Entagma
Unreal Engine Blueprints - Pt.2: Physics
updated
00:00 Intro
00:53 Normal Maps and Coordinate Systems
03:57 Setup Preview Texture
05:00 Prep Work on the geometry
07:27 Modifying the Normal
10:58 Matrices!
13:23 Blend the Maps
14:30 Thank you :)
Links to the mentioned Videos:
youtu.be/5z18ZJaCUOM
youtu.be/Cq5H59G-DHI
youtu.be/xPW9UOhsNnI
youtu.be/dn5zNrH_cFQ
You also can download the Scene file for free on:
entagma.com
And for more free Tutorials, consider becoming a Patreon:
patreon.com/entagma
We are thankful for all of your support! Every like, comment or subscription helps a ton!
Thank you guys!
0:00 Intro and Credits
1:08 Prepping for Mapping
3:05 Setup the UV Maps
4:44 Fix UV Problem
6:51 Creating the Masks
10:39 Writing Code for Blending
12:55 Make Masks sharper
13:52 Add Noise
14:24 Combine into one Texture
15:55 Improve the Setup
17:20 Remove artifacts
20:18 Thank you :)
You wished for this tutorial, so Chris immersed himself in the topic to create these free Tutorials for you Guys!
Watch it on Entagma.com and Youtube on Monday, the 16th!
See you there and give Chris some love in the comments!
Music by Oncys - Eden
Watch it on Entagma now!
This week you get a crashcourse in setting up your VS Code and Python and then we get started with coding!
Follow us, we will be more active now and notify yo when a video comes out!
See you there! Cheers!
Give him some love over on his channel:
youtube.com/@lueyyys
Download the Project File:
entagma.com/downloads/GameOfLife%20Upload.blend
1. Go to www.patreon.com/entagma
2. Support us at the "In-Depth-Knowledge" tier or higher
3. Go to entagma.com and log in with your patreon account to unlock all Premium Courses
This course is about making Machine Learning the next tool in your Houdini toolbox. Just like some problems are best solved with vector maths or Vex or Python, some problems get way easier with ML and we want to be able to spot these problems and solve them using your own custom networks.
To do so we create our own training data in Houdini, we build and train our own neural nets with Python and PyTorch and finally integrate them into our Houdini Workflows using ONNX. Additionally we want to take really deep looks into how these setups work and why they’re built the way they are to get a ton of intuition for applying these techniques to new problems.
Download Scenefile:
entagma.com/downloads/ML101_Ep00_IntroDemos.zip
Chapters:
00:00 Course Introduction
02:01 ML Deformer End Result
04:47 ML Deformer Pros and Cons
06:18 Why this is our first project
08:41 Why muscle sims can work with machine learning
13:48 Closing Remarks
Fire Hydrant Model:
polyhaven.com/a/fire_hydrant
Download project file here:
entagma.com/downloads/H20-5_COPs_ParticleBrush.hip
Chapters
00:00 Intro
00:20 POP Setup
04:44 Creating Streaks from Particles
06:00 Creating Attributes for Copernicus
10:37 Copernicus Setup
Download project file here:
entagma.com/downloads/H20-5_COPs_PBR-Layering.zip
Chapters:
00:00 Intro
00:33 Base Node Tree
07:02 Digital Asset Conversion
10:03 Adding Parameters
Assets used:
polyhaven.com/a/fire_hydrant
polyhaven.com/a/brown_mud_03
cgbookcase.com/textures/painted-metal-03
cgbookcase.com/textures/old-iron-01
Download project file here:
entagma.com/downloads/H20-5_COPs_PBR-Layering.zip
Chapters:
00:00 Intro
00:49 Model Prep
02:43 Copernicus Setup
05:04 PBR Layer HDA
08:27 ID Masks (Metal)
12:40 Curvature Masks (Metal Scratches)
15:59 Directional Masks (Dirt)
18:52 Decals (Logo)
21:21 Export and Previewing with Karma
Today we want to take an artifact from digital cameras, something that we usually try to avoid, and turn it up to eleven! Make it an effect on it's own! And that artifact is rolling shutter. To achieve this, we'll need to bend time using point attributes, which we'll do with a bit of SOP trickery.
