to stick with the theme to the arches. I've already done that one in my illustrious 20 year career. I do that when it's minimal. I like the theme to the arches. It has to be a tune I know very well. I've not done something of that sort before. Give it a bit of a drum and bass remix or something. Let's go for that. Here we go. Let's fire up the tracker. Let's give a quick burst of this. This is my first ever attempt to do a chip tune under time constraints. This was 15 minutes of American pie. That's as far as it goes. I didn't manage the whole nine minute track unsurprisingly. This is sound tracker. This is the software I use. It's running on the real spectrum or in this case an emulator. This is an original bit of software created back in 1991. There are better trackers around, such as Vortex tracker, which runs on Windows, also on Wine for Linux and Mac people. The reason I'm using this isn't out of some sort of hipster authenticity thing. It's just that this is what I'm used to. There's nothing worse than when you're in your creative flow, having to stop and remember how the software works. This is just what I'm used to. Let's clear this song. As you might have just seen there, this is the bit in the middle. This is where all the actual real music writing happens. It's divided into these three columns, which they correspond to the three same generators that you get on the spectrum. What you can do is, and the vertical axis is time. If I start entering some notes, these are just using the standard musical note names. Let's spread them out a bit here. As I saw there, if I start this playing, then you can see that the time is scrolling past and the white row is the one that's showing the note that's playing at the moment. The reason you didn't hear anything playing right there is that I haven't actually set up any sounds yet. This is where we need to go for the sample editor. Samples is a bit of a misnomer here, because this kind of originated from the trackers on the Amiga, where this whole concept started off. Samples in music usually means a clip of recorded audio that you're triggering at precise times, perhaps slowing it down, speeding it up to vary the pitch. In the case of the Amiga, that was the building blocks that you're using to build up your song. Since we're restricted to the sounds that the sound chip can generate, the meaning of samples is kind of a bit hacks around a bit in our case. What we have, I'll edit sample number one here. This is, rather than it being sort of a real audio waveform, it's a graph of how the volume will vary over time. As I said, we're pushing data to the sound chip every 50th of a second, so each of these bars is 50th of a second slot. Let's start creating a waveform that starts loud and dies away. The top half is the volume over time. Since the arches theme, it's very staccato. We want to note that kind of dies away quickly. I think it's going to be this sort of, fortunately, being a spectrum program, it's all kind of very fiddly to actually do the editing. I think there's a bit more volume there. We've got the feeding away fairly quickly sound. You'll notice it is sounding a bit sort of noisy there. That's because the AY chip can also do noise, which is what this second half is. It's control link. I'll set these to one to mute them out. It's also very annoying how you get this click that's actually louder than the music you're playing. I might get around to explaining why it's one to turn this off. It's because it's talking to the chip at a fairly low level and this happens to be how it works at that level. We've masked out the noise and now we've got this pure tone. Again, it's because it can only produce square waves. This is a square wave sound that's dying away fairly quickly. I'll skip past the screen. This is actually showing how the tone will vary over time. I think we want this to be kind of a flat note at the moment. Now we can get on with editing, putting in our notes. Let's clear these out. We've created that as sample one. Let's see what note do we want to start on. I think that works. One important decision at this point is what the tempo is going to be, how quickly we scroll past here because you can only put the notes on the, you can't put them between rows. You probably don't really have much choice here. It's going to be like... I did a second bit while I'm here. That's the one. Incidentally, it's the bottom row of keys here that CVBNM is like the musical keyboard. It's all very low-fi, no midi or anything. Let's hear how this sounds. It's probably going to be way too fast at the moment. Not too bad. Good start. I think I will slow that down slightly. Delay six. Since everything is in units of a 50th of a second, that will be... Advanced I think it's every six 50th of a second. That's one X. I think there's a couple of notes. What's booking me slightly is there's some longer notes that could do with another sample. In trackers, you don't really have a concept of setting the length of a note. If you want a longer note, that just means a sample that takes a long to die away. Let's make it so it sticks at that level here. Again, this is very fiddly and clicky. Again, we want to mask out all of the noise. That's more like it. So far, we've just specified sample one at the beginning there, and then the zeroes means it will carry on playing sample one. That's the one who wants to be a longer note. I've kind of forgotten to actually set it back to one again for the shorter notes. Let's do that again. One, two, one, two, one. Oh, I need to do it for the second half as well. Let's hear that again. You can hear the difference between those two note lengths there. Since unlike a lot of the music that