Glove Emulates Musical Instruments

Zothecula writes "Bridging the gap between computer generated music and real-world instruments, the 'Imaginary Marching Band' is a fledgling, open-source project that allows music to be created by imitating the actions of playing the real thing on a sensor-equipped glove. The work of Scott Peterman, a Masters student at Parsons New School Of Design in New York City, the prototype system uses MIDI data output from the gloves via USB to reproduce the full range of notes from instruments such as the trumpet and trombone."

Comments

Air Guitar!

[LordStormes]

Win! This is even better from the Air Guitar Shirt on ThinkGeek ;)

self-love

[Anonymous Coward]

I wonder what it sounds like when you masturbate with that musical glove on.

I bet it sounds like Lady Gaga being raped to death.

Re:self-love

[LordStormes]

Probably a lot like the trombone.

open source

[Anonymous Coward]

thermoforming machine

Two observations

[Barbara, not Barbie]

Young children will love this.
Parents will hate this.

So now you know what to get for your evil twin sisters kids for Christmas. This has the potential to be even worse than giving them a chemistry, carpentry, or woodburning kit.

Re:Two observations

[chemicaldave]

Parents will hate it? Please explain.

Re:Two observations

[BattleApple]

I don't understand that comment either. Parents sometimes regret buying instruments because of the noise, but real drums don't have volume knobs or headphone jacks. Plus, they're more expensive.

Re:Two observations

[Barbara, not Barbie]

Parents will hate it? Please explain.

1. Kids will love it because it's NOISY. I guess you've never had kids play "cymbals" with the pot lids, or "drums" with the pots and wooden spoons.

2. Parents will hate it because it's NOISY, and when they turn it off, the kids will be equally noisy whining that they want to play with their game.

The only thing missing is a "pull my finger" fart-noise generator.

Re:Two observations

[Beerdood]

Actually, parents will love this. You can control the volume on this thing. A real trumpet is LOUD, and the learning curve is very slow on making a good sound, so your kid will play a lot of rusty notes for that first year.

Re:Two observations

[chemicaldave]

so give them headphones

Welcome to the late 1980's

[WatcherXP]

Bands have been doing this for decades

Re:Welcome to the late 1980's

[Anonymous Coward]

If only someone would translate this idea into a video game controller!

Why the glove?

[courteaudotbiz]

When you could use the Kinect?

Re:Why the glove?

[Nerdfest]

I don't think the Kinect even approaches the resolution required to pick up the finger movement used to play musical instruments.

Uhh, no

[Anonymous Coward]

By definition, any digital sampling is incapable of producing "the full range of notes from instruments such as the trumpet and trombone". Especially for the trombone.
And it's not possible to imitate the actions of many musical instruments, as the tactile feedback is 100% integrated into the production of the sound. For example, it's virtually impossible to do a 'drumroll' without the rebound action of the sticks off the drumhead.

When this guy can demonstrate his glove playing a live, 100% perfect rendition of a Rage Against The Machine, track, call me. Good luck with simulating pulling the mic cord out of the guitar output jack and using it as a pick while it's still hot.

Re:Uhh, no

[spaceplanesfan]

I am putting my karma at risk here, but you sound awfull like:
"CDs can't reproduce full fidelity of vinyl"

Although I somewhat agree that this glove won't have enough resolution to be perfect,
theoretically it is possible to create such one that would.

Re:Uhh, no

[Anonymous Coward]

Good luck with simulating pulling the mic cord out of the guitar output jack and using it as a pick while it's still hot.

I want to know what the hell the microphone cord is doing plugged into the guitar output in the first place. Nevermind that you're also talking about an XLR vs 1/4" plug disparity...

Re:Uhh, no

[BattleApple]

GP reminds me of an interview I heard in the 80's with a synthesizer player.. I think it was Richard Wright from Pink Floyd. The interviewer made some comment about synthesizers being artificial and not legitimate instruments, and the keyboardist said something like "Why don't I give you the instruments and you make the music?"

Look at keyboards now.. They're practically indistinguishable from some the instruments they emulate. Anyway, it's about the music. I don't care how the sound is produced if it sounds good to me.

I think something like this could be good for parents that can't/don't want to spend the money on a real trumpet that their kid is going to give up on after a month. If the kid shows some proficiency with it and still wants a real trumpet after a while, then get a real one.

