Ossia score – a sequencer designed to create interactive audio-visual shows | Hacker News Reader

Ossia score – a sequencer designed to create interactive audio-visual shows

github.com

51 pointsapp4soft13 comments

It would be helpful to have some examples of what is meant by "interactive audio-visual shows". Is a streaming session on OBS such a show? What interactions are supported? How does MIDI get involved here?

I could understand only after seeing this Twitter thread: https://twitter.com/jcelerie/status/1303674506951757824

Here is a video with ossia score v2.5 "in action".[0]

JFTR, There is comment by ossia score dev[1] on HN 7 month ago on another visual multimedia software thread:[2]

> If anyone's interested - I'm working on ossia score (https://ossia.io ; https://github.com/OSSIA/score) which is a bit like Max & PD, but the dataflow is integrated directly in the timeline which allows for better expressibility of evolution of time in the artwork.

[0] https://www.youtube.com/watch?v=sqIxJuKAcy8

[1] https://news.ycombinator.com/user?id=jcelerier

[2] https://news.ycombinator.com/item?id=22348463

Hello ! dev here.

Good point and thanks for the feedback ! The best there is currently is the gallery, with artists who sent us artwork they did with the software: https://ossia.io/gallery.html - theater plays, merry-go-rounds, static museum installations...

> Is a streaming session on OBS such a show?

I'd say so ? it's "show" in the general, artistic sense so it depends on the meaning you put in the word :-)

> What interactions are supported?

We tried to make something generic enough as for you to not feel restricted - so yes, MIDI, but also OSC, DMX, some people directly control servomotors & such through serial port connection, gamepads, wiimote etc.

The idea is that the timeline of the software allows you to write a score for those interactions. That is, write a score such as "when everyone finished playing part 1, move to part 2" or "when the dancer steps into the circle, make a fade out on the lights".

> How does MIDI get involved here?

You can say for instance, "when CC 35 on my keyboard is greater than 64, trigger this branch of the score".

The documentation on this website is not finished yet but we are aiming to complete it during next month.

I'd recommend watching this recording of a talk I gave at the Audio Programmer Meetup which is the most recent overview of the software: https://youtu.be/P_qT2a_2jlo?t=1394

Great! Especially thanks for provided AppImage-builds for Linux ;)

It sounds like this could be used to trigger complex lighting scenes and effects by listening in on the MIDI for a hardware-only music set, is that the case? Because, if so, this may be exactly the kind of software I've been keeping an eye out for.

> It sounds like this could be used to trigger complex lighting scenes and effects by listening in on the MIDI for a hardware-only music set, is that the case?

Yep, that's the idea - and to be honest one of the reason I've originally involved myself in this project is because I wanted to score lightings from Ableton Live with my band and that just did not work as well as I wanted it to. 6 years later the band hasn't released a single thing but at least we have our DAW now :-)

Note that the DMX editing is fairly bare-bones - my recommendation for advanced shows is to design the light scenes in e.g. QLC+ (https://www.qlcplus.org/) and then control them from ossia score at a higher level through OSC messages. But it's possible to bang raw ArtNet bytes if you want - we had a small exhibit doing just that a couple days ago.

This looks really similar to Hyperproduction[1], wondering if there's any connections or relation between the two projects?

[1]: https://ben.ai/hyperproduction/

Is there anywhere where one can try this out?

Hello! Hyperproduction was my MS thesis. I can answer questions if you'd like. It's a fairly simple platform based on Node/Electron and focuses only on control automation rather than any type of "renderer" such as audio or video outputs. Furthermore, it was primarily developed in rehearsals for various productions, so it's quite rough around the edges and serves as a working toolkit for a few of us who find ourselves doing a lot of interactive performance design. Looking at Ossia, I think a fundamental philosophical difference is that Hyperproduction focuses less on being a "score" and more on mapping and completely flexible interconnection. It's a simple forward propagation network. Although there are many processing blocks, there isn't currently any enforced typing along "noodles," which means that it's a somewhat difficult environment to train new users on. You need to have a basic understanding of what all the blocks do and what types of input and output they take before really getting going.

If you would like to see some of the productions we put together, you can check out my dissertation here: https://ben.ai/research

Hi!

Well, yes, ossia is initially based on a score paradigm, but our main dev @jcelerier added a node-base system recently, which kind of turns it to a node/patching system for mapping and interconnection. However, as I said, we don't nearly have such a variety of processing blocks yet - there is a clear API to make new ones, though.

For those interested in such mappings systems, there is also http://libmapper.github.io/ that might be of some interest...

Hi! Another member of the ossia crew here: so, no, there's no connection between both projects (so far). However, even if both projects seem pretty similar at first sight, they might be quite different (I say "might", because I didn't find any way to download or try hyperproduction). ossia score is initially oriented towards managing temporal behaviors and evolutions (aka "interactive scenarios") whereas hyperproduction seems more oriented towards creating and managing conditioning and mapping of control data and signal. ossia score can also do (some of) this, but so far hyperproduction seems to have a larger base of data-processing modules. Maybe we should try and see if both can be connected/made interoperable ?

We need interactive media that is a movie you watch and a camera tracks your eyes and bodiku reactions and can dynamically change the story and portrayal to try the elicit the maximum response along some metric from the viewer

Post such job proposition in "Jobs".[0]

[0] https://news.ycombinator.com/jobs

Reply on news.ycombinator.com