Download the scenefile here:
entagma.com/downloads/RollingShutter.zip
Chapters:
00:00 Intro
01:17 Base RBD Sim
04:17 Calculate Time Offset
08:03 Time Shift A: Create Motion Trails
10:16 Time Shift B: Reconstruct Motion From Trails
13:33 Time Shift C: Offset Time Per Point
Bubbles! We have a bit of a thing for bubbles here at Entagma. But the one classic Houdini setup we have not done yet is a bubble pop. So let's fix this right now!
Download the scenefile here:
entagma.com/downloads/SoapBobblePop_v02.hip
Chapters:
00:00 Building the base bubble
02:11 Creating the pop mask via a pyro spread sim
05:06 Splitting the bubble along the pop line
07:44 Flip Sim A: Sop preparations
10:44 Flip Sim B: Basic parameter tweaking
13:22 Flip Sim C: Driving the sim with our pop mask
16:32 Flip Sim D: Adjusting droplet detection, surface tension and reseeding
20:59 Flip Sim E: Caching and Meshing
22:35 Merging bubble and Flip sim
27:54 Rendering in Solaris and Karma
In an earlier episode we took a first look at the new ONNX inference node, a node that allows us to use trained neural nets inside Houdini. And today we want to use this node for something usefull - Volume Style Transfer! We want to apply a pretrained style to voxels instead of pixels and give our simulations a unique look and much more detail. And we're going to do all of this using just a tiny bit of machine learing and a lot of SOPs!
Watch our previous ONNX episode here:
youtu.be/aCAatiY53s8
Download the scenefile here:
entagma.com/downloads/Free_ONNX_VolumeStyleTransfer.zip
Chapters:
00:00 Intro
02:18 A quick and dirty Pyro setup
04:57 Creating volume slices
09:38 Writing voxel data to slices
11:22 Deleting empty slices
12:43 Style transfer
16:48 Assembling a volume from slices
19:00 Mixing with original volume
Chris Hoffmann (UglyStupidHonest): instagram.com/uglystupidhonest/?hl=en
ComfyUI: github.com/comfyanonymous/ComfyUI
Wundt Curve: frontiersin.org/articles/10.3389/fncom.2021.698252/full
Pixel Composer: makham.itch.io/pixel-composer
Our new master's program: https://www.design-digital-futures.de
Our master's insta: instagram.com/thn_design_master
With Houdini 20 we got our first official machine learning node from SideFX. While we previously had example scene files that did machine learing related stuff or plugins like mlops, this is the first machine learning tool that you get just by installing Houdini. So we should take a look at what it does and what we can use it for!
Download Project File:
entagma.com/downloads/Free_ONNX_Intro.hip
Links mentioned in the video:
github.com/onnx/models
huggingface.co/julienkay/sentis-MiDaS
Chapters:
00:00 Intro to ONNX
03:14 Getting ONNX Models
05:29 MNIST Model
09:00 Style Transfer Model
15:13 Depth Estimation Model
17:37 Limitations
Welcome to the second and final part of our mini series on how to melt complex objects with fine details, textures and multiple materials. All without melting your pc as well! In the first part we built a setup to retain geometric detail and now we're bringing materials, uv coordinates and textures over to our fluid and finally render everything in Karma.
Watch part 1 here:
youtu.be/fbI91vzBmKo
Download Project File:
entagma.com/downloads/Free_MeltingComplexObjects_Part02.zip
Chapters:
00:00 Intro
01:27 Creating Material Masks
03:51 Transferring UVs
08:28 Solaris Setup
10:55 Karma CPU Shader
14:54 Adding Glow
16:32 Karma XPU Shader
18:21 Alternative Point Color Workflow
You guys seem to like melting things! So over the next two videos on our Youtube channel, we're going to answer one of the more common questions: How to melt more complex objects with fine details, textures and multiple materials. All without melting your pc as well!
