I’m creating a composition for string octet and want to spatialize the spectral information of the strings
with a realtime fft patch and multi channel surround speakers, such that it as much as possible gives the audience the impression of being in the construction of the timbre space.

I’m interested in hearing from people who’ve tried similar ideas and how successful they were in their
approaches. Also, what kind of multi channel approach would be most suitable.
Apologies if this is not too Max specific enough but as I’m contemplating achieving this through Max
I thought the forum would be a good place to make enquiries.

Thanking in advance for advice / information…

do the speakers reproduce the string sounds right away but split one audio signal into 4 or 8 or something?

I’ve done work that creates the impression of an immersive sound object, but I use granular freezes, reverb and buffer schuffler type things along with filters.

though you can use ffts to rearrange the spectral information itself, like inverting the lows and highs and so on.

I know this doesn’t apply to realtime sound but you should read this article by Charles Bascou from GMEM Marseille.
He obtained good results with a kind of timbre / space position correlation.

http://start1g.ovh.net/~gmemzeuq/_media/_dev/_publi/ADAPT-Holo-Edit-.pdf

http://gmem.org/index.php?option=com_content&view=article&id=315&Itemid=305

Léo

I’m looking to aim for something resembling tonal fusion. But in order to parse out the spectral content towards this goal I’m probably looking at wavefield synthesis or ambisonics.
So if any composers have tried this or similar approaches do let me know.

Thanks for the replies so far.