Hnns Hlgr Rtz @sciss@octodon.social
Sound Artist, Composer, Researcher
http://www.sciss.de
Hey #Caturday - it's time to #OccupyHanns
After a week of extraordinary sun, #Bergen is back to the weather we've been warned of :-O Still having lots of fun, today is last workshop day.
#Körper (1) is up and running at #Lydgalleriet Bergen - in the next days I will post some photos of all the other pieces that make up 'Thresholds of the Algorithmic'.
http://bek.no/en/thresholds-of-the-algorithmic/
Now busy preparing our workshop that begins tomorrow.
We're prototyping the sound interaction with #Körper now (next week Friday it will be shown in #Bergen).
Yesterday I was sceptic whether moving my experimentation from IntelliJ into a Mellite, which meant restructuring lots of glue code and introducing many constraints, was good idea. Now that everything indeed runs in Mellite (I have simple view wrapper for Sphere GNG, as shown in screen shot), I am confident it was the right decision. I can now tweak parameters while installation is running.
SphereGNG working finally after lots of bug chasing 😑
We had fun yesterday at #MinuteConcerts - I'm waiting for the copy of the recording to get a better impression, but I sensed people liked it. #AnemoneActiniaria
It's that time of the year again
#Graz tomorrow 11pm at #ForumStadtpark - #Anemone Actiniaria performance for impuls #MinuteConcerts festival. We're gonna rave!
Fiddling around with the monitor mount; I would improve the construction if I did the mold again, but for now it should be "ok". Just have to find a good way to hide two corners of the frame. (Video image is from other project, just for testing).
Poor man's parallel programming, I just split the batch and ran the shit three times. (The forth core is for tooting)
Now using stochastic sampling - much faster, and actually nicer. Only again something weird is happening somewhere with the spherical coordinates, but perhaps I leave it like this, as the scattering is nice.
Some funny things happening when playing back the spectrally zoomed constant Q transformed doppler signal mapped to voronoi coodinates (in other words - what's the bloody perceptual link here?)
Something went wrong with those trigonometrics (the sphere is supposed to rotate)
This but at higher resolution, and using an actual Voronoi to cover a 1/5th of a sphere.
I was ignorant for many years – only just now did I realise why these are called #Cat cables
Through out these PNGs to test the correctness of a new RotateFlip UGen in FScape; out of laziness wrote them to the home directory, now every time I open the file manager, I see these colours, and that makes me happy LOL
Some beautiful Doppler analysis; I added the Constant-Q filter to FScape, currently in the SNAPSHOT of the work branch (new release soon) - the README has more or less the corresponding code: https://github.com/Sciss/FScape-next/tree/work
So I can clearly see the Doppler shifts, but those spikes are really annoying. I tested multiple channels and cable now, it's always the same. Will try a different audio interface today, but I'm not convinced that it would be the source of this behaviour...
I'm getting some strange crackle in my ultra-sound signal, with a spacing of around 130 to 190 ms. I checked all sorts of things and can't find the culprit. It must be either an electrical problem or something of the DAC/ADC. Anyone have any ideas...?
(Depicted is 40 kHz signal, where the disruptions are visible).
Experimenting with GNG-U again, using an old video of mine. Figuring out how to create the particular texture I'm looking after.