Watch Mo's original video here:
youtu.be/B3W-S0EW9xw
Download Project File:
entagma.com/downloads/Free_MeltingComplexObjects.zip
Chapters:
00:00 Intro
01:11 Starter File Overview
03:58 Fluid Prep
07:46 High Res Object Prep
09:37 Fluid Meshing
13:11 High Res Object Deforming
19:07 Blending Both And Finishing The Effect
In this episode we're using a little geometry processing trick to build a sci fi panel deformer, much like some of the tools in the Tissue addon for Blender. We'll also learn how to use some of the OpenSubDiv functions within Vex in Houdini.
We highly recommend this additional OpenSubDiv tutorial by Konstantin Magnus:
youtube.com/watch?v=c8ZiYXbN3y0
Download Project File:
entagma.com/downloads/Freetut_OSD_Deformer.zip
We are happy to bring you another Blender Tutorial with the help of Louis Brodnig. He will show you how to create a differential line growth set up with Blender Geometry Nodes that he used in his Youtube Video about Emergence.
Check out Louis Youtube Video:
youtube.com/watch?v=cXVTIpYDJa4
Download Project File:
entagma.com/downloads/Differential%20Line%20Growth%20BlenderGeoNodes.blend
Links mentioned in the video:
ventrella.com/Clusters
youtube.com/watch?v=p4YirERTVF0
Download Project Files
entagma.com/downloads/Freetut_ParticleLife.zip
Timecodes:
00:00 Intro and Theory
07:14 Houdini Prep
12:19 Coding the Algorithm
19:26 Playing with the Algorithm
Houdini 20 comes with a completely rebuilt cloud workflow, which makes it a whole lot easier to create a wide variety of realistic clouds. So let's build one from scratch and take a look at the new cloud shape tools, noises and shaders!
Downlload Scene File: entagma.com/downloads/Freetut_H20_Clouds.hip
Download Project File:
entagma.com/downloads/H20_SopsDopsKarma.zip
Download Scene File:
entagma.com/downloads/H20_SopsDopsKarma.zip
In this episode we're creating our feather templates from a photo reference you can download here:
featherbase.info/en/species/Erithacus/rubecula
entagma.com/downloads/H20_Feathers.zip
Video Chapters:
0:00 Importing the Feather Template
1:06 Creating the Feather
2:53 Tweaking Parameters
4:01 Closer Look at the Feather Geometry
5:13 New Nodes to customize Feathers
8:50 Manipulate Feather Geometry
9:50 Naming for multiple Feathers
10:36 Using the Template as a UV Texture
12:58 Cache the Result
In this episode we build a plumage from the templates we created in episode one using the standard Houdini grooming tools as well as something a bit more procedural.
entagma.com/downloads/H20_Feathers.zip
Chapters:
0:00 Set up a Groom
1:21 Groom the traditional way
2:18 Guide Groom Node
4:52 Get a high Poly Version
5:55 a more procedural Approach
7:19 Turn Guide Curves into Feathers
8:43 Fix Orientation
9:12 Deintersect Feathers
10:30 Fix UV Texture
In this episode we'll simulate our groom using Vellum and finally bring our character into Karma for rendering.
entagma.com/downloads/H20_Feathers.zip
We’re thrilled to have Andreas Catucci as a guest! Somehow aside from creating amazing work, he managed to record a video on a technique very dear to our hearts: Dart throwing. Sounds like simulating aerodynamics? Not quite. It’s a method to densely pack a surface with non-overlapping geometry.
Not only does Andreas cover the implementation but also does an amazing job explaining the why.
