Added a "Camera Preview" feature in Puppetry, to allow the user to see the live view from the scene Camera, while recording.
This is helpful in avoiding surprises after recording, for e.g. realizing that the movements weren't captured correctly by the scene Camera.
#blender#xr
Freebird v2.5.0 released - Adds the ability to add and edit `Text` while inside VR. This is useful for labeling and making notes inside VR, without having to sketch notes by hand. #vr#blender#3d#xr
@TSKelso @bojoku_66 @Starlink@SpaceX LL2 isn't 100% reliable with aborted launches, in my usage. But it's far better than doing it 100% manually (which I was doing until recently), and using it to complement my workflow than replacing it entirely
@TSKelso @bojoku_66 @Starlink@SpaceX I use the Launch Library 2 API for auto-updating launches on my site (https://t.co/Pv8H1BM4kG). Not sure if this helps you, but it could serve as another source of data for ref. I still manually check every launch. LL2 got the G9-13 launch correct - https://t.co/ewd1aLbLiz
One month since Freebird v2 was launched, so here's a quick recap of what's happened with @freebirdxr since then, and what's coming up!
https://t.co/y1cprhqIQt
#VR#Blender#3D#animation#3dmodeling
Freebird 2.0.2 is now available! You can now pose and edit armature bones in #Blender using #VR. It's really easy to grab and pose your characters into natural-looking poses. Feedback welcome on Discord!
Get started at https://t.co/d8xJlkGYLE
#3D#Modeling#Animation
Easy Diffusion 3.0 is live, on it's first birthday!
Major features: SDXL, ControlNet, multiple LoRAs, Embeddings, seamless tiling. And significantly faster and more optimized for GPU memory than v2.0.
https://t.co/J1DSht08Za
#StableDiffusion#AI#LLM#AIArtworks
Added support for SD-XL 1.0 (base and refiner models) to Easy Diffusion (beta).
Supports text-to-image, image-to-image and inpainting, with VRAM optimizations to run on lower-end PCs.
Available at https://t.co/45Fj3mMo6l
#stablediffusion#AIArtwork#AI
Might be interesting to use RealESRGAN at runtime (maybe during scene load) to upscale textures close to the camera (esp in VR). Texture resolution vs disk space is always a tradeoff, and this might help smaller-sized textures look sharper at runtime. Maybe via PyTorch Mobile?
What exactly counts as "large", for a model's training to require approval? Is it a trillion parameters?
Why that arbitrary number? What if tech improves & it can be trained cheaply?
And does fine-tuning existing large models count? Can already be done in a few hours on PCs.