Video yükleniyor...

Video Yüklenemedi

Bu video yüklenirken bir sorun oluştu. Bu geçici bir ağ sorunundan kaynaklanıyor olabilir veya video kullanılamıyor olabilir.

Ana Sayfaya Dön

I’m using Blender natively to import Apple ML-Sharp Gaussian Splatting. ✨ Controlling focal point, camera parameters and parallax directly inside Blender opens the door to entirely new workflows. The result is seriously impressive. 🔥 #b3d #MLSharp #GaussianSplatting

Thomas Marcos

6,372 subscribers

75,320 görüntüleme • 8 ay önce •via X (Twitter)

Sanat Bilim & Teknoloji #b3d #MLSharp #GaussianSplatting

Anya Rossi• Live Now

Private livecam show

21 Yorum

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

Here it’s still a simple setup, but I find this use case really interesting, especially for adding beautiful defocus to something like a portrait photo. At @benutsVFX, we’ve been experimenting with Gaussian Splatting more broadly for several months, exploring how it can fit into cinematic production. Follow us to stay updated on our research and developments. 🎬✨

BLUE NILE 3D profil fotoğrafı

BLUE NILE 3D8 ay önce

Niiiiice, with the Mediastorm setup?

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

I didn’t know it! But it looks super handy

Nitin Sharma 🔸 profil fotoğrafı

Nitin Sharma 🔸8 ay önce

how does it behaves with DLSS?

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

Oh that’s a great idea. I need to test it soon. I’m not sure if DLSS requires normals, depth or motion vectors for this setup.

Massimo 🫧 profil fotoğrafı

Massimo 🫧8 ay önce

Is this octane renderer? Kiri? Something custom?

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

Here it’s rendered in Cycles, starting from a Gaussian Splat setup built in Geometry Nodes. The splats does not update the camera data dynamically when you rotate around the scene.

BLENDER SUSHI 🫶 X - 24/7 Blenderian profil fotoğrafı

BLENDER SUSHI 🫶 X - 24/7 Blenderian8 ay önce

Could be interesting to make stop motion animation

Tensor Z🏳️‍🌈 profil fotoğrafı

Tensor Z🏳️‍🌈8 ay önce

This is so creative

internet profil fotoğrafı

internet8 ay önce

that was my motivation for using splats as well, making it possible to take photos with arbitrary cameras (I’m using octane here)

Naman kapoor profil fotoğrafı

Naman kapoor8 ay önce

Woah

DoctorDisco profil fotoğrafı

DoctorDisco8 ay önce

What are you using for the dof/lens distortion? I don’t think you can do that vanilla, but I’m not sure

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

You can use the “Fisheye” lens model instead of “Perspective.”

Marchen profil fotoğrafı

Marchen8 ay önce

blender native splat import is huge. the depth of field control alone makes this workflow worth it

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

There is no native Gaussian Splat importer in Blender yet, but with Geometry Nodes it is absolutely possible to build a splat reader. 😉

Marchen profil fotoğrafı

Marchen8 ay önce

ah gotcha, geometry nodes approach makes sense

Count JJ V3D-Artist profil fotoğrafı

Count JJ V3D-Artist8 ay önce

The splat is native in blender? Or is there an add on I’m not aware of?

Thomas Marcos profil fotoğrafı

Thomas Marcos8 ay önce

It’s done through Geometry Nodes. Since ML-Sharp is built for a single camera, it makes the setup much simpler to implement.

Count JJ V3D-Artist profil fotoğrafı

Count JJ V3D-Artist8 ay önce

You have a tutorial or something on setting up the geometry nodes? Looks a lot nicer than KIRI’s

Jeremiah Warren ◡̈ profil fotoğrafı

Jeremiah Warren ◡̈8 ay önce

This is so fucking sick

oliverfranklinanderson profil fotoğrafı

oliverfranklinanderson8 ay önce

I pulled this off with a splat import addon and MLsharp a few weeks ago, are you sharing your geonodes setup in the future?