Check out Andreas' Portfolio:
andreascatucci.com
Download Project File:
entagma.com/downloads/A_Dartthrower_v01.hipnc
For more visit: www.entagma.com
Today we’re revisiting an old setup of ours: Creating intricate interference patterns using circular waves. Time has passed since Mo built the first version of this in 2018 and now, with Houdini 19.5, we got a new node that finally allows us to calculate those patterns on 3d surfaces as well.
Download Project File: entagma.com/downloads/Freetut_InterferencePatternsOnSurfaces.zip
In this video Chris covers a simple way of simulating origami like patterns in Houdini and Vellum, just using edge groups and (depending on how you count them) three or six lines of vex.
We can very much recommend watching Junichiro Horikawa's previous explorations on the topic here:
youtube.com/watch?v=VKO4DUso3Bs
youtube.com/watch?v=J91mQ9OJq6s
youtube.com/watch?v=2f3joj3RUCo
Download the project file here:
entagma.com/downloads/VellumFolding.zip
Kapitel
00:00:00 Setting up fold groups
02:24:18 A bit of theory
05:48:17 Transferring fold groups to bend constraints
09:07:06 Animating the folds
Get MLOPs: www.bismuth.at/mlops
Get MLOPs: www.bismuth.at/mlops
More on: www.entagma.com
There are some setups that start out looking quite easy and then get harder the more you start thinking about them. And this belt adjuster is one of them because we got quite a few elements interacting with each other. As usual with Vellum, the solution here is choosing the right constraints and then tweaking lot's of values, which is exactly what we're covering in this episode.
We just released a major update for Machine Learning Operators (MLOPs) in Houdini. Here's the rundown of its major new features. There's more, just go to our Github and join our Discord to find out!
00:00 - SAM (Segment Anything Model)
01:03 - Remove BG
01:45 - UMAP Dimensionality Reduction
03:07 - Image Similarity Measurement
04:20 - Prompt Similarity Measurement
05:00 - BLIP Caption & CLIP Interrogate
06:07 - Camera To Points
07:40 - CLIP Image Embeddings
08:20 - RESRGAN Upressing
09:03 - InstructPix2Pix
09:51 - Stable Diffusion Pipeline Node
10:43 - Teaser: Training
11:00 - Infamous Last Words
https://bismuth.at/mlops
https://www.bismuth.at/mlops
Installing MLOPs:
youtube.com/watch?v=rtgghkYKKLY
First Setup:
youtu.be/3kVB-bNWeYA
00:00 - Intro/Setup
00:43 - Setting Up Stable Diffusion
01:42 - Depth Map From Viewport Camera & ControlNet
03:16 - Backprojection / Texturing
05:02 - Backsides (Clip & Flip)
06:25 - Infamous Last Words
This week we want to give you a lot more insight into what a latent space actually is and how it gets created. In this case not by using Stable Diffusion, but by building and training our own neural net inside Houdini, who’s only job is to create a very simple latent space for us to explore.
In this video we’ll go over the main concepts and the theory of our latent space. And on Thursday in the Advanced Setups Course on Patreon, we’ll build the entire setup using Houdini and Pytorch.
You can find all visualisations from this video to explore by yourself in the scene file. Make sure you have access to Pytorch and Numpy in your Houdini Python environment, for example by installing the MLOPs Addon.
Download MLOPs
github.com/Bismuth-Consultancy-BV/MLOPs
Parts of this setup are inspired by this article by Eugenia Anello
medium.com/dataseries/convolutional-autoencoder-in-pytorch-on-mnist-dataset-d65145c132ac
entagma.com/downloads/MyFirstLatentSpace_Part01.zip
Now that MLOPs is released and available for everyone, let's create some motion graphics using those new tools and combine them with other workflows we already know from Houdini. In this video Chris shows you how he created these strange forest renders while beta-testing MLOPs and how to fake depth of field inside point clouds.
