I've just invented a new musical instrument in my BRAIN. (Well, more, an interface. But.) All I need is a large multitouch screen and an SDK. IGOR! BRING TO ME $10,000!
No? Foo.
Hm. Okay, I just need cheaper ways to do it. What does fibreoptic cost in bulk these days? It's gotta be cheap, right? Sensors are gonna be spendy tho'. On the other hand, that could be crazy kinds of analogue, which has its good points. And wouldn't be as mobile, no flick thing. I want that flick thing. Foo.
Okay, really, it's very simple. So if you took a piano keyboard and warped it into an arc, yes? And made it flat and touch sensitive (see above, multitouch screen), and tone-wrapping rather than tonally fixed by position of finger on key, but with a kind of snap-to-centre functionality (over position... and time? That should be a setting) so it doesn't turn into a Theremin. Then add the ability to flick the "keyboard" so that it'd move underneath your fingers, without having to lift your fingers. And you could have volume by area-of-touch as an option, and lockdown and snapto positional points and things like that. (eta: note that the "keyboard" is the "playtable" and part of the multitouch screen, not a separate keyboard or keyboard-like device.)
The base sound would be synth, of course. You could have waveform editing on the fly by drawing them on other parts of the touchscreen. And you could have a bunch of saved FM waveforms, and drag and stretch them over parts of the playsurface, with all kinds of combinatory options. (TBD: if you flick, does the affected region stay associated with screenspace or with virtual playtable space?) And you could do sampled noises as well, of course. Modified, maybe, by the FM waveforms as overlays. The permutations are of course obvious. You could have pedals as options switches, for likes like "combine" and "multiply" and "subtract" and "sustain" or "extend" and and and.
For maximum operatic effect, allow dragable/applyable stage effects controllers.
eta a few minutes later: force-feedback (rumble) gloves are an option for tactile response.
No? Foo.
Hm. Okay, I just need cheaper ways to do it. What does fibreoptic cost in bulk these days? It's gotta be cheap, right? Sensors are gonna be spendy tho'. On the other hand, that could be crazy kinds of analogue, which has its good points. And wouldn't be as mobile, no flick thing. I want that flick thing. Foo.
Okay, really, it's very simple. So if you took a piano keyboard and warped it into an arc, yes? And made it flat and touch sensitive (see above, multitouch screen), and tone-wrapping rather than tonally fixed by position of finger on key, but with a kind of snap-to-centre functionality (over position... and time? That should be a setting) so it doesn't turn into a Theremin. Then add the ability to flick the "keyboard" so that it'd move underneath your fingers, without having to lift your fingers. And you could have volume by area-of-touch as an option, and lockdown and snapto positional points and things like that. (eta: note that the "keyboard" is the "playtable" and part of the multitouch screen, not a separate keyboard or keyboard-like device.)
The base sound would be synth, of course. You could have waveform editing on the fly by drawing them on other parts of the touchscreen. And you could have a bunch of saved FM waveforms, and drag and stretch them over parts of the playsurface, with all kinds of combinatory options. (TBD: if you flick, does the affected region stay associated with screenspace or with virtual playtable space?) And you could do sampled noises as well, of course. Modified, maybe, by the FM waveforms as overlays. The permutations are of course obvious. You could have pedals as options switches, for likes like "combine" and "multiply" and "subtract" and "sustain" or "extend" and and and.
For maximum operatic effect, allow dragable/applyable stage effects controllers.
eta a few minutes later: force-feedback (rumble) gloves are an option for tactile response.
no subject
Date: 2008-08-03 06:02 am (UTC)Hm, google found me a paper about it from 2005 (http://portal.acm.org/citation.cfm?id=1086019).
no subject
Date: 2008-08-03 06:08 am (UTC)no subject
Date: 2008-08-03 06:12 am (UTC)no subject
Date: 2008-08-03 06:15 am (UTC)no subject
Date: 2008-08-03 08:39 pm (UTC)Nothing remotely like what you're suggesting - it is a single-touch [DS version at least, but I think the hardware ones are too] pad with x/y axes. But it probably has some of the same feel - and you can just get it now. ;)
Dragging cutoff/peak(resonance) is fun; and the crazy solo note generating mode is good too (it lets you select what scale/key you want, then note = x, note length =y, so you can go near the top to have a continuous sound or down the bottom for 'bip bip bip bip'). Although it does end up making stuff that all sounds the same :)
To be honest I think your idea sounds a bit Minority Report - would look great in a Hollywood film, probably no use in practice. Real keyboards work pretty well for 'expressive' performance (including aftertouch, if you want one extra dimension you don't have on a piano, and pitch-bend wheels etc if needed), real knobs, dials, and faders are pretty good for shaping sound, and while 'touch-pad' xy positioning does indeed have interesting results for effects control etc, that's available too without having to make a complex control surface. And the tactile feedback is a pretty big thing that probably can't be addressed simply by rumble gloves. I guess I don't really see what problem you're trying to solve.
(As for live waveform drawing - is that any use? I've tried, and it's pretty hard to draw waveforms that sound anything useful [our brains are more frequency-domain]. I can see using such a panel for envelope-shaping, but even for envelope-shaping, probably would be better if rather than actually drawing it, you could just drag the waveform to the desired shape, so as to ensure a smooth transition. That could be pretty cool. Like when you drag part of a Bezier curve in Illustrator/Inkscape. But, on the other hand, attack-decay-sustain-release has actually done us pretty well for the last 40 years or whatever, and it's only four knobs...)
no subject
Date: 2008-08-04 05:03 pm (UTC)I don't know that I'm trying to solve a problem as much as I'm making something (in my head, anyway) that I think would be fun. I think this would be fun. It may not be practical; won't know until I play with it. But I think it'd be fun.
no subject
Date: 2008-08-16 10:37 pm (UTC)Did you see (I forget if I linked it) my brother's pet rock thing? This was part of his PhD... it was a (real) rock in a cage, with the door open so you could reach through to touch it. You could pet the rock and it would make noises. The noise it made was different depending on how you stroked it and how hard you pressed. (This was the actual research innovation it was supposed to demonstrate - it worked by processing webcam images, and your fingernails change colour when you press hard on something, so if you have a camera you can make any surface pressure-sensitive.)
Anyway just saying it could be a lot of effort to get your idea done and you might end up with something not much more use than a pet rock. Or a theremin, to use your earlier example. :)
(And I remembered now - it was a VERY long time ago but as a kid, I did indeed once write a [DOS] program where you could draw a waveform with a mouse and it would play it back looped at different pitches etc. This is how I discovered that drawing waveforms doesn't work very well, which I had vaguely remembered in my earlier comment.)
no subject
Date: 2008-08-18 04:27 am (UTC)Drawing with mice is t3h suxx0rs, tho'. (yes yes I know different but don't care YOU CAN'T DENY MY DREAMS or something) @_@
no subject
Date: 2008-08-06 01:13 pm (UTC)