À Clermont-Ferrand ce week-end se tient la compétition vidéo du festival Vidéoformes. Mon (modeste) film « Ensemble, ils dérivèrent » y sera projeté dimanche après-midi. Le reste de la programmation me semble très prometteuse, j'ai hâte !

In a somewhat recent Tsoding stream, I've heard about venation pattern algorithms, and tried to implement one (Runions et al.). You may try it yourself at chalier.fr/generative-a...

Inspired by Actinomy's server charm manifesto, I've tried to build my own. Very pleasant experience, got to learn FreeCAD, how to solder things. If you want to know what's on it, you'll have to get near me! To build your own: codeberg.org/actinomy/ser...

I've been experimenting a bit with Sharp, Apple's open model for generating 3D models from single photos. It's quite good, and certainly allows for various cool effects. Here are in an example rendered with Blender and the 3DGS Render add-on by the Kiri Engine team.

Turing patterns emerging from a repeated blur-sharpen process, a WebGL implementation: chalier.fr/generative-a...

Quelques effets maison dans ce clip ! www.youtube.com/watch?v=YOiN...

Inspired by a recent Spectral Collective video published on 3B1B's channel, I've implemented a simple version of the liquid-vapor model: chalier.fr/generative-a... Move the mouse to change params, scroll for speed, left/right click to add/remove particles, press S/R to screenshot/record the screen.

Très cool à suivre (en replay malheureusement) ! Je partage à nouveau une implémentation rapide sous WebGL de ces bassins de Wada : chalier.fr/generative-a...

Lately, I've been refactoring and cleaning up my FFtools repository: github.com/ychalier/fft...

A few weeks ago, I published the version 1.11 of Transflow, which generalizes the way pixel sources behave with a nice 'Compositor' framework. Mainly, Transflow should now be easier to use, even for more advanced effects. I've also updated the documentation with more examples, for you to try it out.

I've just pushed the v1.10 of Transflow. Not a lot a new features, I've tried to clean up the code and make it more structured: github.com/ychalier/tra... The next update should also be a refactoring, I would like to unify the multiple accumulation methods into a cleaner, general class. Stay tuned!

Depuis quelques temps, je me suis remis à jouer aux échecs et je fais régulièrement des parties sur l'appli @chess.com qui permet de trouver des adversaires en direct un peu partout dans le monde. Or j'ai remarqué un truc :

Also, due to the current heatwaves here in France, I'm turning off my web server during heat peaks, I would not want to fry it. Sorry for the inconvenience!

Just released v1.9.0 of Transflow, with a brand new GUI! If you're afraid of command line interfaces, this should make your life easier. Beware, though, this is still experimental, bugs may/will arise.

I've just added an audio reactive feature to my randomart webapp. In short, you can use specific variable names in the grammar to refer to the (normalized) intensity of ranges of the frequency spectrum (e.g. low, mid & hi). Here's a sneakpeak: 🎵 Mega Hyper Ultrastorm (Kevin MacLeod, 2017)

2y ago, I built a tool to generate halftone images, inspired by a paper about halftone QR codes. Lately, I've wanted to use this technique to embed a QR code within an halftone image. I found an old repo on GitHub, refactored it and added my scripts: github.com/ychalier/hal...

Je partage mon implémentation (faite un peu à la va-vite), réalisée pour WebGL (rapide et à peu près portable !) : chalier.fr/generative-a...

I tweaked Transflow to replicate the 'sticky texture' effect of regular datamoshing. The code is a bit dirty but the result is nice! Here are three clips pasted together:

Just learned about LiDAR HD data by IGN, a point cloud mapping of most of surface of France. It's really cool, and freely available! Visualization made with QGIS 3D Viewer.

Motion Interference. Applying the reversed flow on the same video makes pixels relatively static while revealing some of the transformations – e.g. through perspective change or flow inaccuracies. Thought for later: is there a motion equivalent of the Moiré effect? Below, view from a moving train:

Been trying out Deforum for generating videos with Stable Diffusion. It's great at textured and abstract drawings. (Side note: I need a better GPU)

I've been playing around with a public webcam recording 30s clips every 5min of a pedestrian area, and blending those clips together. Here's an example with an (about) 8 hours span (94 clips). Also, you may try to find that place, it should be pretty easy!

