So yeah, I guess we can start. Yeah, so... Alright, thank you everyone for coming. I'm very happy to be here in the Tangible Music Lab. So I'm Moon-Chi from Beijing, China. I make instruments, I play them, and I teach. So my main instrument works including the wind p, Wingly, and Upcoming Horizon that I designed with Trent from Whimsical Reps. All right, so my topic today is called the Design Methodology of General Purpose Musical Interfaces. So it will be about three things. The first is that introduce a bit about my new prototype which is called SkySys right here. And two, some of my thinking processes along the way when designing interface. And the third is like I like to share some ideas for general music interfaces with you. general music interfaces for you. So as we all know, the decoupling of interface and sound engine is the biggest difference between acoustic and electronic musical instruments, because you can do whatever interface and control the same sound engine or do vice versa. And there's a lot of freedom in it, which can be considered as an advantage and also a disadvantage, because it's ironically, so a lot of electronic music designers, such as us, all have a dream to build electronic instrument that has the interface and sound coupling as deep as an acoustic instrument, right? Because an acoustic instrument can do whatever to it and it will make different sound. Sometimes the differences will be nuances but they will be different. But you can't play the shell of a MIDI keyboard, right? But you can play a shell of a guitar, right? So at the same time, it's maintaining the versatility and all the possibilities that electronic technology has offered. So the very reason why I'm here, so why I get connected with Enrique Martin was because of the one particular work by the Tangible Music Lab, which was called Tangible Scores. You know that? It's like a piece of ink carving, right, with contact microphones inside. And you stretch it with your hand to make some acoustic sound based on the surface, based on the textures and the shape and it will be possessed by the program. So it's very much acoustic. Interface sound is very deeply coupled. You know that, right? It looks good. They're like paintings, right? That you can scratch. Yeah, maybe we're all cats. Trin and I have been working on the Horadon, which is a stand-alone instrument, a complex one, for about almost six years. We just gave a lot of our time into it. And it is a standalone instrument, so it has the interface and a sound engine. So we did a lot of hard work to connect them, to make them, to make it make sense, right? them to make them to make it making sense right and after is almost finished or now it's almost finished so maybe like a strike striking for balance so I started thinking like meditating about a general purpose musical interfaces which in other words is just a controller for anything they can connect it with. Well, the goal of genetic musical interfaces is to provide the right amount of functions and intuitiveness so it can be specific in a player's sense. It's much like a science module. So it has some functions, has a good interface. So people can use it to build their own system. And eventually, it will become their own instrument. So yeah, modules are like a half an instrument. It will complete the other half in the user's hand. it will complete the other half in the user's hand. Yeah, so strike for the balance, I did something like that. So you will see the stars. Now these are stars. It's called skies and this is sky. And you can switch skies. So it's called skies. That's how easy it is. Well today, besides the interaction of this module, I'm also talking about an abstraction layer that gets me through the design. So I don't get lost when I'm just focusing on the technical details. Because we all have too much freedom these days to do everything whenever we want with coding and circuit. So it's very easy to get lost sometimes. You want to add this, you want to add that, you want to change the design to be something else. So there has to be an abstract layer on top of it, that's a restrict yourself so you can stay focused and finally have some readout in the end, right? So today it's mostly about that. Okay. So my procedure of thinking about general musical interface is that first there needs to be an anchor point. So, basically what the design is based on. So, for a standard instrument, it always, almost always, based on the sound engine because you need to, like is based on the sound engine because you need to like a dedication between the interface to the sound engine. But for a controller, you also need to have some sort of anchor point. For example, if it's a note interface like a black and white keys, MIDI keyboard or like an isomorphic MP or whatever interface it's a note, musical notes is the point. It's based on it. And if it's a dial controller, it's mostly, you can find that most of the dial controller has a mixer-like interface, right? Because it's, the dial is based on like a multi-track idea, so there's multi-track fader to control the volume and all the FXNs almost always mix so for my anchor point for the