Re:Uhh, no

[jeyk]

Good luck with simulating pulling the mic cord out of the guitar output jack and using it as a pick while it's still hot.

I want to know what the hell the microphone cord is doing plugged into the guitar output in the first place. Nevermind that you're also talking about an XLR vs 1/4" plug disparity...

Even if we ignore the discrepancies and assume it should read "...pulling the cord out of the guitar output jack..." the result would be... silence. I don't believe the sound of an unplugged electrical guitar could be heard in the rest of the band still playing at full volume.

Re:Uhh, no

[maxume]

I think the point is more that if you want to simulate a trombone slide, the most likely way to succeed is to build a trombone slide.

Re:Uhh, no

[digsbo]

You and GP both sound like people who have never studied music seriously (hours a day), while GGP likely has. The level of tactile and other feedback from physical instruments affects performer behavior in ways you can't possibly imagine. The responsiveness of a reed on a mouthpiece, the delicate subtlety of hammer action on a piano while one tries to play legato chords with the pedal which changes the weight of the keys due to the dampers being lifted, the vibration of the low brasses and woodwinds which have an almost massaging effect on the player's facial muscles are all examples I've personally experienced. You may not care, but a musician knows that a digital piano is NOT a digital equivalent of a piano. It is a slightly different instrument which is not an acoustic piano. That doesn't mean it's not a great thing (I love having a digital to play on), but it is not, and will never be an acoustic instrument, and will not produce the same output. That you don't care for some value of "close enough" is a testament to the engineering, not proof of the perfect equivalence of the two.

Re:Uhh, no

[Anonymous Coward]

Even if we ignore the discrepancies and assume it should read "...pulling the cord out of the guitar output jack..." the result would be... silence. I don't believe the sound of an unplugged electrical guitar could be heard in the rest of the band still playing at full volume.

You clearly have never touched the tip of a live guitar cord, let alone run it across other metal objects. But for some reason you felt it necessary to offer your opinion on a subject that you obviously have absolutely no knowledge of. That's what I've come to expect from slashdotters.

Re:Uhh, no

[SleazyRidr]

Real instruments will still exist for people who are serious about playing them, they'll never go away.

This is basically just a toy, a way to 'try out' a bunch of different instruments without the capital outlay of having to buy them all.

Re:Uhh, no

[Anonymous Coward]

Dude, everything you just described, or will ever describe can be simulated.
It's as if you never heard of physical modelling. ...or proper programming.
And there is a "close enough", even for you: When you, from the way it behaves and sounds, can not distinguish it from the "original" anymore. Which is way less close than $400-wooden-volume-knob-"for-warmth"-buyers and similar "audiophiles" would ever admit. So there is no point discussing quantum-physics exactness. (Which you'd still not accept.)

Also, who said one has to go for exactly simulating one single limited instrument, when you can do so much more.
Synthesizers like Absynth already sound better than every instrument ever invented. There are already controllers allowing more creative playing styles than any instrument ever built. "Real physical" instruments are the second class now.

I wonder if you refuse calling playing music with microphone and speakers "real music making".

Luckily, nobody asks you.

Re:Uhh, no

[BattleApple]

I own 5 guitars, a digital piano, and an old upright piano. I'm not saying this could ever replace real instruments, but it could be a cheap alternative for people that aren't that particular about the mechanics. I play the digital piano more than the acoustic, because the strings are too old to stay in tune, and it really wouldn't be worth it to replace them. I'll always rather play a quality real instrument.
Don't conclude that I'm not an experienced musician because I can understand that there are people that would like to have something like this. I wouldn't even want one of these things, other than to play around with for a few minutes. There are also other applications.. people with disabilities, etc.
Another thing.. just because someone appreciates a real instrument over some kind of imitation doesn't mean they're any good at playing it. As far as listening to music, I'm pretty sure anyone would rather listen to good music played on fake instruments than bad music on real instruments.

Re:Uhh, no

[CannonballHead]

Look at keyboards now.. They're practically indistinguishable from some the instruments they emulate.

That is pretty false. I've heard some pretty expensive keyboards as well as digital/electronic organs (the pipe-organ-reproducing kind). They are quite good, and perhaps when recorded, hard to distinguish. When played in a live room, it's a lot different. Even with great speakers, an acoustic piano sounds different. Better? Worse? Well, that's an aesthetic, and I'm not arguing about that. If you like keyboards better, that's totally fine with me. :)

I think something like this could be good for parents that can't/don't want to spend the money on a real trumpet that their kid is going to give up on after a month. If the kid shows some proficiency with it and still wants a real trumpet after a while, then get a real one.

