Interesting. It seems like you want a value judgement capable of marking an axis between live coding and live arrangement (DJing, triggering clips in Ableton, triggering pre-written functions in SC, etc.). Is it a question of the behavioral complexity the performer is making use of (limited of course by what the software supports)? Would you be looking for a metric that can determine to what extent a piece is authored "live"? Is it a product of both of these factors, or more? I'm guessing that nobody would suggest analysing the resultant sounds (sample time-series) to measure it, but where would you measure it? Translate the code into a neutral representation of the algorithm (as a time series) to compare performances?
A related question: do we care whether a performer authors a reverb live or just uses one that already exists? Is it only interesting if the reverb is significantly (and/or perceptually?) different from known or recognizable reverbs (relative to audience familiarity?) (The same could go for tropes of musical structure). So then, is it more about improvisational novelty? A product of algorithmic complexity, degree of live-authorship, as well as the creative novelty?
(BTW, why not point & click? I could point-and-click a recursive function, for loop, coroutine whatever. I could move a bunch of stones on an augmented-reality table to define a recursive function etc. Maybe one interface is faster for some things and slower for others, more or less suggestive to the performer or prone to error, but text vs. non-text doesn't seem to be the interesting distinction.)
On Aug 11, 2013, at 9:25 AM, alln4tural-list_at_gmx.net wrote:
> At 20:05 10.08.2013 +0100, alex wrote:
>
>> I think of live coding as improvisation. I make sure things are live
>> by writing some new functions in advance of a live coding performance,
>> to create new space to explore. Otherwise it can just feel like going
>> through pre-practised 'licks'.
>>
>> From the original post:
>>> i myself use SC. i've spent a lot of time, though never near enough, poking around
>>> the docs and listserv and built a collection of tools that i can invoke with very
>>> abbreviated code. over time i've collected a little library of such tools and
>>> accompanying code snippets, and when i sit down to make noise, i'm mostly sitting
>>> in front of a page full of these snippets, "just" deciding which of them to execute
>>> when, and the only code i write in situ, if any, is in the form of small routines, made
>>> with copy and paste, that endlessly loop over a couple of these code snippets.
>>> is that live coding?
>>
>> I wouldn't call it live coding. If you're not changing the code, by
>> combining and abstracting things, then it's not programming.
>
>
> sounds reasonable, but doesn't that mean you can only tell after the fact whether this has been a live coding session or .. a set of licks?
>
>
>
Received on Mon Aug 12 2013 - 19:17:45 BST