From the conceptual to the concrete, by the end of the workshop you will have both worked with abstract concepts in the textport and created a functioning tool for saving presets. From using storage to writing your own extensions we’ll work through the several concepts that will help you better leverage Python in TouchDesigner for installations and events. This is the exact kind of process I follow if I’m confronted with if I have to implement something in TouchDesigner that isn’t already in the collective common knowledge of TouchDesigner developers.īuilding off of previous Python workshops, this class aims to demystify a few of the elements often used when doing advanced Python development work in TouchDesigner. And it starts with simple steps like reading articles and making notes. Being able to have the patience and process to take other platforms tutorials and translating them into the steps we need to implement something in TouchDesigner is invaluable.
More often than not, if you’re looking to implement something in TouchDesigner that hasn’t already been explored by other pros, the only resources you’re going to find are going to be made for other platforms. Our TouchDesigner community is still comparatively small when compared to Unreal or Unity communities. Then we should be in the ballpark of a cool effect! Now while this may seem a bit pedantic and needlessly “documentation-filling” but there’s an important skill to practice here. Add some noise per block (what would we do without noise?!).Quantize the sampling blocks of our movement vector to create a blocky effect.Add a mechanism to blend between unprocessed and processed pixels.Find a way to recycle the pixels on screen / don’t clear screen every frame.Displace our colour image sampling UV position.Get/create a set of motion vectors for movement in the image.So our first step to implementing something like this is to break down their article into some key steps:
KANYE WEST DATAMOSH CODE
If you’re new to GLSL, don’t worry, in the next blog post we’ll explain all the lines of code we’ll be writing. In our implementation we’re going to use a mix of GLSL and some pre-built tools in TouchDesigner to make the magic happen. While we can’t use everything from the Unity implementation, the author does a great job of walking us through the high level concepts of what needs to be happening. The both show how in many cases a video will actually have different types of frames holding different amounts of information:Ĭreating Your Own Datamoshing Effect Thanks to my good friend, Landon Thomas, who found me this write up Here are two images that show these concepts at a high level just for visualization purposes. That’s why files with not much motion in them (most of the frames are the same) can be compressed to very small sizes while textures like noise or grainy videos (where lots of information per frame is changing) have a tough time being compressed to smaller sizes. Without getting into the weeds, the important thing to note is that many compression algorithms try to only hold the amount of data in each frame that has changed since the last frame. If you’re reading about datamoshing online you’ll often hear about I-Frames, P-Frames, and B-Frames.
KANYE WEST DATAMOSH FULL
Not all frames in the video are actually full frames of data. Datamoshing effects in their original appearance are created by issues between the different types of frames in compressed videos. Some effects are easier to brute force your way to intuitively, but others require a bit of reading. Half of the battle of recreating any effect is knowing how it works. Now it’s a creative effect that has a lot of potential! How does it work? You may have seen effects like this before if you had a corrupt video file or low powered hardware having a tough time decoding heavily compressed videos.
0 kommentar(er)
