A video synth patch I put together last night with some audio made in vcv rack.
A video synth patch I put together last night with some audio made in vcv rack.
I cleaned up my OBS configs and got a video recording. It's all needlessly complicated.
The visuals are kinda being controlled by midi via *deep breath* TidalCycles midi > Univer Inter midi to CV module > 3trinsRGB video synth > video output to capture card > composited under TidalCycles code in OBS.
I've owned most of my video synth for two years but every few months I reread the instruction manuals for different modules and discover something new. In DWO3 I finally now understand what the RST input does.
In this video I mostly combining DWO3 with Solaire and DSG3.
Also adding in that recent VCV patch for audio.
Last night I took out the video synth and tried animating some of my old blender renders. I might be onto something
The new year got me started on a small creative-coding project, which is a multitrack A/V compositor. It's not meant to be scalable or high performance, rather the idea is that the type of video I'd want to make with it is something that Love2D could easily do with a fast turnaround - karaoke text, looped animations, some procedural stuff - so why not make an editor out of that, made for that, instead of trying to cobble together other tools.
My inspiration for this is "Cartoon Cat Karaoke" - so simple, so effective, I think of it a few times a year
https://www.lexaloffle.com/bbs/?pid=31029
In three hours I bashed out 320 lines of code which do all of this:
Play a WAV music track, displaying what frame it corresponds to
Move a playhead through the track and loop or not loop it
Mark events on each frame on one of eight tracks, one event per frame per track, each event is "just a string"
Save and load the marks to a text file("lousy CSV" since it uses the first two commas semantically and consumes the rest of the line as an event string)
Select mark data from a palette and auto-advance through the palette so that you can time all the lyrics in order.
Cache and compute the spans between each mark so that you can do bouncing-ball text, etc.
The secret sauce of the output, of course, is that the mark data and spans are just going to go straight into hardcoded rendering that triggers effects and animations. Getting rid of formal structure and letting it be very messy and global allows everything as a possibility - wacky state-variable physics stuff, triggering samples over the main track, custom parse strings for one-off effects, and so forth. I dub this approach "A/V jamming" since it's meant to proceed roughly at the pace of a game jam - still edited and curated, but also performed, improvised and bespoken in many respects.
#gamedev #musicvideo #videosynthesis
I found this video on my computer of a anti piracy campaign from 1992 and decided to run it through my video synthesiser. Its slogan "Don't copy that floppy" warned that if piracy continues creative industries will fail.
Fast forward to 2025 and tech bros are being given free reign to steal from creators to feed their AI slop machines and are being hailed as innovators. Clearly it's only stealing if we do it.
Might go copy me a floppy or two.