You can get a student trumpet for ~$150 (super cheaply made, from China) and a decent student trumpet for ~$300. I doubt the glove will be much cheaper than that, probably has a lower resell value, and is probably a lot more easily broken. And you can't play it in a band. And it requires speakers. And midi cables. And electricity...

Re:Uhh, no

[CannonballHead]

Revision: last I saw, I think you could get the Chinese made super cheap trumpet for like $70 on ebay.

Re:Uhh, no

[MarkvW]

This is one of the best slashdot posts I have read in a long time.

Re:Uhh, no

[Anonymous Coward]

the sustain pedal doesn't change key weight. The soft pedal, however does as the hammers get closer to the strings by default.

Re:Uhh, no

[digsbo]

My bad, I thought you were saying it was a replacement. I agree with you that digital stuff is an excellent substitute for many, or even most, situations.

Re:Uhh, no

[digsbo]

Yes, it does change key weight. The damper is lifted when you strike the key. When you use the pedal, the pedal lifts it, meaning that work isn't done when striking the key. This is most obvious if you're playing soft passages and trying to emphasize certain notes over others.

Re:Uhh, no

[digsbo]

Ok, AC, I apologized to GP for misunderstanding him, but WTF? What would be the point of simulating the vibrations from a bari sax? Wouldn't it make more sense to have an actual instrument? All the physical things that happen provide that extra level of uncertainty and expressive flexibility...they will never be 100% recreated because there is simply no point. Past a certain level of mimicry, it will be as hard to play and as expensive to make as the actual thing, so the actual thing (as opposed to a digital reproduction) will be so much cheaper there is no economic reason to try and simulate it more accurately. Thus, it won't happen.

Re:Uhh, no

[BattleApple]

yeah, definitely not a replacement.. I probably didn't convey that well. I do think there's something about real instruments that can't be quantified.

Re:Uhh, no

[BattleApple]

Actually I was thinking about recorded instruments when I said that.. I don't know if a live piano could ever be simulated. You have long strings inducing vibration in other strings, the soundboard, and other parts of the piano. Maybe this could all be digitally modeled someday. When you play the upper register on a digital piano, it does simulate the resonance of the surrounding notes since they don't have dampers on a real piano. It's pretty rudimentary though, at least on mine. (cheap)

I didn't realize you could get a trumpet for $70, but a really poor quality instrument can cause a lot of frustration for a beginner. I don't really know anything about trumpets though. I know a crappy guitar can be discouraging, mainly because of fingertip pain, but that's going to happen with any guitar to some degree at first.

Re:Uhh, no

[BattleApple]

well if this thing is going to be amplified, there's gonna be a cord to pull out somewhere. Not sure what you'd run it across though.

Re:Uhh, no

[hcdejong]

And it requires speakers.

That would be an advantage in some cases (e.g. practicing in a non-soundproof room in an apartment building). Finally, a brass instrument that can be played quietly if so desired.

Re:Uhh, no

[Anonymous Coward]

You just repeated the exact point I made. (See line 3 of my above comment.)

No idea how you managed to see it as the exact opposite of that. Wishful thinking?

Re:Uhh, no

[CannonballHead]

Hum. Yes, digital recordings of a piano ... it's hard to distinguish a keyboard from a piano, but I'm not sure if that's just the way it's recorded or what. If comparing digital recordings to digital instruments then yeah, it's almost the same medium to begin with I guess :) Your explanation of the string vibration thing is one of the reasons I would think it would be quite difficult to realistically digitally model...

Yeah, you can get pretty cheap trumpets. Poor quality? Yes.... but you can still learn pretty much all the basics, you won't hurt yourself, and it'd be more versatile than a glove. The glove is missing some player input that the player does with his or her lips, tongue, and air... e.g., articulation (how hard to attack the note, how long to hold it, even how to end the note - end it by stopping air flow at the throat or end it by stopping air flow with the tongue? makes a different sound on the trumpet).

