John Paszkiewicz

One more terrain study with GAEA and Blender of course https://lnkd.in/gcvdCV5h

10

1 Comment

https://lnkd.in/gxmZ2Sei Ink dissolving in water while passing through wobbling cells. Simulated in Houdini DOPs, for achieving mutual interaction of Vellum and Pyro solvers. Rendered in V-ray for Houdini, composited/color corrected in After Effects. #Houdini #Simulation #CGI

9

2 Comments

The VFX industry is buzzing about USD (Universal Scene Description), but what exactly is it? This week on #VFXTuesdays, I'd love to dive into a great video from Foundry exploring the capabilities of USD and why major studios are embracing it. Is USD the key to streamlining workflows and boosting collaboration? Watch the video and see for yourself! #vfx #usd #compositing #3d #animation

722

17 Comments

🎥 Proper documentation is key, and this tutorial will walk you through every step of installing and setting up das element Watch here: https://lnkd.in/eVCzhA8J #vfx #visualeffects #documentation #daselement

27

DiffIR2VR-Zero: Zero-Shot Video Restoration with Diffusion-based Image Restoration Models 🔍 🔍 🔍 This paper presents a zero-shot video restoration method using pre-trained image restoration diffusion models, addressing the limitations of traditional methods that require retraining and have poor generalization. By employing a hierarchical token merging strategy and a hybrid correspondence mechanism, the approach achieves superior performance and generalization across various datasets and extreme degradations. The method works with any 2D restoration diffusion model, providing a versatile tool for efficient and high-quality video enhancement without extensive retraining. project: https://lnkd.in/dm4_Nkgu paper: https://lnkd.in/dghiYuTh code: https://lnkd.in/dmdHcSKj demo: https://lnkd.in/dBAUeB7X #VideoRestoration #ZeroShotLearning #DiffusionModels #ImageRestoration #AI #MachineLearning #OpticalFlow #SuperResolution #VideoEnhancement #ComputerVision #HighQualityVideo #TechInnovation

6

📣 REACTO: Reconstructing Articulated Objects from a Single Video 📣 This paper discusses the challenge of reconstructing 3D articulated objects from single videos, highlighting the limitations of existing dynamic neural radiance fields in modeling piece-wise rigid objects. They introduce Quasi-Rigid Blend Skinning, a novel deformation model that enhances part rigidity while allowing flexible joint deformation, integrating enhanced bone rigging, quasi-sparse skinning weights, and geodesic point assignments. Their approach significantly improves the fidelity of 3D reconstructions of articulated objects, as evidenced by tests on real and synthetic datasets. project: https://lnkd.in/dAk-8qvc github: Code not yet available. paper: https://lnkd.in/d6u7udMR #3DArticulationTech #AdvancedDeformationModel #RigidBlendSkinning #HighFidelity3D #NextGen3DModeling #DynamicNeuralFields #ArticulatedObjectsRecon

7

🚀 Transforming Manufacturing with Spark Vfx Studios 🚀 At Spark Vfx Studios, we're revolutionizing the manufacturing industry by creating engaging, informative 3D animated videos. Our cutting-edge animations help manufacturers showcase their complex processes in a way that's easy to understand and visually captivating. 🏭✨ 🔹 Why 3D Animation? Simplifies Complex Concepts: Break down intricate manufacturing processes into clear, digestible visuals. Enhances Communication: Visual storytelling helps convey your message more effectively than traditional methods. Boosts Engagement: Dynamic animations capture and retain the audience's attention, making your pitch unforgettable. 🔹 Benefits for Manufacturers: Sales Enablement: Equip your sales team with powerful visual tools to demonstrate product features and benefits convincingly. Training & Onboarding: Accelerate employee training with visual aids that simplify learning. Marketing & Promotion: Stand out at trade shows, on social media, and in marketing campaigns with eye-catching animated content. 🔹 Success Stories: Our collaborations have helped clients reduce training time by 30%, increase engagement rates by 50%, and drive a 20% boost in sales conversions. 🌟 Ready to elevate your manufacturing business? Let's bring your processes to life with Spark VFX Studios. Contact us today to get started! 🌐✨ #3DAnimation #Manufacturing #VisualStorytelling #SalesEnablement #IndustrialMarketing #SparkVFXStudios #InnovationInManufacturing #DigitalTransformation Feel free to share your thoughts or reach out if you’re interested in learning more about how we can help your business thrive! 🚀 https://lnkd.in/gFptJKPv