the youth is listening to today, this is in three, four times. Well, in compound time, which means that by default, you have 64 rows in this. Since we all probably want to change that to a 48, because it's all in threes, or 47 in fact, because it's zero based, just to confuse you. Now that's about to repeat again. As a space saving measure, because memory is always very tight on the spectrum, the song you're making is divided into what's known as patterns, which kind of musical phrases. Since a lot of music is quite repetitive, if you've got a standard song structure where it's verse, chorus, verse, chorus, you don't want to have multiple copies of that data. You can actually define your song as a sequence of these patterns possibly repeating. In the case of verse, chorus, verse, chorus, it might be one, two, three, four, one, two, three, four, and then three, four if there's another chorus, and that way you're not storing multiple copies of the same data. At the moment we're just only working with one pattern at the moment. We'll see if we get time to do a bit more of the song, but let's make this be a bit more adventurous about the samples here. So this long note here, we can add a bit of vibrato. Let's imagine we're playing the arches theme on an overdriven electric guitar. It might be a bit too much to hope for. We can have a bit of, when it's doing the long note, so we can do a sort of a bit of vibrato effect on there, so that adds a bit more life to it. As I say, these columns on the left here, it's actually just one column. Again, it's sort of defining what happens over time, and every 50th of a second at the moment, we're not adjusting the pitch at all, it's all just one flap tone. Let's say, let's shift it up a bit here. These are in units of, it's complicated, because again, because this is all addressing the chip on a very low level, which is kind of incidentally why we're using this sort of interface and not kind of conventional like music sort of, yeah, musical notation. So this is kind of adjustment to the period of wave in clock cycles, which also means as you're going to the higher frequencies, these notes have, these little adjustments have a bigger effect, so it's all a bit of a blackout to adjust it all, so hopefully you'll, so hopefully you might be able to hear if I stop it at this point. It's very subtle, but it's kind of, it's, I'll move a bit more to, you might, you might just be able to hear it's, it's sort of rampings of up in tones at the end slightly, that's a bit more, a bit, a bit. So let's say that's, so it's going up and down again and then, it's all very subtle, we don't want it to actually change the note we're playing in any fundamental way, and let's do that in the negative direction here. So, so you can hear this sort of sound there. So the other thing that is kind of what we're going for there, so we can see if we play that back we can hopefully hear the effect on the actual song here. So yeah, it's a bit, hopefully you can hear the effect there. Let's get a bit of a baseline sort of going here. So let's start entering stuff into the second, this is a second channel of audio. So, so down to sort of octave two, I think I should do it. Let's, I'll use sample one for this just because we don't really want the vibrato thing going on here. So, let's see how this is, it probably goes down, down to, down to A, down to A, down to A. Okay, so let's see what that's giving us. Okay, that's kind of working. So, let's onto drums which I'm sure the original theme didn't have, but let's go in wild here. So, this is where we start sort of getting the noise side of things into play. So, although we still need to sort of define something here, because it's still using the same volume control. So, let's see, we want to, let's get, so let's do a snare drum here. So, we want that to be, again to fading away fairly quickly over time. Let's do something like this. And so, again, you've got that sort of, that's a quite noisy thing. We want a bit of tonal content in there. So, I'll leave that, I could mask this out completely and then it will just be noise, but I think we need, we need a bit of tone as well. So, the second half is actually controlling the, again, this is where it kind of gets easy and low level, the pitch of the noise. I know white noise doesn't really have a pitch, but because it's actually done by adding sort of random ones and zeros and this is controlling, yeah, how quickly it changes from one to zero. It kind of changes the kind of, the flavor of the noise, I suppose. So, if we, so here we can hear that, that's kind of a very sort of, quite a high noise and then if I take that down, you can hear that's kind of going a bit deeper as you go down until the right of the top is that. So, rather than So, we've got some extra white noise coming from outside by the sound of things. Okay, so, as a sort of snare drum sound, you want sort of that sort of slowly fall, so it goes, that's it. So, let's go start. Again, this is sort of backwards and confusing, just, but again, it matches the way that the chip works. So, let's, so let's get that one to sort of go down to the lowest frequency there. So, so can kind of hear a bit there, maybe we'll make it reach the, the lowest bits faster. So, it's kind of in a way, so that's kind of what we're aiming for, but I think what we also need to do here is make sure that the tone content also sort of goes right down to zero, well, not just zero, it plummets down. For some reason, you get the options of sort of adjusting this by either one or by 128 and nothing in between. So, let's do this a few multiples of 128. So, the idea here is that we're kind of making it stop sounding like a