[shufflebeat]

I've been following the development of IR technology as a musician with a healthy obsession with(other people's) good ideas. As we've gone along together I've constructed my own way of understanding the process which may well be complete nonsense but here goes.

Imagine you're in a tiled bathroom where your fiddle sounds excellent, reflections bouncing around and coming at you from all directions. You may or may not be a fiddle player, I'm not. You would love to take this acoustic response to all your gigs but that would be impractical.

Really, you don't want to take the tiles, bricks and mirrors, you just want the effect they have on sounds you introduce in the space.

If you could add the mono or stereo reverb tail to every sample in a digital recorded signal then the end product would be a bit like the original sustained sound in the space.

So, you take a balloon and a recording device into the bathroom and record the sound of the balloon being popped and the resulting reverb tail.

Because the pop is such a short sound (an acoustic "impulse", so to speak) there isn't much of the original in the reverb tail, just what has reflected round the space and back to the mics, early reflections, standing waves and all (the "response" of the space to the impulse, do to speak).

So you go to your PC and edit the original "pop" off the recording and save the result. Now if you substitute the pop for a studio recording of the fiddle through the Impulse Response recording sample by sample you will synthesise the response to the fiddle using the response to the pop, crunch the resulting numbers and convert to analogue sound - Viola!

Now do the same by tapping your guitar with a spoon (or better still, someone else's guitar) and record what comes after.

Somebody may come with a much more technically correct explanation but this one has worked quite well for me so far.