skies is preset it has a thing for preset, actually my first Eureka module is called voltage memory, it's a preset saver. I created it to play melodies on plum butters, on plum butters. And this is like my latest iteration of a preset manager. Okay, so on this screen the presets are called stars. As you see, the blue one, the green one, the red one. And there's a white one in the middle it's wire shape okay and also I have some like description of it that I want to add to it so it's the anchor point of preset how I want it to be I want it to be gestural, simple, and intuitive. Gestural, I mean, you can play it with this joystick, so you can strum it. It's a very natural gesture. By simple, I mean it's easy to understand. So what's the most we do when we create a preset? We tweak the sound. If we like it, we tap it to save it. And if we want another sound, we tap it to save it. And we can edit the preset. And we can come back to it, right? It's a very natural thing that you do this. It will pick the nearest three stars and interplay between them. And if you swipe the star, it will be randomized. Just swipe the star, it will be randomized. So that's a very natural workflow we're working with Preset. First you have the sound and then you decide where you want to put it and you record it with this joystick. So to make it simple the joystick always work no matter what mode you are at. Right so it's this is like the user can easily grasp it that so there's no multi-mode switching event during this. And it has a parallel output with five parallel outputs, so they all work the same. With the third output, which is speed, this cursor moves, and there's a gate output, which will go on when the cursor is in the stars. And there's another operation mode, which I call Blink. And that, it will be quantized to the stars. So this is where I came to setting up, thinking about a few pairs of opposing attributes. Like the first one I always have in mind is continuous and quantized. But they're very different, right? So on the start I made it very extreme. Let me explain it to you. So when it's in this morph mode, everything is continuous. But you can notice, if you are familiar with this kind of thing, you can immediately notice that there is actually a switching event happening. In fact, when I'm tweaking this encoder to set a new sound, the joystick is not moved, and the CV output should be determined by the position of this joystick. So if I move the joystick, it needs to go back to the sky value and now from the encoder value it's like an encoder will take priority will take it and the sky will take it back but what I did here to make this continual process is that it's not a switching event I use the delta of the joystick to morph it to the sky bed. So it's also continual. When the delta is zero, the morphing stops. And when it's moved, you can't hear it because there's no gate. Alright, you can't hear it because there's no gate. I'll turn this up. Alright, as I turn a little circle in this area, it morphs back to the real value in the sky. To this point. So this is continuous. Also, you have this continual gate when you're on the stars. this continual gate when you're on this stars. Continual movement, continual values, everything's moved. And also this delta is continual value, but in this blink mode, the delta output becomes the distance of the stars you just moved from. As you can see the size of this right cursor directly reflect how high the delta output is. Directly like this output and you can see size changes if it is this distance, right? This is like medium distance. It's like medium size and this is small distance and smaller size And if I have a very far star and go there, right? It's very big So it's also became a non-continuous switching event Became the definite stair when it's in this blink mode and the preset will have a step output the data will have a steppy output and the gate becomes triggered so it's also become as the continuity was broken, right? So on this I have this two different type of mode unity to its own idea. If it's discrete, it's entirely discrete. Nothing is continual. It's all switching infinite. With the morph mode, there's no switching even even if you taking preset that's it that's the idea of it and the other like upholding attribute thought about is is a serial and a parallel right? Because parallel is easier. Serial takes a bit more thinking, like man energy to use. So for a controller like this, it's mostly parallel. And those five output act exactly the same. And what is the serial thing I added to is that I have relative stars. You have seen this blue stars they're all absolute. They contains a single value but this green one Every time you go through it, it changes the middle point. You can hear it better with this one. All right, let me just make it automatic. So it has a CV input. The green one is descending and the red one is ascending. And because when it bumps to the red one, the envelope generator will just push it to the blue one to the right so it will jump here the cv input as the cursor offset is also such a simple idea so you can just understand it immediately, but it can