Or do you practice at all ? Do you start from scratch or with prepared code ? Do you keep the results ? Do you record your sessions ?
I try to get into a routine of doing one session without prepared code each day, record it, and listen to it a few times over the next days. I think that helps. But, as it is common I guess, I tend to get distracted by improving my tools, rather than using them.
The worst thing for me is recording screencasts … having your sub-par tying skills immortalized in a video throws me off the track way too easily.
I always keep the code versioned with git.
Often I record my sessions, but rarely listen to them.
I practice mostly to try freshly recorded samples or to experiment just learned features or to apply some ideas that came to mind.
Right now I’m practicing towards an hour-long solo algorave set at a
festival. As @parkellipsen says, what ends up happening is that spend
too much time tinkering with my setup and not enough time making music
and listening back to it! To combat this, I’ve been setting myself the
task of playing non-stop for half an hour at a time, only allowed to
fiddle with stuff under the hood between times.
I love to practice! (But I haven’t built my own tools so I don’t have that distraction).
For me I don’t pre-prepare any code, but I like to try out different sounds and have a rough ‘palette’ of sounds that I know work together.
I also don’t really practice a full set but will experiment and try things out to see how different things sound together, maybe try some syntax I’m not too sure of or try out some weird combinations of stuff.
I basically use practice as a time to get inspired so I can feel confident in my ideas in a live setting
This is such an important topic and all the replies give valuable insight!
I use Sonic Pi for live coding and for me, one thing that I like to do is write functions that I can use in my performances. So when I practice and find something cool that I think I’ll want to reuse on stage I’ll usually make a function from that block of code and design it so that I can reuse it in different ways by changing the sounds or the speed or whatever parameter I find interesting.
This gives me a nice middle ground between coding everything from scratch and preparing all the code beforehand. I can improvise while minimizing the amount of code I need to write live on stage.
I definitely practice a lot when I have a gig coming up! Also, like @tedthetrumpet, I try to set aside time to just play my set and not fiddle with functions. I was just re-reading an article from Sam Aaron and Simon Blackwell (Craft practices of language I think) and Sam has a nice reflection on this.
I also try to separate out phases between warm-ups (typing basic building blocks of my set and keyboard warmups), free experimentation, composition, and more concrete rehearsals.