tone, because you could still kind of identify that as a specific musical note, and that we're taking this down, so it's sort of going sort of down to the ground really quickly, so that you can't really identify it as a particular note. So, I think that might be enough. Okay, it's kind of turning a bit bass, bass note-y there. Let's, no, wrong way. Let's move that up a bit. So, I've kind of got a bit of a snare, sort of, it probably could have done with sort of not fading out in volume quite as fast, but it'll do for now. So, let's also do a bass drum. In this case, we don't really want it to have that sort of sort of white noise quality, so we can get this whole percussive thing just by just using this second trick of sort of ramping the tone down really quickly. In this case, a bass drum that sort of does, that's kind of a very short stabby sort of noise, so the volume should fall away quickly there, and as I say, we don't really want any noise there. So, then we've got this very blippy sound, and then, again, we'll take this down to, yeah, runkeys, make this pitch slider way down when you go a bit faster than that. Okay, that's kind of more what I'm looking for there. So, this is all a bit of a sort of trial and error sort of thing. It's not like we're physically modeling these instruments or anything, so, and we're only going to come so close to it actually sounding like a real drum, but I think that's probably good enough to work with now. So, now I've got to decide what a drum beat for the arches theme is going to sound like. So, okay, I'll try something like that. So, it's, four is my bass drum, so it's, boom, two, boom bass there. Let's see, how does that go? Okay, and because I can't be bothered to figure that out for the second half, let's do a spot of cut and paste, so it's still very low level again. So, that was, so we're still working in pattern one, copying the paste from one, and that was zero to 23, paste to 24 to the end. So, it's like copying from the first half of this pattern to the second half, and that was channel C. So, yeah, it takes a lot of experience to kind of remember what numbers to put in there. So, okay, so I think we're definitely getting somewhere here. So, what next? Probably wants us a bit of accompaniment. So, it's a bit, a few chords to add to that, except this is where we kind of run into a bit of a problem, because we've, now that we've got the melody, the bass, and the, and the drums, we've run out of channels to actually put notes into. And so, it is kind of the, the sort of perennial problem we run it and run into with, sort of chip music, if, if, if it's the, the real thought where you have these constraints. So, we'll need to find ways to kind of insert our, our chords around these. And obviously, chords consist of multiple notes being played at once, which, I know we're going to have enough trouble sort of inserting a single note in, in, around what we've already got. So, this is where we actually get to what's probably the, the most distinctive thing about chip tunes, which is arpeggios. So, if you've got sort of several notes to play at once, like a C chord, then the, then the trick is to, to cycle between, very fast, except faster than that. So, that's, it's, well, it doesn't really sound like they're all playing at once, but there's enough of an effect to pass, and this is kind of the, the, the sound that you would always associate with kind of Mario soundtracks or whatever. So, this is under sort of ornaments here. So, ornaments again, this is another way of the varying pitch over time, but this time this is measured in actual semitones. So, we're working out the actual musical level, rather than the, rather than this abstract frequencies thing. So, at the moment we're just, that's all zeros. So, so I probably want sort of a C major chord. So, I think that's, yeah. Now, I, I'm, I'm doing, so it's sticking at each note for two fiftieth of a second. I've found a three experience that, if you do it, this kind of leaves it with a nice melodic one. If you cycle through the notes too fast, it kind of becomes a bit harsh. So, let's go for 047, and let's repeat this a couple of times. Usually, I'd sort of go through this whole, the whole list of notes, but given that we're kind of a bit pushed for time, that's probably good enough. So, there, you've got this son of major chord there, and it's, and you can come here while it is sort of cycling through the notes there. Yeah, it's, it's sort of a decent, a decent enough recreation of actually playing that as a chord. So, let's get another one. So, that's a du du du, let's, we probably want to do du du du. There. Okay, so we'll start with those. You might find that we need more of these, as in pretty much everything in the soundtrack, there's some of a quite a sort of severe limit to the number of samples and ornaments you can have. We've got sort of 15 or 16 to play with. I think we're not going to have time to fill that up, but once you're actually doing a sort of full length song, you can find yourself running into these, and you have to make creative compromises. But so, let's, let's see, this is going to be a tricky one. Is it where, where to insert these? So, I can probably, let's see, it's, this is kind of a different problem solved every time you write a song. I think we can probably get away with the bass notes being sort of shortened, so it's, it'll be du du du du du. It's, I guess it's kind of a ragtime sort of thing or something. So, okay, so we've got so a bass note there, and now we want this to be, let's see, it's probably going to be this, that octave. So, let's, so that was, oh, we're probably going to want another sample at some point because