be very versatile. For example, if you have an envelope generator swiping through a few stars, it can be a sequencer, right? And if you have a manual CV or what, swiping through a series of stars, it can be a quantizer because you can set precise those values, and some of it can be a quantizer because you can set precise those values and some of it can be related so with any CV feedback from the other part of the system it can be generated kind of like on song so sick that's basically very much it just saw the function there. Well, actually I want to thank a guy from during the Super Bowl. So he suggests about this. He kind of like wanted his cursor to be snapped to the stars because he was afraid that he would miss it with his neural movement. So I kind of expanded on this idea to have us set it up, all this switching event across the output. So that's right. I want to hear some comments from you. Yeah. Questions? Yeah. I guess that for this kind of layer of abstraction, to say that you have chosen the metaphor of the stars and the distance between the stars, how you navigate that. That is a very particular decision, how to create a way to navigate the natural parameters or signals of the synthesizer. Obviously, you see this as a way that maybe new metaphors, new ways of navigation would be developed in the future, not only by you but by other designers as well? Yeah, I was about to, my goal is actually to provide a thinking, like maybe something like a thinking habit so you don't really get lost when you're along the way because I know there's a lot of instruments designers, especially the ones that are looking to have them product out. You will have this problem that they're kind of stuck, always stuck in the middle somewhere, that they don't know where this particular product should go because they always have too many options available and this is a very flourishing market the center of the market because human interface is just there's no right or wrong because it's almost emotional right the instrument is kind of emotional thing there's no or wrong. So everyone is coming out with their own ideas and the technology we use is very accessible, allowing us to do whatever we want. something that restricts, that's restriction behind the techniques can actually help a lot to for you to concentrate to be something like a final product. So that's what I'm thinking about. I know Tangent Music Lab is more about experimenting so you're more like on the scientist side, right? You don't care whether, you just care what is happening in front of you and to possess right maybe maybe we never become something like like a sample per lab but like engineering is quite a difference someone's like an like and said it's never compromised by engineering is an art difference. It's almost like, like sentence never compromise, but engineering is an art of compromises. We compromise everything. So yeah, it's just some sort of a compromise. I hope it's kind of helpful. So if you have any questions, let me know before I move on. Yeah, like you, Let me know before I move on. Each knob does exactly the same, tweaking the corresponding output. They work the same. The 1 for the 1, output 2 for the 2. I use this indicator just similar to a mahjong. indicator just similar to a mahjong. Since this is so abstract, it's very inviting to see this Moise platform and everyone could write their own kind of film for different modes. Is this also something you had in mind? Yeah, of course, if you want to just make it open? For instance, yeah, or maybe just for you to see more as a platform. This is what happens for some of the instruments, like for instance, Bustle instrument people, they also kind of have this one module and they use it for multiple products now? Well, you know, the actual, the hardest thing about an open instrument is not the product itself, it's the support. Without enough support, it will make no sense, right? Because people will encounter every problem they the users will have different technical backgrounds and they may face all the problem because they have different operating system the computer setup and that support is actually the the real deal when you have an open instrument product. That's why Monom only has a handful of products because each one they need to spend time to build all the community. And the technical support of the documentation so is very crucial it takes a lot of time and yeah maybe I'll just make it a black box of people and just in grab and play to make some noise the right the for mister now just you the DC voltages in each input. Any plans to make some kind of built-in, a bit more complex modulations? That's a great question. That is exactly what a maker would face when he designs his own thing. Of course I can. I can make each input output in different kind of waves, different kind of envelopes, right? Make it a digital sound generator. For example, if I move to this and this becomes a sound wave and then it can jump to another wave and morph between them, which is like fantastic if you kind of implement it right. But it's very like, it's very hard for this interface. I mean technically you can make