7

Character subtraction. Clean up in Nuke #Nuke #Cleanplate #Cleanup #VFX #Paint #Openforwork

19

If you're in London before the start of July. Why not visit Cooke Optics Gallery in Newman St W1. Our current exhibition FOCUS delves into the ways in which cinematographers us focus as a tool and technique to communicate how the audience engages with a motion story. Motion is just one aspect of consideration when designing a lens to convey dimesionality. At Cooke our lens design philosophy is to create lenses for motion capture that create a 3d feel on a 2d plane. Otherwise known as The Cooke Look. Which is why we only design and manufacture lenses for motion capture. As for the use of organic lens use in VR, VFX, gaming and animation. We recently set up our new Cooke Digital division to advance our unique mapped lens data for integrated use in the all aspects of motion capture across the entertainment industries. Since our /i lens data protocol is an industry standard in motion picture capture. The next exhibition COLOUR, will review cinematographers use of colour in the story telling process.

13

I got together with our CEO Rob E. to weigh in on the largest change in our industry in decades. Here’s our take at Massive Mission on AI for post-production Consider Jurassic Park in 1993. Their plan was to film models as had been done for decades. When the first CGI dinosaur test was shown, its advantages were obvious. A threshold had been crossed. Many model and prop makers  adopted the new tools and picked up their art in the new medium. This is where AI is at today for the media production industry. We all will need to adapt because this is a sea change of our industry, bigger than even film to digital. With just the AI tools brought to market in the last 18 months we’re now able to: Do 50-80% of the audio cleanup work in minutes vs. hours Replace, remove, or create visuals in our videos on the fly by describing the changes with text  Correct eye line when people are reading offscreen text Speed up rough storyboards from days to hours And more, higher quality tools are being created weekly. It often feels like too much to keep up with. These tools are coming so fast, it’s like drinking from a fire hose. Which ones are worth testing, and which ones are just a re-skinned chat-GPT? Which ones move the needle meaningfully in the right direction and what ethical and environmental boundaries do we need to consider as we navigate this tech? Our approach is to test and learn. Explore these tools' strengths to avoid their weaknesses. With so many options in the market, it can also be hard to sift. Matt Wolfe, Mike Seymour, Michael Kammes, Katie Hinsen, & Two-Minute Papers, are great people to watch to filter through the news. A helpful lens we’re using to digest all this is ‘progress’. With each new tool presented, consider, “Is this better than what I’m doing now?” Is the end result progress over my previous process? For many (not all) of these tools, the answer is yes. Ultimately, we all want to know: will using AI help give us more free time to spend with loved ones or pursue passions? That’s the goal, right? Bottom line, consider if AI empowers your art and improves your life. From a business and ethical perspective, pay attention to standards. Initiatives like the Coalition for Content Provenance and Authenticity are helping establish much-needed guardrails. The AI Act in Europe seems like it could pave the way for legislation worldwide, much like it did with GDPR. The 2030 Vision Initiative from MovieLabs is gathering inertia from industry leaders. Keeping track of these high-level decisions helps set a good framework for other new developments. AI is a double-edged sword. As professionals, we need to be shepherding these tools to create great work, while being mindful that powerful tools have powerful consequences if used incorrectly. #ai #mediaproduction #videoproduction

2

1 Comment

Here is a new tutorial. It's a big one. ADVANCED CHARACTER RIGGING IN AFTER EFFECTS. Hope you all will like it. https://lnkd.in/dxrf8yCF #animation #tutorial #design #motiongraphics

9

DiM: Diffusion Mamba for Efficient High-Resolution Image Synthesis 🔔🔔🔔 Diffusion models have excelled in image generation, but the computational cost of Vision Transformers poses challenges for high-resolution images. To address this, we introduce Diffusion Mamba (DiM), combining the efficiency of Mamba, a sequence model based on State Space Models (SSM), with diffusion models for high-resolution image synthesis. DiM incorporates multi-directional scans, learnable padding tokens, and local feature enhancement, achieving efficient high-resolution image generation, further improved by a "weak-to-strong" training strategy and training-free upsampling methods for even higher resolutions. project:https://lnkd.in/gxgVrJPv paper: https://lnkd.in/gr8tgKrG #DiffusionModels #ImageGeneration #HighResolution #VisionTransformers #DiM #StateSpaceModels #EfficientAI #ImageSynthesis #MachineLearning #DeepLearning #AIResearch