Benzer Videolar

📢 I’m testing ViggleAI new AI tool, PINOC, and it’s seriously impressive. You can create 3D Gaussian splatting models from just one image and animate them with a video! It’s the first AI I’ve tried that can animate Gaussian splatting models. I also talked to someone from Viggle, and they might add the option to import your own Gaussian splatting models and animate them. If that happens, I’ll probably make a video tutorial showing how to boost the detail level of a Gaussian splatting model by 4x to create highly realistic animated characters.

📢 I’m testing ViggleAI new AI tool, PINOC, and it’s seriously impressive. You can create 3D Gaussian splatting models from just one image and animate them with a video! It’s the first AI I’ve tried that can animate Gaussian splatting models. I also talked to someone from Viggle, and they might add the option to import your own Gaussian splatting models and animate them. If that happens, I’ll probably make a video tutorial showing how to boost the detail level of a Gaussian splatting model by 4x to create highly realistic animated characters.

Alex

29,776 görüntüleme • 5 ay önce

Parallax Occlusion Mapping in Blender 5.0 🔥 A single plane, no subdivisions & displacement, just shader math. The new closures and repeat zone now make it possible to implement it as a reusable nodegroup. #blender #b3d #eevee #shaders

Parallax Occlusion Mapping in Blender 5.0 🔥 A single plane, no subdivisions & displacement, just shader math. The new closures and repeat zone now make it possible to implement it as a reusable nodegroup. #blender #b3d #eevee #shaders

Quackers

36,720 görüntüleme • 1 yıl önce

3DGS Render Addon for Blender is out! New features include: - Paint and use image textures to edit colour - Mesh to 3DGS conversion - Transformations and edits are exportable - Lot's of other boosts and little features! Link for the addon and a Quick Guide tutorial below #b3d #blender #3dgs #gaussiansplatting

3DGS Render Addon for Blender is out! New features include: - Paint and use image textures to edit colour - Mesh to 3DGS conversion - Transformations and edits are exportable - Lot's of other boosts and little features! Link for the addon and a Quick Guide tutorial below #b3d #blender #3dgs #gaussiansplatting

KIRI Engine - 3D Scanner App

19,019 görüntüleme • 1 yıl önce

4D videos, are we cooked? 😳 uses 4D Gaussian Splatting to reconstruct real footage into fully explorable scenes, letting you move the camera freely through a moment after it’s been filmed. Combine it with tools like Seedance for character replacement and motion control, and Beeble for relighting, and you unlock entire cinematic workflows. You can even step inside your scenes in augmented reality with the Apple Vision Pro. #4DVAI #GaussianSplatting #AIVideoWorkflow #CreativeAI #SpatialVideo

4D videos, are we cooked? 😳 uses 4D Gaussian Splatting to reconstruct real footage into fully explorable scenes, letting you move the camera freely through a moment after it’s been filmed. Combine it with tools like Seedance for character replacement and motion control, and Beeble for relighting, and you unlock entire cinematic workflows. You can even step inside your scenes in augmented reality with the Apple Vision Pro. #4DVAI #GaussianSplatting #AIVideoWorkflow #CreativeAI #SpatialVideo

Rourke Heath

69,983 görüntüleme • 3 ay önce

🚨 STOP learning Blender !! Learning Blender in 2026..... Is officially obsolete. MiniMax Design can now automate Blender to create ANY 3d SCENE !! I just gave it a PROMPT.... And MiniMax Design used GPT-6 Astra to handle everything like: 1️⃣ Opening Blender 2️⃣ Generating the 3D elements 3️⃣ Fixing lighting 4️⃣ Camera Tracking Data etc.... It built the entire scene inside Blender. Finally, I used MiniMax H3 Max to add colors to the 3d SCENE !! Then I just rendered the 3d scene !! Best part? Using GPT-6 Astra through MiniMax Design is way more affordable than running it directly.

🚨 STOP learning Blender !! Learning Blender in 2026..... Is officially obsolete. MiniMax Design can now automate Blender to create ANY 3d SCENE !! I just gave it a PROMPT.... And MiniMax Design used GPT-6 Astra to handle everything like: 1️⃣ Opening Blender 2️⃣ Generating the 3D elements 3️⃣ Fixing lighting 4️⃣ Camera Tracking Data etc.... It built the entire scene inside Blender. Finally, I used MiniMax H3 Max to add colors to the 3d SCENE !! Then I just rendered the 3d scene !! Best part? Using GPT-6 Astra through MiniMax Design is way more affordable than running it directly.