it do everything, but to make it making sense on this interface would take a lot of pre-planning you need to take a lot a lot of thing out of your ideas and actually I think it's almost like cheating when we design modules right because if you say if the user wants a function they always have another module right so make a module make the function simple the function simple functions are simple and so they they have the mind capacity to trace everything right yeah and one thing i always enjoy is um One thing I always enjoy is the patch programming possibility of a module. Right from search, it's like the USG, it can be patched for, patch itself for, I think maybe like 11 to 9 to 13 different functions. So, it's a module with simple electronic functions that you can just patch it to become everything between oscillator to analog generator to low pass filter. This always fascinated me. I want to bring something similar to a controller. So, for example, this precision input, it's like you can make it doing all sorts of different functions. Just as I said, it can be a sequencer, quantizer, or it can be a hot letter for your hand maybe you have some no signing horizon delay and uh... alphal is the happens weapon sweeping through it like a strongman and you move up and down to change court this is i mean there's also many ways to use it and it's all come from one simple function that this kind of thing that's fascinating me. I like this thing, not a very definitive, like a music function. So there's a user with a, be another layer between the simple function and the music op. Right? Yeah. Well, this is just about everything I want to say. Finally, are you going to show something in the oscilloscope? Yeah, I just want to show how the output looks like on the oscilloscope. It appears like a sun wave because of this. Maybe the ground is not there, so there's no ground connection, so it's very noisy. Okay. All right. so it's very noisy. You can see this is very, like I find it very interesting. So I have this cursor here because it's quantized to stars. It's like stick to it. But if I move this joystick the bass position of it. Before the quantization, the CV is offsetting it. So this can also be used for performance if you want. And one other thing that you may already notice is that this joystick is exactly the same size as this screen. Which is good, why? Because if you have this head of this joystick just So, which is good, why? Because if you have this head of this joystick just slightly off this black circle, it will be slightly off the screen as well. So it will have a very direct visual feedback. I mean the size. So exactly copy of your precision of your joystick. So this is kind of like a design detail to share. All right, so why am I presentation says this? Yeah, actually this morning I've been programming this new feature. Everything so last minute. Everything is so last minute. And, right, if you go it's just half an hour. Can you explain again the outputs? So there are five controller outputs, but there are two more? Yeah, there are two more. This one is data and it is the speed of this movement when it's in the MRF mode. And this is the distance between the one you just moved from when it's in this mode. So it's not the real data anymore. It's just a distance output from here. And it's reflected by the side of the circuit. So yeah, so this is maybe like my idea for the so-called preset management. I want it to be like much more beyond the set of parameters I recall. But anyway, it's not a real preset manager because it can only save like five voltages. I can use it anywhere. So it's kind of limited, but I think the limitation is good. If you like morphing the whole state of a machine, maybe like a real, maybe like a subtractive sentence, like a few tens of controls, it will be, well, it can be interesting, but maybe overwhelming too. Right. Oh, yeah, yeah. My stuff is very time saving. So what? I just have another question. Did you directly start programming it on the hardware, or did you do some prototyping in C++ or some PD patching before? Or how does the process look like? Oh, yeah. That's because I do have the experiences in this chip and stuff. So I just directly draw the circuit. I just draw the circuit because I know it will work. And then I start programming on the platform directly. And are you working first on like stripboards or are you directly ordering PCBs because you know this will work? Yeah, I directly order PCBs, yes, because it's very convenient in China. So like it's a week and it will be ready. Solder. So yeah, that's actually a good thing. So with a standard instrument, with sound, you always have a software prototype before you actually move it onto the actual hardware because you want to know whether it's a, if it's interesting, right? But with controller you don't need to because you don't need what users will use it with. There's a, there's a Sam from, from 1974, I met him because he had a lecture in Beijing, Max, obviously. And he told me there's like some sort of neural network of this interpolation, right? This kind of thing. because it's controller that I said is there's no better