Get MLOPs for Houdini: https://bismuth.at/mlops
MLOPs Installation Instructions
youtube.com/watch?v=rtgghkYKKLY
MLOPs Txt2Img
youtube.com/watch?v=3kVB-bNWeYA
MLOPs Controlnets
youtube.com/watch?v=_bxir7qGjzQ
entagma.com/downloads/MLOPsStrangeForests.zip
https://www.bismuth.at/mlops
Installing MLOPs:
youtube.com/watch?v=rtgghkYKKLY
First Setup:
youtu.be/3kVB-bNWeYA
00:00 - Intro/Setup
00:24 - Theory
00:43 - Implementation
02:05 - Bugfixing
02:25 - Parameter Fiddling
03:22 - Examples
03:45 - Infamous Last Words
https://www.bismuth.at/mlops
Installing MLOPs:
youtube.com/watch?v=rtgghkYKKLY
First Setup:
youtu.be/3kVB-bNWeYA
00:00 - Intro/Setup
01:54 - Prompt Blending
04:50 - Prompt Arithmetic
https://www.bismuth.at/mlops
Installing MLOPs:
youtube.com/watch?v=rtgghkYKKLY
First Setup:
youtu.be/3kVB-bNWeYA
00:00 - Intro / Standard Setup
01:03 - Simple ControlNet
03:45 - Multi ControlNet
05:31 - Inpainting Using ControlNet
06:40 - Infamous Last Words
https://www.bismuth.at/mlops
Installing MLOPs:
youtube.com/watch?v=rtgghkYKKLY
First Setup:
youtu.be/3kVB-bNWeYA
00:00 - Img2Img Workflow
01:54 - Masking/Inpainting
youtube.com/watch?v=rtgghkYKKLY
Download MLOPs:
https://www.bismuth.at/mlops
00:00 - Intro
00:32 - Prompt Create
00:41 - Tokenizer
00:53 - Text Encoder
01:25 - Negative Prompt
01:42 - Scheduler
02:08 - Latent Noise Generate
02:35 - Solver & Scheduler Parameters
03:25 - Running The Solver
03:45 - Visualizing Latents
04:07 - Image Decoder
04:28 - Image Export
05:15 - Infamous Last Words
The idea for this video came from a question on our Patreon Community Hub, asking us how we'd go about creating curling ribbons in vellum, similar to a very beautiful setup by Playback Design. And, as it turns out, this is a quite fun and simple build!
00:31 Single Ribbon: Build
03:45 Single Ribbon: Sim
08:30 Creating Copies
10:50 Many Ribbons: Animation Attribute
14:37 Many Ribbons: Animating
15:24 Many Ribbons: Sim
17:30 Fixing Index Issues And ReSimming
Original project by Playback Design:
behance.net/gallery/160665727/Phyllotaxis
Download Project File: entagma.com/downloads/VellumCurlingRibbons.hip
What do you do when you need fast info passes rendered out but are too stupid to set up something in ROPs and too lazy to use Solaris? Right - you build yourself a 1970s render engine straight in SOPs. Clever? Not sure. Useful? Definitely. Make sure to check back on entagma.com on Thursday for this week's premium tut to see what we can use these passes for.
Download Scene File: entagma.com/downloads/diy_renderer_02.hip
00:00 - Intro
00:54 - Setting Up Something To Render
01:48 - Building The Renderer: Image Plane
02:55 - Transforming Image Plane To Camera Space
04:25 - Ray Directions
05:57 - Merging In Render Geometry
07:24 - Geometry Intersections / Rendering Depth
12:22 - Writing Our Rendering Into Volumes
14:22 - Rendering Normal Map
17:12 - Rendering Segmentation Map
19:16 - Rendering A Cheap Diffuse Pass
22:50 - Aux Maps To Volumes
24:28 - Importing Volumes To COPs for Compositing
25:22 - Anti-Aliasing
26:23 - Recap / Infamous Last Words
Of course Mo could take the high road and build a stable diffusion pipeline using Huggingface’s Diffusers library… But let’s not kid ourselves. Thanks to Automatic1111’s API Mo can duct tape together Houdini and WebUI to use both of these awe inspiring powerful tools to generate… dog images. Welcome to the brave new world that is design!