3

🚀Week 10 at VFXTalk! 🪐 Join us!: https://lnkd.in/gwywTJsh 🧠 Key Metrics 📈 398 total members! 331 Pros 41 Students 26 Educators 1 new Admin (Welcome Hugh!) 2 new Moderators (Thank You Tim and Genevieve!) *** 📣 Hope you’re all doing well after a jam-packed week. Just wanted to share some quick thoughts before work today. It’s pretty cool to see how our community boards give us a snapshot of the industry worldwide, way beyond what you usually find on other social sites. I hoped for this but didn’t expect it. With “Survive till ‘25” as our current theme, it’s no shocker that things are a still slow in the US and Canada. But the UK is starting to buzz again, and Australia and NZ are full of activity! Looks like things might really kick off once the IATSE stuff settles down. 🤞 Also, we’ve got a new server neighbor —LightLinks! Last week Lighting Artist Karen Kuzmiak (https://lnkd.in/gPVXRzjN) started up her own Discord with a similar mission to ours. So swing by and say hello! She’s sending out private invites. I chatted with her a bit and she’s super nice and all about helping the 3D community. Remember, high tides raise all boats, so let’s keep spreading the good vibes! There are plenty of conversations to be had. Lastly, I want to say that I really love seeing all the ways artists are benefiting from this project of ours. Whether you’re a student getting solid feedback on your project or a pro expanding your network and wanting to chat about this or that, it feels great to see our community mission and tone take off and become something special. And it’s all thanks to all of you! *** ⚡️Some of our popular threads are: ⚡️ 1. “Having ADD in the VFX industry”: In a surprise to no one, there’s a lot of artists with “neurodivergent” wiring in our industry :) 2. “Custom settings for Nuke nodes”: We’ve all got our own secret sauces, but are your tools battle tested by years in production? Ours are! Come check out what everyone is sharing. 3. “How often should an artist publish work for Feedback?”: This is a great conversation started by a student that turned into a great talk about proper note taking practices. I promise it’s more interesting than it sounds lol. 4. “What have you seen recently that you wish you got to work on?” It’s like fantasy football for VFX! We’ve all got those shows that made us want to work in this crazy business. Come and add to your favorite projects to the big list. 5. “Fallout Thread”: Just like everyone this week, we had a ton of members binging the show and talking about it! Some even worked on it! We’ve got a lot planned for the future and there are SO many more conversations being had too. Do you have something you’d like to talk about? Come over and post! Every voice matters and all are welcome no matter what your skill level is. *** See you all next week! Signing off… #VFXTalk #CommunityBuilding #VFXCommunity #VisualEffects #VFX #CG #IndustryInsights #JoinUs

11

1 Comment

Here’s a pack shot render of my completed Bandai CG asset. Textured in Substance Painter. The game's demo sequence was created by constructing the sprites as vectors, then sequencing an animation precisely like the original. #VFX #VisualEffects #3D #CGI #B3D #Compositing #SubstancePainter #Blender

4

StoryDiffusion: Consistent Self-Attention for Long-Range Image and Video Generation 🔔 🔔 🔔 This paper introduces a new approach called Consistent Self-Attention that improves the consistency of content across a series of images generated by diffusion-based models. Additionally, the paper presents a Semantic Motion Predictor for long-range video generation, which estimates motion conditions in the semantic space between images, allowing for smooth transitions and stable subjects in video sequences. The integration of these technologies into the StoryDiffusion framework enables the generation of consistent images and videos that narratively align with a text-based story, advancing research in visual story generation. project: https://lnkd.in/d8ZBGesG paper: https://lnkd.in/dN3KeBJj github: https://lnkd.in/d-r5PuGR demo: https://lnkd.in/dUSN-khP #ConsistentSelfAttention #SemanticMotionPredictor #StoryDiffusion #VisualStoryGeneration #DiffusionModels #TextToImage #VideoGeneration #AIResearch #GenerativeModels #InnovationInAI #AIStorytelling

12

3 Comments

My esteemed colleague, Alban Bergeret has written an awesome new EDC tutorial on high precision rotations in Unreal... Check it out here: https://lnkd.in/gbKtMnY9 #unrealengine #simulation #visualization #engineering #gamedev #animation #storytelling #filmmaking #realtime #UE5

32