interpolation because you don't know what the user will use it for right it's a one specific kind of interpolation only is better when it's controlling a specific sound a specific set of parameters for For example pitch. For pitch you may need a special curve right because that's why a hard drive sounds great. Like if you morph between two sets of drone notes, like morph between two chords, maybe you want a curve like this so it stays more in the stable zone but if you don't have a definite set of parameters that you are working on controlling the there's no better interpolations I just use linear yeah I think that's pretty like all around a simple thing to do. And it's only the reason that I only choose three of them to morph between. Oh yeah, maybe in this blink mode I should remove the lines. Now the lines are a little bit misleading. I'll do it when I get back to my room. Okay. So with these three lines you can see it indicates the three nearest stars are moving the output from... I mean for 2D interpolation, the three is like the lowest number, right? Because with two, with two, there's no 2D, right right if there's only two stars it will be like a 1d interpolation so three is the lowest number the reason I chose it because this will give the most versatility like if you have a very big value changes between all these adjacent stars when the when the focus I would say so I would call those three star focus when the focus, I would say, I would call it a three-star focus, when the focus changes, you actually have a big change of the target value, but it won't be going to the value immediately because, as I said before, in this morph mode, the output is interpolated by the dirt. Right, so that's like interpolation speed. So there will be no jump, no instant change when this line jumps. But it can be a big change, let me see yeah, it is like that you have big value changes but it is also great One thing I found pretty interesting that because I'm working with Trent from Mannequins, women's car wraps, that all these three modules, the is just friends and uh so that they're all based on parallel ideas because the precise matter is it's obviously parallel right this just friends it's like six parallel altars it's controlled by this knob that they can span the span parameter, the span is the speed of six voices. Like controlling a group of parallel stuff with one parameter kind of thing, like group controlling. And this silhouette is like, it has six inputs, you can scan through it to become the focus of the... It can be focus, it can be background, it's also about parallel signal management. So all these three modules share the same idea. So that's why I really like this one last point I made, which is like opposing attributes. It's a very fun thing to think about. You have the continual part in your interface, you have the context part, that's a pair. You have the serial part and parallel part, that's a pair. You have the user set part, and you have the preset part, that's a pair. For preset manual, you always have enough user set part, right? Because that's where they put in their preset. But for a standalone instrument, there's a lot of preset values inside, so it's more or less like modular patch, that's with all the tunators set just the way you want it, and you just encase it into a standalone shell right and what what else I can think about yes and the other is a tonal and tonal pair right which we saw really a lot for designing that horridum I bet you can find more that this kind of opposing attributes when you're making your own design. This is actually very helpful. Because for some parameters it's better to have a context. For example, like a page. People always want to play notes. Not everyone's a ceramic player, so. But for some other tonal parameters, timber parameters, you may want it to be continuous like this. Right, so this is one other point I want to make. Cool, I think, I don't know what to say. Oh please, Marty? Can you maybe share some of your approaches in literature teaching and workshop teaching? So what are the topics you're really addressing from the technical or conceptual side? So maybe from your practice and teach. So how you're... What, how I teach in class? Yeah, so what topics are you teaching and what are your focus is more... conceptual science, so how you're taking... I don't know how conceptual it can be because I don't... my mind is empty now, but for university I usually... I teach this thing, that's sound synthesis, that's one topic I teach this thing, that's a sense, sound synthesis. That's one topic I teach. Max, it's kind of another thing. And the other is like, I do like DIY workshops. Not much of the C++ programming because it mostly happens at music school and art university. I teach the students which actually have a painting background. So basically you are facilitating technical knowledge to music and art students? Yeah, yeah. Basically I'll teach art students. I think maybe some instructors. Scrap sharing. Yeah, I thought about that. Yeah, I think. That's it. And in the workshops you're doing, it's mostly settled around the instruments or... Yeah, yeah. Sometimes