Darshal Jaitwar

86,087 görüntüleme • 1 ay önce

New paper from Apple - Sharp Monocular View Synthesis in Less than a Second Mescheder et al. @ Apple just released a very impressive paper (congrats! 🎉🥳). You give it an image and it generates a really great looking 3d Gaussian representation. Uses depth pro. It's really good. The model is about 3GB, and takes ~5-10s on my M1 Max per image. Single feedforward pass network. The video is using "Metal Splatter" to view the .ply from ml-sharp on Apple Vision Pro. Some really wow moments today trying it on different scenes.

New paper from Apple - Sharp Monocular View Synthesis in Less than a Second Mescheder et al. @ Apple just released a very impressive paper (congrats! 🎉🥳). You give it an image and it generates a really great looking 3d Gaussian representation. Uses depth pro. It's really good. The model is about 3GB, and takes ~5-10s on my M1 Max per image. Single feedforward pass network. The video is using "Metal Splatter" to view the .ply from ml-sharp on Apple Vision Pro. Some really wow moments today trying it on different scenes.

Tim Davison ᯅ

133,076 görüntüleme • 10 ay önce

Breakdown Rive : Create the 3D asset in Blender → rig it with bones → import it into Rive → use the Rive CLI to expose the bones → then animate everything directly in the Rive Editor. The cool part is that the 3D bones can be fully controlled and animated directly inside Rive.

Breakdown Rive : Create the 3D asset in Blender → rig it with bones → import it into Rive → use the Rive CLI to expose the bones → then animate everything directly in the Rive Editor. The cool part is that the 3D bones can be fully controlled and animated directly inside Rive.

Angga Motion

19,432 görüntüleme • 26 gün önce

My recent #deltarune model motivated me to implement a new feature for my upcoming Freeform-Facial-Feature tool Using #Blender 4.5 beta and it's manifold boolean The transparency mask should be a lot more reliable with sharp corners and hard surfaces now! #b3d #geometrynodes

My recent #deltarune model motivated me to implement a new feature for my upcoming Freeform-Facial-Feature tool Using #Blender 4.5 beta and it's manifold boolean The transparency mask should be a lot more reliable with sharp corners and hard surfaces now! #b3d #geometrynodes

Krysidian

77,171 görüntüleme • 1 yıl önce

Using Kushiro's new 'Shatter Cut' addon for Blender to quickly generate a random floor plan. Takes no time at all. Feels like cheating. The addon is free btw ✨ Tip: To quickly call the Shatter Cut operator press "Right Click,C" #b3d

Using Kushiro's new 'Shatter Cut' addon for Blender to quickly generate a random floor plan. Takes no time at all. Feels like cheating. The addon is free btw ✨ Tip: To quickly call the Shatter Cut operator press "Right Click,C" #b3d

passivestar

125,752 görüntüleme • 3 yıl önce

After testing img2threejs by Nick, I used the exact same Canyon gravel bike image with Blender MCP. The difference is pretty clear: Blender produced a much more detailed and faithful reconstruction, especially around the frame geometry, cockpit, drivetrain and braking system. img2threejs remains incredibly fast for generating a procedural Three.js result directly from an image, while Blender MCP offers much more control and precision when you want to push the reconstruction further. Same reference image, two completely different workflows. Which result do you prefer?

After testing img2threejs by Nick, I used the exact same Canyon gravel bike image with Blender MCP. The difference is pretty clear: Blender produced a much more detailed and faithful reconstruction, especially around the frame geometry, cockpit, drivetrain and braking system. img2threejs remains incredibly fast for generating a procedural Three.js result directly from an image, while Blender MCP offers much more control and precision when you want to push the reconstruction further. Same reference image, two completely different workflows. Which result do you prefer?