this is the same volume as the main melody, which is, is no good. Now I have to say, so F is the, that's the code to say, let's turn this into, let's use one of these ornaments, and in this case we're using ornament number one, and if I sort of leave it, so for running like that, then, so we hear that's kind of added the ornament to the bass line as well, which is sort of pretty horrible, so we're going to have to explicitly say turn that off when we're playing a bass note. So let's see, that's probably the second one, and that's off. Okay, let's see if that's sounding vaguely possible now. Okay, so let's hear that again. Okay, maybe what I really need to do is just cut out some of these bass notes so it's, so that we've got a bit more time to actually hear that accompaniment. So let's sort of chop out every other one, and let's, okay, so, so well, I'm going through sort of removing a few of these, these bass notes, so to fit in some of the non-bass ones. Okay, and since that second half is not quite the same as the first one, I don't think I can do my copy and paste tricks, so I need to quickly go through and add those two, and that's, that needs to be played for. I'm going to remove that one. Now we need to sort of mine a note, mine a chord, so off to the ornament editor again. Let's see, let's just do one more repeat of those notes. Okay, so that was on number three, so let's hear where that goes again. I'm going to stop that too early. Yeah, so that's the minor note we want, so now I think it's, right, so I, okay. So yeah, that's actually sounding pretty good. I like that. I thought I would have to take the volume of those accompaniment staves down a bit, but I think that's working pretty well. It's, I think that's quite a distinctive bit of our tune now, so we can sort of bring that up in the mix as it were, and I think the fact that it is also fairly short stabs that then gets interrupted by the bass means that it's kind of not quite as prominent as the melody there. I think it's time for a bit of hardware abuse of the very mild sort. I'm not going to set fire to my laptop or anything, but so those bass notes, while we've kind of jazzed things up a bit, let's just focus on this channel here. They're still sort of pretty plain, so that's the we can do something a bit more sort of exciting here, and this is where, so, so this is where I think we have to go back a bit into the history of like where the AY chip originated from, because I get the impression that when they first designed it, they didn't really imagine that people would use it by pushing data to it every 50th of a second, like we've been doing all the way through this. It seems like they imagined that you would just send data to the chip on every time you wanted to actually play a new actual musical note, because obviously it can play a note under its own steam if you're not bothered about doing all fancy effects to it, so they probably imagined that's how people would use this chip, but then they evidently realised that, well, this is just going to result in tunes where all the music notes are just going to be a flat B sort of sound, so let's throw our users a bone here, and so they came up with the idea of envelopes, which is a bit of functionality on board the chip, which would allow you to vary the volume of that note over time, and so you'd be able to set the, rather than having it just a flat volume, have it say that it will fade away at a certain time, then maybe it will repeat. There's a few different options, which is a nice thought of them, but as it happens this is pretty much completely useless, because we've already been doing our sample editing to vary our volume over time, and doing this in software is a lot more flexible, we could, for example, do something that sort of goes down and then sort of ramps up slightly to get sort of one-one-one sort of sound, which you couldn't do on board facilities, so as I said this bit of this feature of the chip is pretty much useless except for one thing, which is that if, because you can sort of control how long it takes for the envelope to fade away, what if we sort of ramped up the speed of this, so it's actually fading, so the volume is fading away much, much faster than one-fiftyth of a second at that point, and then we set it to repeatedly do that. At that point we've actually moved out of the sort of volume domain, and it's actually become an audio waveform in its own right, so when I said at the beginning that the AY chip can only do square waves, that wasn't exactly true, it can also be tricked into doing sort of triangular and sawtooth waves, but the downside of this is, like so many things here, you have to sort of work with it at a very low level, because this is kind of, as I say, hardware abuse, there's kind of speeds that you're actually setting the envelopes to are sort of right at the edge of the tolerance of what it was designed for, which means that you don't really have much resolution to control exactly what pitch this happens at, so you can't reliably match it to an actual musical tone. I think the Atari VCS has a similar sort of issue for crazy people who write music on that, because that was not designed for music at all, so it's that they kind of have a kind of very, because they're kind of working with this very limited resolution that doesn't always match to musical notes, then they kind of have to compose their music out of this reduced scale, and it's even more hardcore than I'm prepared to go, but so luckily we can sort of, we can make the best of this though by using it for the like the baseline of where, kind of because the notes' lengths are a lot, the periods are longer, then you've got a bit