I have a... Well, that's just... Sometimes I have a PCB that is already made and they only make the interface. I mean, the functions are already fixed. They have maybe like a... There can be like a resistive sensor or something, and they can do like a resistor ladder and some contacts along the way, along the line, to make it like a guitar shape, or it can be anything, like LDRs, whatever. So they basically make the interfaces. And the other type is based on a development board, such as Arduino and Tensie. And for them, on the coding side, mostly just simple playing. Because that's easier, the easiest to grasp, so they can move on to the interface, the design pretty quick. One advantage of the art students is that they are very good at, actually very good at the, the form I shall say so I think it's the amazing form of these interfaces where it's like turning a hand off of a jam music box or I had actually had one student has this the interest in which I always kind of like to remind me, which is great. So he has a lot of gears for the knobs. So for this knob, they have gears. She has gears around the shaft. And he uses a rubber band to put them together. So when she's turning one knob, the other will turn in different ratios based on the size of gears. So it's like very physical micro control. And she actually has no experiences in this musical instrument. But it's actually a very good idea. But you can actually build the micro control physically on the interface, something like that. This kind of thing always impressed me. So I think interface design is pretty much a low tech thing, but it can be very freeing to do. There is a question from the audience online. One person is asking what is the role of physical gesture like touch, tap, twist, drum, strum on your instrument design? Do you start with a physical gesture and build an instrument interface controller or start with an instrument synthesis concept and explore which gestures have the best interaction? Oh, that's why I love questions. I love communication. It's a very good question. Because for us humans, different gestures, we have different amount of control, right? There's a lot of gestures there. But some of them, like precision, we have very precise control, right? We can precisely control the precision on an interface, even for ceremony, we have this in the space gesture. Right? We can precisely control the position on an interface. Even for ceremony we have this in the space gesture. After practice we can handle that, right? But we have very bad pressure control. It's like you can't press a thing and hold it steady enough to stay in that place now. And it can be used studying now mistake and it can be used statically for the interfaces can be fun if you want it premier to be simon bank on probably can put into those last control of all that security put into those less controllable, less accurate gestures. So the question is whether... You start with a gesture or you start with sound synthesis. For any general interfaces like a controller, there's no sound to be considered. So, the first point I made, which is the anchor point, is that you need a, so this one can be, can make sense. So, for this, it's preset. For some of this, MPE is note, it's musical, obviously musical notes. There's a... if you really like purposeless interface, it's only a bunch, it's always matrix. You can see there's a MIDI controller that's like a matrix of knobs with nothing else, just all the same knobs in a matrix. And there's a matrix of buttons, like the Monum one, Monum grid, which is like functionless and definitionless. So it's a, I find it's very interesting. They always a matrix of the same thing, same size, same height. But for musical interface, especially for stand on instrument, you always have different size of knobs, right? Some is more important in the front, some is less important away from you and smaller. I mean, any, in this case, gesture is just, you you know limited to this exploration in two dimensions yeah yeah but obviously we are talking about pressure that it could be you know multi-dimensional right yeah yeah so they use of gesture of sensors whatever yeah it's something you want to explore in the future or it says I think for a product it's difficult somehow to explore many degrees of freedom. But can you share maybe some thoughts on that? Well, there's a few concerns based on that. There's actually a four-dimensional joystick, or three-dimensional one, where there's a turn and press. But for me, an instrument needs to look good, but that's too chunky to look good. You really need a musical instrument to look good, so we want to play it. Maybe I'll make it smaller, but still leaving enough size in hand. And the other thing that I don't have any momentary pressure sensor on this. I know someone on Superboost is doing this joystick type of control with a string that he can just pull. Almost like Ernest Macdonald, that kind of control, pulling string control. But that is, that is exactly another set of opposing attributes, which is like a momentary and toggle. So that is so momentary, which go exactly against to the idea of preset, because with