Scientists Create Air Guitar T-shirt

onco_p53 writes "Australian scientists have invented a T-shirt that allows air guitarists to play actual music as they strum the air. The shirt has sensors in each elbow and sleeves to detect and interpret the air guitarist's arm movements — one arm chooses chords and the other strums imaginary strings. The gestures are then connected wirelessly to guitar audio samples to generate the music. Check out the video!"

Comments

Major Vs Minor

[eldavojohn]

I'll probably be modded as flamebait but I don't think that shirt was actually controlling that sound. I think this video was intended to be a viral internet video to garner venture capital for developing that kind of technology.

I assume this is intended to only play power chords and on top of that there is no way the machine can determine if you intend the third to be major or minor in the chord. Which is interesting because in the video, the chord changes from major to minor with no change in the performer. That's fine, let's then assume that you can select a key and it will adhere to the chords in that key.

At several points in the video, he strums by flicking his wrist instead of moving his whole arm. That's pretty standard for guitarists but doesn't explain how the shirt (with sensor fibres in the elbow) would recognize the motion. Also, near the end, the guitar that I assume he is playing strums multiple times without him doing anything.

I read the article and, if it has been developed, this short short video did not do a good job of showing it off or selling me on it actually working. I'd rather see an average guy just messing around with it with no back track. I don't care if it's not perfect, it's just that I could make that video with crappy acid trip effects in my basement and my friend on guitar watching me move my arms.

Re:Major Vs Minor

[jimstapleton]

I"m not saying I completely trust this either, but:

At several points in the video, he strums by flicking his wrist instead of moving his whole arm. That's pretty standard for guitarists but doesn't explain how the shirt (with sensor fibres in the elbow) would recognize the motion. Also, near the end, the guitar that I assume he is playing strums multiple times without him doing anything.

Grab your right elbow with your left hand, flick your left wrist. There is movement there, and it is different than when you movce your whole arm. It's possible to differentiate between the two I suspect, maybe not easy, but possible.

Defeats the Purpose

[organgtool]

You still need to know how to perform the correct movements in order to create something that sounds close to the original song. The whole purpose of air guitar is that you don't need to know anything about playing guitar and any mistakes you make aren't usually apparent.

Wouldn't it have been easier to...

[Anonymous Coward]

just buy a damn Strat and a Marshall, and learn the A, G and D chords?

You wanna be guitar heroes, and you don't even wanna put the work in on it.

http://www.youtube.com/watch?v=wBiJ-K0IpDA

But won't it still sound bad?

[jbarr]

Isn't the point of Air Guitar that someone who can't play guitar mimics someone who can?

The 1990's called...

[cr0sh]

...and they want their VPL DataGlove back...

Alright - I realize this isn't a glove interface, but what is it with geeks, "virtual" environments (including those in the head), and air guitar? Arguably, from everything I have read about the history of VPL, the DataGlove's first raison-d'etra (probably munged that!) was so that Jaron Lanier (and a buddy?) could play "air guitar" using their Atari ST and MIDI (google around on "VPL" and "DataGlove" if you don't believe me). Since that time, tons of people, companies, and bands have experimented in one form or another to bring the "air guitar" to life.

I don't play the guitar, nor do I know how, and I won't say I have strummed a bit in the air, too - everyone has, I think. However, I know that if I really wanted to play guitar, no amount of movements "in air" will ever be able to recreate what happens with a real guitar. At best, I will get an approximation of (likely) samples of somebody else's real performance.

Finally, I must admit I haven't read the article - but I would be willing to bet that there are 50/50 odds that the article mentions how the technology could be used to "read" sign language by monitoring the movements of the arms (yet another thing the DataGlove was touted as a useful thing to do with it, and yet another that gets hauled out year-after-year).

I guess what I am tired of is all of this rehash of stuff we already know how to do - let's quit playing and demonstrating dead-end uses for VR (and AR) technology, and let's start using it for real world tasks!