more control, so let's start sort of dropping in some of these then, so these are, rather than using the F command which is for the ornaments, we can, the ones to use it's A and C, I think C is kind of, that's like the sawtooth which is a harsher thing, so let's see how that goes, and because this is actually happening on top of the existing square wave, you can actually get some interesting distortion effects, where if it's not exactly in tune then it will kind of go in an outer phase, so let's see how this goes then, so that's our our note at the moment, so let's try adding the C, so a moment so zero is that's the fastest, so let's slow that down to so here you can see as we're setting the speed of it to like 24, and oh yes by the way this is in hex just to throw another spanner in the works, so we can, so as I say we can't reliably tune this to actual musical notes, so we're required to do this in hex values and kind of tune it by hand, so we want to try and find a frequency that sort of fits well with that, so let's see, I don't know and that's quite well, let's see what the ones on the side are, okay let's go one octave down actually, I don't know what 2d multiplied by 2, it's probably 5 8, actually no I like the first one better, so yeah so what's it, so 2d we're going for there, so yeah let's so find all of our notes here are f's so that's a G, so that will be a slightly higher and faster one, so it will be so it'll be C, so like 2, so that sounds right, so 2d again, so down a bit more, yeah that sounds about right, and up again, so that's what that's to, so you can see I'm kind of having to sort of adjust this by hand to look sounds right because there's no guarantee that you will actually match this up to a real musical note, okay so let's see what our newly pimped up theme tune sounds like here, okay so that's sounding a bit better, yeah that's it, so now we've got a bit of time left, let's try and move on to the second half of the theme, and let's actually, I'm enjoying that bass line quite a bit, so let's try and make this one a bit more basic then, so let's so Melody first though so, so that's, so rather than that I'll just do the Melody first then add these samples, I was adding the samples so it's 2 for the along notes, how that's sounding, okay let's see so let's say a bit of time, let's copy and paste the drums from pattern one, so that was channel C and luckily we didn't have to stick any of our chords into that one, so that should be okay that's sounding good, so as I say we're going to make this a bit more sort of bass heavy this time, so let's see, so I think we went for like note and tones around the 20 mark here, let's take that down, so okay so let's go for sort of, play that for one note and then they have the accompaniment, let's do all the bass notes first, that's sounding right, and then 70, there we go, we had for F that was 58, 78 there, okay let's see if you can make that, actually let's stick that an octave down, ah I should have done that, nice fat note shouldn't have done that all the way through really, okay and let's stick in our accompaniments here so then let's another of those chords, then it's going to be, oh no that's not quite the one, it's so it's another one that we don't have, okay let's just stick that one for now, let's not bother repeating, okay yep there you go, okay so let's put both of those halves together, so as I say, so now we've moved on to a second pattern, so in position one we have pattern one and in position two we have pattern two, so we didn't actually get to do any repeating there, but okay so let's see what our finished tune sounds like, thanks to Matt Westcott, we have five minutes for questions, so if you've got a quick question and it should be quick please come up to the microphones, okay we're just saving it first because I think this is a good one, I'm going to keep it, yeah okay uh no hang on, there's a weird book there, question at the back, I don't have a question, I just wanted to say it's brilliant thank you very much for this big show, thank you come to the microphone please, I have a question for the audience actually, do the non-brits in the audience actually know what the arches is, good point, show fans, hands up if you actually know what the arches is, okay that's um, hands up if you know and you're not British, yes it is a very British thing, it's a soap opera on the radio, it's sets of in the countryside so most of the story lines are, actually I could be horribly misrepresenting it because I'm not a follower of it myself but it's all about sheep getting stolen and things and uh yeah it's like say very British, so microphone, can we have the fun mic up please, I'm not sure if this is question for you, how does this compare to for instance punch cards on dance organs, can you I suppose yeah I can kind of see the similarity of uh yeah it's um music, yeah it's vertically over time, I suppose in that case you could sort of think of that as kind of an extreme version of this, where we've got three channels, I suppose like a player piano would be like 80 channels except each one is just one tone but uh yeah it's an interesting thought, yeah there's definite parallels there, yeah. Do you happen to have a copy of your Derude Sandstorm cover that the audience could listen to? Oh yeah okay I think we've got time for this yeah okay, particularly if you've got the video to go with it, yes I do I believe we can do that yes yes um okay so it's um yeah good okay um it's gonna do top, sorry what's the song called? okay um yep this'll be the one, all right it's okay um actually need to be on okay it has lyrics, I'm afraid that's all the time we've got so huge thanks to Matt Westcott