Givros

31,979 görüntüleme • 2 ay önce

Big announcement🚀 Our comprehensive course on 3D Animated Character Creation using Blender 3D is officially online. If you're new to the 3D world and keen on crafting your very first animated character, this is for you. #B3d #animation #CGI #Blender3D

Big announcement🚀 Our comprehensive course on 3D Animated Character Creation using Blender 3D is officially online. If you're new to the 3D world and keen on crafting your very first animated character, this is for you. #B3d #animation #CGI #Blender3D

Ram Singh

36,246 görüntüleme • 2 yıl önce

Blender 3.5 is released and I worked on the new hair node system! 🎉 Here are some clips of how using it looks like, watch the full overview video here: Big shout-out, of course, to the rest of the #geometrynodes team and Daniel Bystedt 🔶 🙌 #b3d

Blender 3.5 is released and I worked on the new hair node system! 🎉 Here are some clips of how using it looks like, watch the full overview video here: Big shout-out, of course, to the rest of the #geometrynodes team and Daniel Bystedt 🔶 🙌 #b3d

Simon Thommes

102,064 görüntüleme • 3 yıl önce

Gaussian Splats are quickly becoming one of my favorite ways to move complex 3D assets between workflows. For this test I started with Panthera Spelaea, the Blender 5.2 splash artwork by Joanna Kobierska, available from under CC-BY. Using GPT-6 Astra, I had it build a 200-frame camera rig and send the renders directly to The Render Network through the API and export a COLMAP data set. NEXT: COLMAP → Postshot → Gaussian Splat. The result is a real-time, high-fidelity version of an insanely dense asset - millions of hairs and all - that I can now bring back into OTOY Octane, relight, reframe, drop into other scenes, or even publish straight to the web. Capture the visual complexity once, then carry it anywhere. Check out the splat:

Gaussian Splats are quickly becoming one of my favorite ways to move complex 3D assets between workflows. For this test I started with Panthera Spelaea, the Blender 5.2 splash artwork by Joanna Kobierska, available from under CC-BY. Using GPT-6 Astra, I had it build a 200-frame camera rig and send the renders directly to The Render Network through the API and export a COLMAP data set. NEXT: COLMAP → Postshot → Gaussian Splat. The result is a real-time, high-fidelity version of an insanely dense asset - millions of hairs and all - that I can now bring back into OTOY Octane, relight, reframe, drop into other scenes, or even publish straight to the web. Capture the visual complexity once, then carry it anywhere. Check out the splat:

Spenser Dickerson

11,923 görüntüleme • 1 ay önce

GPT-6.1 Sol, running in Ultra mode, has been autonomously iterating on this project for more than 16 hours now. It’s using Blender and Three.js, and the entire thing runs directly in the browser. Like I said earlier, this model is genuinely impressive. OpenAI built something here that feels extremely capable, surprisingly cheap, and now fast enough to make long-running agentic workflows much more practical. (The project is still in progress...)

GPT-6.1 Sol, running in Ultra mode, has been autonomously iterating on this project for more than 16 hours now. It’s using Blender and Three.js, and the entire thing runs directly in the browser. Like I said earlier, this model is genuinely impressive. OpenAI built something here that feels extremely capable, surprisingly cheap, and now fast enough to make long-running agentic workflows much more practical. (The project is still in progress...)

Token Gremlin

18,553 görüntüleme • 19 gün önce

Blender just became 10x more powerful with AI. 🚨 Blender + GPT Astra = a completely different way of creating visuals. Instead of choosing between traditional 3D and generative AI, you can combine both. → Build your scene in 3D → Control the camera, depth, and composition → Use AI to transform the result → Create visuals that would normally take hours or days This is where creative workflows are heading. Comment “LAB” and I’ll show you how to do this.

Blender just became 10x more powerful with AI. 🚨 Blender + GPT Astra = a completely different way of creating visuals. Instead of choosing between traditional 3D and generative AI, you can combine both. → Build your scene in 3D → Control the camera, depth, and composition → Use AI to transform the result → Create visuals that would normally take hours or days This is where creative workflows are heading. Comment “LAB” and I’ll show you how to do this.