Download Project File: entagma.com/downloads/SD_txt2img_003a.hip
00:00 - Intro
00:50 - Setting up Automatic1111 WebUI's API
01:49 - Houdini Basic Scene Setup
03:10 - Python Imports & Writing Images To Volumes
08:07 - Main Routine: Controlling WebUI With Python
13:25 - Extracting Our Data From WebUI
15:38 - Moving The Data To Compositing
Jeroen Claus of phospho.co.uk happens to know a particular kind of software that's pretty good at applying a set of geometrical rules. So he sets out to take you on a journey of building Escher inspired tesselations.
Download Project File (.hiplc): entagma.com/downloads/Entagma_EscherTessellation_v01.hiplc
More on: www.entagma.com
We continues poking into stable diffusion by giving a brief overview of what tokens and embeddings are and how the can be manipulated to blend between prompts. Finally we create some nonsensical animals.
Huge thanks to Chris Hoffmann (ugly stupid honest): instagram.com/uglystupidhonest
Notebook: Stable Diffusion Deep Dive: github.com/fastai/diffusion-nbs/blob/master/Stable%20Diffusion%20Deep%20Dive.ipynb
The Illustrated Stable Diffusion: jalammar.github.io/illustrated-stable-diffusion
Automatic1111's WebUI: github.com/AUTOMATIC1111/stable-diffusion-webui
00:00 - Intro / Theory
00:55 - Visualizing Tokens / Embeddings using WebUI
03:03 - Stable Diffusion Deep Dive as Basis
03:55 - Modifying The Notebook
05:22 - Custom Compact Notebook
08:16 - Why Not Use WebUI?
08:55 - Infamous Last Words
Download Project File (.ipynb): entagma.com/downloads/StableDiffusion_BlendEmbeds2_003.zip
He's been working together with http://tape-lab.com, UCL Cancer Research's Cell Communication Laboratory, to visualise data from their study "A Single-cell Perturbation Landscape of Colonic Stem Cell Polarisation", which is now available as preprint: biorxiv.org/content/10.1101/2023.02.15.528008v1
In this video, Jeroen shows how you can use single cell RNA sequencing data to create a data-driven version of the Waddington Landscape, a classical graphic from the 50s representing the possible developmental paths of a cell. This new way to visualise single cell RNA sequencing data helps to make dimensional reductions of huge amounts of data more intuitive to interpret and understand.
While the data might be complex, the Houdini setup is rather straightforward: this tutorial will give you an overview of how to get csv data into Houdini, and with only a few nodes build an effective data visualisation.
Download Project Files (.zip): entagma.com/downloads/Waddington_Landscapes_Tutorial.zip
In part seven of our free course on building a rolling object solver, Bastian J. Schiffer (https://prdx.de/) modifies our solver to take into account random initial rotation and so it works with cubes instead of tubes.
And with this final bonus tutorial - it's a wrap! Thanks so much to Bastian for sharing his knowledge!
Visit his website to check out some of the amazing projects he 's worked on: https://prdx.de
00:00 - Bonus: Rolling Cubes
02:08 - Deep Dive: A More Precise Rotation Approach
10:36 - Up-Facing Discs on Displaced Surfaces
16:07 - Random Initial Rotation on the Discs
Download Project Files: entagma.com/downloads/Rolling-Objects-Tutorial.zip
In part six of our free course on building a rolling object solver, Bastian J. Schiffer (https://prdx.de/) modifies our solver so it works with discs instead of spheres.
00:00 - Intro / Building the Test Geometry
02:42 - Extending the Solver
Download Project Files: entagma.com/downloads/Rolling-Objects-Tutorial.zip