I guess I would put it this way.... to me, as a trumpet (and bass guitar, anyways :) ) player, among other instruments, I would see a single "glove" instrument functioning as a trumpet roughly the same you might see a "Guitar Hero" [toy] guitar functioning as a real guitar. You may be able to get it to sound the same in some ways, but it's missing a whole lot :D :)

Re:Uhh, no

[CannonballHead]

Hehe, touche... on the other hand, it's possible to play quietly. But who wants "quiet" nowadays besides the old boring classical music types ;) (that would be me.)

All you need is Glove

[digitaldc]

Watch out! The Glove is really a weapon of the Blue Meanies.

Marillion?

[dunkelfalke]

I think Steve Hogarth used something like that

Re:Marillion?

[LizardKing]

Yup, he did. Made by Kenton Electronics in London - I remember reading about them when I was still at school in the 1980's.

Brass?

[cforciea]

Do you wear sensors on your mouth for the trumpet and trombone to change buzz frequencies so that you can play more than like 7 notes?

Re:Brass?

[operagost]

It looks like it has an air pressure sensor, which would allow one to adjust volume but not pitch. I can't look at the video now, so I can't say whether this is how they interpreted the data or if they erroneously use air pressure to adjust pitch as can be done with recorders (but not brass instruments).

I feel like such a looser...

[Anonymous Coward]

To think I've been practicing every day for years to master the trumpet...what a total waste of time it has been.

I have seen something similar

[senorpoco]

There used to be a system whereby instead of blowing into a plastic tube you blew into a coiled brass or chromed tube which was flared at one end and instead of glove sensors you pressed on metal valves to change the note being played. It was quite ingenious but I can't for the life of me remember the name of the dang thing.

Re:I have seen something similar

[Anonymous Coward]

There used to be a system whereby instead of blowing into a plastic tube you blew into a coiled brass or chromed tube which was flared at one end and instead of glove sensors you pressed on metal valves to change the note being played. It was quite ingenious but I can't for the life of me remember the name of the dang thing.

It's quite clear that you're talking about the Mellophonium.

Re:I have seen something similar

[khr]

I think modern technology trumps anything as barbaric as what you're describing!

Re:I have seen something similar

[Speare]

I think modern technology trumps anything as barbaric as what you're describing!

I see what you did there.

Synthesizer

[Pikkebaas]

So this is really like an extremely complicated, barely functional synthesizer?

Re:Synthesizer

[jeyk]

Basically, yes.

Re:Synthesizer

[royallthefourth]

Actually the synthesizer is often a separate device from the controller, which is typically a keyboard or sequencer or sometimes something more novel like a specially equipped guitar.

So, they've got a new controller but the sounds are probably the same as any other synthetic instrument. I don't know what the point of this exercise is besides obvious self indulgence. The keyboard is the standard synth input because it is not only versatile for the virtuoso types, but anyone can sit down at one and make music without needing to learn loads of cumbersome technique.
Of course, there is a correct technique for playing a piano, but its enforcement is not necessary for making pleasing sounds.

Re:Synthesizer

[Anonymous Coward]

Basically, yes.

100% wrong. This device makes absolutely no sound on it's own -- it is purely a MIDI controller. A synthesizer uses oscillators to create sounds, which are then modified further by running the signal through filters and other oscillators. This glove just sends MIDI signals, which are totally noiseless until combined with some sort of sound synthesis hardware or software.

Re:Synthesizer

[MarkvW]

Dude!
The synth can be in your computer. It can be anything you want: Sampled piano, sampled drums, or synth sounds. Then there are the effects that you can apply.

The glove is just the interface. You can use the device to create your own instrument! You won't be limited by pads, a midi keyboard, or your computer keyboard.

Re:Synthesizer

[Pikkebaas]

Alright, sorry for my inaccuracy.

SO it's like a really shitty keyboard?

Re:Synthesizer

[MarkvW]

That really depends upon the sophistication of the design.
It could be shitty, or it could be really cool!

Slapstick, here we come

[Moraelin]

Aye, just wait until the kids discover how to set them for bongo, and then start slapping each other silly. It's slapstick comedy potential at its finest.

Re:Slapstick, here we come

[jeyk]

In the video from TFA, the creator mentions that they can already do snare drums and cymbals, which also has a lot of slapstick potential.

ObMovieReference

[Cyrano de Maniac]

This one time, at imaginary band camp...