Sebastien Jefferies

29,008 görüntüleme • 29 gün önce

🏗️ Most 3D reconstruction papers evaluate novel views by applying an optimal color correction using the ground-truth image. That's circular and unrealistic. PPISP directly predicts camera parameters from the rendered radiance alone, using an ISP controller trained on input views. ✨Result: metrics that don't require access to target images, aligning evaluation with real-world deployment. ➡️

🏗️ Most 3D reconstruction papers evaluate novel views by applying an optimal color correction using the ground-truth image. That's circular and unrealistic. PPISP directly predicts camera parameters from the rendered radiance alone, using an ISP controller trained on input views. ✨Result: metrics that don't require access to target images, aligning evaluation with real-world deployment. ➡️

NVIDIA AI Developer

12,013 görüntüleme • 8 ay önce

This is the power of VFX and AI now.. Creating volumetric effects in AI and then using AfterEffects and Premiere to layer and edit + Topaz Labs to upscale, you can visually achieve the same effects that would have taken me A LOT longer in Cinema4D/Blender. This is really amazing, only because if I can do this now with what we have, imagine where it will be in 6-mo 🔥 it will only get better. Soon we’ll be able to do a lot more within 3D environments with gaussian splats and volumetric effects. It won’t be long after that before Cinema, blender and others integrate gaussian splats and generative prompting directly into the software design and render process. At that point things are going to get REALLY BADASS! I’m chillin tonight, finally a night off lol. But tomorrow I’ll show you guys how I did this ❤️

This is the power of VFX and AI now.. Creating volumetric effects in AI and then using AfterEffects and Premiere to layer and edit + Topaz Labs to upscale, you can visually achieve the same effects that would have taken me A LOT longer in Cinema4D/Blender. This is really amazing, only because if I can do this now with what we have, imagine where it will be in 6-mo 🔥 it will only get better. Soon we’ll be able to do a lot more within 3D environments with gaussian splats and volumetric effects. It won’t be long after that before Cinema, blender and others integrate gaussian splats and generative prompting directly into the software design and render process. At that point things are going to get REALLY BADASS! I’m chillin tonight, finally a night off lol. But tomorrow I’ll show you guys how I did this ❤️

Dustin Hollywood

19,760 görüntüleme • 11 ay önce

After seeing some primitive but impressive examples earlier this week, I sent Astra on a goal to model our football stadium entirely in blender using reference pictures that I provided. Pretty crazy 🤯 There’s some clear issues to work out. But this was impossible for a model to do just days ago. It also animated the camera and rendered out this example video for me!

After seeing some primitive but impressive examples earlier this week, I sent Astra on a goal to model our football stadium entirely in blender using reference pictures that I provided. Pretty crazy 🤯 There’s some clear issues to work out. But this was impossible for a model to do just days ago. It also animated the camera and rendered out this example video for me!

Trey

21,544 görüntüleme • 1 ay önce

This is much worse for 3D artists than another AI image generator. Claude takes one architectural reference, connects directly to Blender, and starts rebuilding it as an actual editable scene. It blocks out the buildings, matches the camera, adds materials, vegetation, lighting, and then renders the result. The interesting part is not the first generation. Claude inspects its own render, notices the composition and lighting are wrong, then goes back into Blender and keeps changing the scene. The final result is still rough. The materials look basic, the geometry needs cleanup, and an experienced artist would make it far better. But AI is no longer producing a flat image and pretending it understands 3D. It is operating the software, judging its own work, and iterating inside the real project. The “AI cannot do their actual work” argument is getting harder to defend.

This is much worse for 3D artists than another AI image generator. Claude takes one architectural reference, connects directly to Blender, and starts rebuilding it as an actual editable scene. It blocks out the buildings, matches the camera, adds materials, vegetation, lighting, and then renders the result. The interesting part is not the first generation. Claude inspects its own render, notices the composition and lighting are wrong, then goes back into Blender and keeps changing the scene. The final result is still rough. The materials look basic, the geometry needs cleanup, and an experienced artist would make it far better. But AI is no longer producing a flat image and pretending it understands 3D. It is operating the software, judging its own work, and iterating inside the real project. The “AI cannot do their actual work” argument is getting harder to defend.

Rina

34,431 görüntüleme • 2 ay önce