Follow

It blows my mind that the standard way to sequence a string quartet basically involves getting a computer to believe they are a piano believing they are a stringed instrument.

There's at least one step that should be removed from that process.

@paul like, writing for a MIDI module or something?

@hecanjog like an orchestral mock-up with an emphasis on fidelity.

@paul ah right on. yeah, it's weird. I'm working with the east/west library now doing some fake string stuff. Every articulation and playing style is a different instrument, so to have some kind of fluidity of gesture for anything that changes a bit, you have to load like 5+ instruments and spread the score across as many channels. It's awkward AF...

@hecanjog
Right? It's a weird gesture translation.

Imagine starting with a virtual bow and fingerboard control interface, and building up from there. I think that would be an interesting approach.

@paul totally, starting from the perspective of the physical object is probably the only sensible approach if you want to go any further than a cartoon representation... Looking at Helmut Lachenmann’s string writing might be inspiring to that end, his approach goes well beyond the piano cartoon and makes use of the entire instrument from its properties as an object rather than a general timbre smooshed into note form...

@hecanjog
I need to start building a physical reading list from all the recs you give me...

@paul I’m out of touch with this world, likely there are young people now taking these ideas even further but I really love Lachenmann. This is some fantastic string writing: https://m.youtube.com/watch?v=GZt3UfNDNV0

@paul also I hope it's not obnoxious when I barf citations at you like that... in my defense you tend to toot interesting ideas about things I am interested in 😛

@hecanjog quite the contrary! I sometimes make posts in the hopes you'll have something interesting to share.

the bigger issue for me is keeping track of it all and actually reading it.

Perhaps I can start storing some links/references using twtxt and tags. My wiki system would then be able to automatically aggregate them in wiki pages.

@paul it mirrors the traditional workflow of composers using the piano to compose and writing down as sheet music, then giving this sheet music to the string quartet to play it. The reason it looks like this now is _because_ steps were removed/mashed together, not the other way around.

@electret Yes, I absolutely understand the craft and tradition. My interest is exploring alternative workflows that could potentially yield better sounding performances and better leverage the capabilities of computers. Sort of something along the lines of Bret Victor's "Stop Drawing Dead Fish", but with music.

@paul the part that surprise me here is that the end goal is "sequence a string quartet". If you want to leverage the capabilities of computers, why not synthesising the sound from scratch? Also, what is "better sounding"? What capabilities do you have in mind? (for context, I'm an experimental musician -> my questions are out of genuine interest, not trying to shut down the idea!)

@electret well, the string quartet is the gateway idea. It's really about challenging ideas of what it means to compose music on a computer, but through re-interpreting old ideas rather than throwing them outright. As a string player going through an electronic music program, it was hard not to feel the keyboard bias in the industry tools we were taught how to use.

@electret String quartets have some interesting properties that are generally interesting to explore. Highly limited voices that are highly heterogeneous. Limited multi-mono polyphony in this manner is a much more computer-friendly thing to model than the massive polyphony of a piano.

The base characteristics of a string instrument are also quite unique too. They are actually more continuous than discrete instruments. The bow is an incredibly expressive device, capable of smoothly going from legato to a staccato sound. There's a whole world there in articulation patterns alone. Then there's the fretless nature of the instrument things like vibrato. A performer's natural vibrato is a very personal thing.

All of this of course ties to the true instrument of what we call Western Music, which is the human singing voice. Everything mimics that thing. We just went with keyboards because they were the best simulators until computers were invented.

@paul there might be two interconnected problems here: the problem of input, and the problem of simulation.

I think the problem of simulation is decently solved (not perfect, but enough to have some realist results).

The problem of input is also a problem of competence. The richer the data is, the more complex it is to be produced.

To have a ground breaking solution you'd need to have an input solution that close enough of the skill gap that it can be used for various types of instruments.

@electret It's not exactly about fidelity for me though. Nor is it about solving problems. It's more along the lines of asking myself things like "why does this string quartet tug at my heartstrings", then trying to hobble together unique computer-generated sounds and structures inspired by the answers I come up with.

@electret Oh, I'm also not thinking about *sequencing* for samplers and synthesizers in a high-quality orchestral mock-up, not necessarily making sheet music for human performers.

Because MIDI is so hardwired to be a keyboard (written by/for keyboardists), it's really easy for synthesized string instrument performances (step-sequenced or recorded live) to sound keyboard-y. A lot of work and craft that goes into professional orchestral mockups basically goes into massaging keyboard-centric performances and representations of music into other instruments.

My thought experiment would be: how would you sequence a realistic sounding string performance on the computer without being proficient at keyboard? What would be the best way to do it if MIDI didn't exist?

And that's the particular cloud my head is in today. Apologies for the word flood.

@paul before learning enough keyboard skills, I just mouse clicked my way to sequence what I needed (still do sometimes). But I don't think that's the answer you're looking for :D
Unfortunately I'm too removed from "realistic sounding composition" to have an interesting answer.

The kind of questions I have floating these days are more around "how would I make the kind of music I make, if computer didn't exist?"

@electret I was a point-and-clicker too myself before I got onto trackers. this starts to leer into HCI territory, at which point I begin to think about the nature of computer mice in music. That's another cloud I sometimes put my head through...

"what would I do without a computer" is actually great way to think too. Either way, I think it's trying to tell the story of "what exactly are these computers doing for us".

@paul I've always loved FamiTracker because it's effects system allowed me to use it like a string instrument. In DAWs, similar effects would probably require a bunch of automation curves.

I personally think that violin playing, with all of it's techniques, would be best represented with some sort of description language. If my computer is performing a double stop, I want it to know exactly how it should be played.

Sign in to participate in the conversation
post.lurk.org

Welcome to post.lurk.org, an instance for discussions around cultural freedom, experimental, new media art, net and computational culture, and things like that.