Same theme, different domain: www.youtube.com/watch?v=EmPZ...

Recently, I've enhanced the Transflow web demo and added a few features, such as thresholding, iterative Horn-Schunck method or screen recording.

Well, my FlowNet2 integration attempt was not successful: I had a lot of issues with installing dependencies and finding compatible module versions. However, I managed to integrate LiteFlowNet, a more recent alternative, which also provides very high quality results!

I tried to use a MPEG-4 Part 2 video codec (Xvid) instead of h264 for my datamoshing script that removes I-Frames from the bitstream. Results are much more interesting! h264 really was too resilient! and frames would heal instantly. Now they're crushed, as they should be.

Transflow update v1.6! Lately, I've been working on new features, mainly to support other optical flow algorithms. Then, the only option was Farnebäck's algorithm, OpenCV's solution for dense flow fields. I've now added support for Lukas-Kanade and Horn-Schunck methods.

Been working on a new feature for Transflow, to provide a way to move things more progressively and avoid unpleasant blinking. In the example below, new pixels are introduced in the rightmost column, and then moved around with the optical flow. Will release it soon with other minor fixes!

I learned that Audacity can be somewhat automate with the mod-script-pipe module, so I tried to produce frame-by-frame moshed videos with it. I added it to my datamoshing repo on GitHub: github.com/ychalier/dat...

In last Transflow version, I added an automatic upscaling feature for using small video inputs. E.g. a 128*72 flow can be applied to a 1920*1080 image by counting 15*15 blocks. Besides being faster, it also gets closer to the original datamoshing look and feel, which mostly relies on macroblocks.

Recently, I stumbled upon a video by Japhy Riddle called 'Making Audio Waves That Look and Sound Like Things': www.youtube.com/watch?v=qeUA...

This open source framework for generating procedural 3D scenes developed by the Princeton Vision & Learning Lab, “Infinigen” infinigen.org , seems almost too good to be true. www.youtube.com/watch?v=6tgs...

For anyone also interested in databending/datamoshing: a great starting tool is audacity. Save an image as a .bmp or .tif Use audacity to import raw data, u-law, little-endian Leave the first 3 or so % alone Mess with 'Filter Curve EQ' img1: Low end img2: middle img3: High end

I recently heard about granular synthesis and I wanted to experiment with it. I've just read about the principle and tried a basic implementation, it's fun! I heard that Microsound by Curtis Roads was a good introduction to the principle, so I guess I'll read that now.

The tool is available in Beatviewer v1.1.0! bsky.app/profile/chal... github.com/ychalier/bea...

Someone recently incited me into adding a video rendering feature for a past projet, Beatviewer. You give it an audio track, and a video, and it edits it automaticlly. Here's a sneak peak with a slow-down effect (better experienced with sound 🔊):

Transflow v1.3.0 with a new feature: input repetition! github.com/ychalier/tra...

I tried to apply different convolution kernels to the computed optical flow, and see how it impacts the result. Most interesting results came from using the gradient to directionnaly select the movement.

When I thought of doing transitions as in this example bsky.app/profile/chal..., I wanted to try a new way of moving pixels, so they leave an empty spot behind them (which could later be covered by another moving pixel). Here comes the Crumble Accumulator, for Transflow v1.2.0!

Just released v1.1.0 for Transflow, with the recent changes mentionned here: support for multiple flow sources, better gradient algorithm, and controlled bitmap alteration. Here is a recording of the alteration GUI tool:

I've changed the algorithm used in Transflow for generating gradients. The previous hacky solution would generate blobs and blur them, which was neither efficient nor pretty. The new approach is based on a simplified version of Randomart. It's faster and a lot nicer!

I wanted to try controlling the output of Transflow renderings. It appears that with enough motion, the output is only determined by a couple of pixels. This means that subtly altering the input is often enough. E.g. this deer image contains only 80 pink pixels, but that's all what matters.

First post ✨ to mention a new feature for my optical flow transfer tool (Transflow): you can now merge multiple flow sources and control how they interact. I've pushed my changes to github.com/ychalier/tra...