Power Glove

[MonsterTrimble]

'nuff said

A better link on the same page

[iMadeGhostzilla]

Guy plays piano with his mind: http://www.gizmag.com/music-with-the-mind-brain-computer-music-interface/18489/

I would imagine that the lag between thought/intention and detection by the EEG device and would be too high but he plays some really complex and dynamic tunes.

PS. Technically even when you use your fingers, you're at the bottom of it still playing with your mind. :-)

Re:A better link on the same page

[BattleApple]

That looks impressive, but I think they're not telling the whole story. I think there are different pre-recorded passages, and the guy is just switching between them, or playing the next passage in a sequence.

Re:A better link on the same page

[iMadeGhostzilla]

Yeah on second watching I realized it's true, they even say something to that extent. Knew it was too good to be true. :-( Still it's probably satisfying for the person doing it.

I tried it last night

[larry bagina]

best skin flute ever.

So thats the 1812 overture then

[Anonymous Coward]

When Im playing my Flute

Re:I love the Imaginary Marching Band Glove

[kbrasee]

YESSSSSSSSSS

Great Idea

[b4upoo]

Although the musical artists frequently take a beating from new technology this is a tool that might come to be important for brass and woodwind people.
                          Also if someone would just build a valve cluster that is electronic and will show the score, the note being played and ask the user to quickly find the correct fingerings it would be a boon for trumpet, French horn, valve trombone. Sousaphone, Tubas and more. A three valve and four valve cluster could be made so that if the players real instrument is a three valve the fourth valve could simply be turned off. The idea is to get a metronome like response with the fingerings landing at the right moment. This would allow extending practice time in which no noise is made and also getting concentration on timing and correct fingerings. This could be a fairly inexpensive product that would aid millions of players.

Re:Great Idea

[CannonballHead]

Also if someone would just build a valve cluster that is electronic and will show the score, the note being played and ask the user to quickly find the correct fingerings it would be a boon for trumpet, French horn, valve trombone. Sousaphone, Tubas and more.

I used to do that, and still do when practicing. You sit there with your instrument and look at your music and go through it, in time, just as if you were playing, ecept you don't actually play it. You finger the valves harder than normal.

A three valve and four valve cluster could be made so that if the players real instrument is a three valve the fourth valve could simply be turned off. The idea is to get a metronome like response with the fingerings landing at the right moment.

That last bit may be somewhat novel, at any rate. Until the conductor changes pace :)

This would allow extending practice time in which no noise is made and also getting concentration on timing and correct fingerings. This could be a fairly inexpensive product that would aid millions of players.

That's already possible though. And if they already play, they already have an instrument to practice with... seems this owuld be an additional device requiring more purchasing...

No controller needed

[slickrockpete]

This system was originally invented by Harold Hill in Gary, Indiana and required no glove. It's called the "think system".

Octaves?

[captaindomon]

Don't trumpet and trombone (and most brass instruments) have multiple octaves that can only be accessed by the player changing their embouchure? How are you going to give them more than a dozen notes or so, without a sensor for their lips?

Re:Octaves?

[Beerdood]

I didn't think that was possible either, but he must have some sort of sensor built in already, because he is able to make at least two tones with none of the valves down in the demonstration video (a C, and a G). He actually appears to be blowing, so perhaps there's some sort of sensor there that bases the pitch on the speed of the air or something

Re:Octaves?

[treeves]

Exactly right. More than just different octaves. One can play the harmonic series on a brass instrument without any changes to valves that are depressed.
The harmonic series are multiples of the fundamental frequency of the horn, so the notes get closer together as you ascend the series.
(e.g. A=110Hz, A'=220Hz, E'=330Hz, A''=440Hz, C#''=550Hz , E''= 660Hz, etc.)

These notes are all accessible by the same valve combination, which means the same length horn. A change in valves just changes the length of the horn, which gives you access to a different harmonic series. (e.g. if you have an F horn, depressing the second valve makes it an E horn, depressing the first valve makes it an E flat horn, depressing the first and second valves (or just the third valve) makes it a D horn, etc.)

In the case of the French horn, which I play, the instrument plays high enough in the harmonic series that one can play a full chromatic scale just by the embouchure and right hand (which also bends the pitch as it goes into the bell to different degrees, the left hand operates the valves).
And I can cover four octaves (plus a little on a good day).

There is no way gloves can simulate the most important inputs to a brass instrument (embouchure and breath), only the less important valve combination.

Major problems...