preset, you need it to stay somewhere, right need those joysticks to stay somewhere if there is a momentary element to it it just feel wrong right so you cannot come back to the same value immediately yeah and if you let's just imagine if you have a lot of pulling string here I mean their product makes sense because it's live controller, right? And it's even going back to the middle. It's like a separate setting joystick anyway. So if this has a pulling string on it or anything momentary, it's very hard to keep it, to keep the position intact when you're playing the momentary thing. So you actually can't use it. If you use the preset, you can't use the momentary thing. If you use the momentary thing, the preset doesn't make sense. It can only be a gestural controller, which is just part of it. So yeah, thinking about all this this recording attributes do helps a lot while I'm doing that question so it's just a two-dimensional joystick and there's no pressure or anything in this country just use the touch display and use two-dimensional tissues, isn't it the same? Yeah, but touch display has more latency than joystick. This is actually responding very fast. It's much faster than the screen refreshes. So it can. Let me see how fast I can make it. Because there's a feedback system, so now it's out of control. And the sky recording, which you can see the bank recording is like instant, so you can use it during play. Right. Right. You can jump between them and stuff. Whoa, this is so big. I think it's very fast. It's much faster than the certain modes. So that's why I didn't use the touch, because touch screen is good for like a simy real time setup than real time play. like latency is a really big thing when it comes to the musicals and stuff okay well it's almost an hour all right thank you do you want to play maybe a couple of minutes of this? Or is it too difficult now to arrange? You guys, what's even more interesting, do someone want to play it? Yeah, maybe. OK, just come and play it for a few minutes. OK? Let's try how this will respond. So thanks a lot to Monty. And thank you. Thank you so much. Tomorrow at 7, there will be a concert at Linz with Monshi and also Sosia and Sandra Mufi. So please, you are all invited to go. Yeah, another half an hour of this terrible sound. Thank you. Okay, so maybe, yes, let's go and try it. Let's try it. So you have a bookla. Oh, that's search. But it can be bookla, because there's like a Of course PSU since I'd come right to MCU power so you can 15 minus 15 plus 15 no problem It's just a matter of that power input There's a just hear the... So the just-run is still on. If I don't have this big bag, it can be... Stable? Oh, and if I turn the knob, it becomes white. It's just indicating you're off from the real value. And it only makes sense in the blink mode because everything is switching events. Like I said, it's like a Unity routine, right? Everything is switching events in this mode so you can have different states of the circle. But in this mode there's no difference in circuit, everything is continuous. No switching at all, just try it out. So this one is just gate signal? Yeah, that's gate and trigger. And this? It's gate in this mode and trigger in this mode. And this? This is data. It's the speed that you move cursor in this mode and it's a distance between this. So it's also DC voltage? It's all DC voltage. It gives you only the voltage when you move? No, because it stays steady for the distance. Like if you have a, let me start with maybe this day. Maybe it's here. You have this far on here, so it's very far. You have here, nearer, so it's smaller. We'll change from here. Here it's getting big because of distance. It's the same. It's the same. It's the same. It's the same. Yeah, yeah, just take a set. Yeah? Take a one. Take a set. It's okay now. Why are those icons so different? I guess it's a kind of a preset. What's the sound? I guess it's a kind of a preset. It's a parallel sound. It's a piece. It's a parallel sound. The signal goes to the DC controller. Basically with the joystick you have two coordinates. They're just not the same way to the screen like the model question so it's not like the screen that we have in between in a way because switching no matter how many stars you have on screen you have two modes to navigate you don't you just don't care how many stars you have on screen. When you press it's like this one. No, it's the same screen. Oh, you mean it's shifting the skies, right? You can save a lot of different cells. Not from the TV. It's almost like saving... There was a mode where you could enter new settings, new presets. And you switch the values. You press and change the preset. For each of these things? Yes. Presses? Yes. And then you have 5 DC voltages just 5 different values but they are fixed in this you can play ok ok 5 tones like yeah you mean, uh... No, no, no. You also said you can, like, randomize it. A different sky. Look at the sky. One sky is one planet. You mean interplay between two you have a position right here, and you interplay, and they start to move towards you. Which starts to move towards you then? Because you have a lot of two.