[CannonballHead]

This, at best, will only produce one side of the (in this case, brass) instrument: the fingering (or slide position).

Any brass player knows that what you do with your embouchure is a bit important... and what about tonguing? or, basically, all articulations? And breath - how does volume work?

Seems like at best it will produce relatively boring all-sound-the-same sampled notes. I don't see how it will get the required input from the player otherwise.

Great for learning, bad for playing / bands

[Beerdood]

I've played trumpet for about 20 years or so (play in a few bands as well), so this was kind of a neat to see. I can see that it has some implications for learning the notes and technique but that's about it. Most new trumpet players (6 months or less experience) have a lot of difficultly making clear sounds or hitting the notes properly - so this device might be handy for learning scales and notes and whatnot. But at the same time, the person learning this isn't really learning how to make a proper sound on the mouthpiece - that seems to negate any sort of technical benefit you might get from this

As for a band application, I really doubt we'll ever any sort of practical use from this. Listening to the audio sample here really made me squirm (yes, I know it's a prototype in the early stages, but I can't stand a synthesized horn sound on a keyboard either). Wind and string instruments really have no business being synthesized, because it's going to be impossible (or very difficult) to reproduce all the sounds that make them awesome. Stuff like a half-valve muffled sound, slight pitch variations, vibrato, trills, tongue rolling etc - i doubt any of this is possible on a synthesized instrument this like one, or ever will be.

Just to be clear, i'm not dissing synthesized instruments in general - a synth keyboard or drums is quite useful and you make a variety of neat sounds and effects with those (I have a keyboard myself with all sort of effects). I just doubt there would ever be a useful application for any sort of synthesized wind instruments (i.e. saxophone, trombone, violin, guitar)

Re:Great for learning, bad for playing / bands

[Mindcontrolled]

I agree, but I'd like to elaborate a bit - the point of synthesizers is actually to make synthetic sounds, sounds that are not confined by the physical configuration of a particular instrument. That's where it gets interesting. Synthpiano and drums are fine, sure, if you don't want to lug a goddamn piano to your gigs, but the real power of synths is to get out of the realm of physical limitations to your instruments.

Re:Great for learning, bad for playing / bands

[TeknoHog]

I agree. It's like complaining that vegans cannot find a perfect replacement for meat, when they are not interested in the taste of dead animals at all.

Not for Woodwinds, then, eh?

[Bing Tsher E]

the prototype system uses MIDI data output from the gloves via USB to reproduce the full range of notes from instruments such as the trumpet and trombone."

The trumpet has three buttons you press. The rest is done using the mouthpiece. The trombone, you just use a slide to set the pitch.

I am fairly certain this thing doesn't do at good job at the g-sharp trill key on a clarinet.

MIDI data?

[Anonymous Coward]

PLEASE thats old atari crap!!! MP3 OR AAC DATA MUST BE DONE or its crap!

Trumpet and Trombone?

[frank_adrian314159]

More like woodwinds, if they want to do a reasonable job. Sure as heck wouldn't work for keyboard instruments.

Keyboard...

[TemporalBeing]

If only they did a keyboard...sadly they seem to have focused on only a couple brass instruments. Hardly revolutionary.

Thanks for the comments

[Anonymous Coward]

Hi everyone - I will get to answering all the questions here ASAP but definitely appreciate the interest. FYI - this definitely does not mirror only fingering/slide. It uses a pretty sensitive barometic pressure sensor by Freescale Conductors to surprisingly accurately (according to the brass players who have been working with me while I develop the project) calculate embouchure and allow glissandos and the like. I'll make sure to post a video or reveillie or taps on the site so you can see the notes change sans fingers. The plan is definitely to tackle woodwinds next, though the need for more buttons than fingers will probably mean that sax and clarinet will be constructed on the torso. Also FYI - this definitely is not intended to replace real, non-synthesized instruments. My work is really more as an inventor and game designer than a musician per se and I am really interested in the places where "playing around" (ie pantomiming a trumpet, or putting on a costume) can transcend make believe and become something "real" (in this case actual MIDI music, which though it isn't as pretty as a real trumpet can easily be printed out as sheet music and played on a trumpet, making it no less "real" compositionally). For more info, definitely check out imaginarymarchingband.com. The site's a bit sparse as the project's still in it's infancy, but I'll be putting up how-tos, videos of the other instruments, and source material pretty constantly over the next few months.

Best
scott