Added symbolic link feature so you no longer have to duplicate your models from your web UI!
DiffEx is an easy to use Desktop UI for AnimateDiff CLI Prompt Travel. It takes out a lot of the headache and labor needed to create animations by doing most of the work for you. A detailed readme is included to get you up and running!
New Features in v1.5
Console-less! A cool new progress bar has taken the place of the console for a less cluttered user experience.
Folder walking! Any sub folders in your Stable Diffusion directories will now be detected and added to DiffEx.
Added a refresh button so you no longer have to restart DiffEx after downloading new models.
Added the ability to use a video as an IP-Adapter for better coherence especially when using Stylize.
Added the Canvas to the Region tab so you can now draw your masks inside of DiffEx.
Added a Separate feature to Regions for better layer separation in the final output.
Added support for Ckpt and Pt files.
More bug fixes and optimizations of course.
Stylize mode added (vid2vid) DiffEx will automatically create prompts out of your videos. You can select them from your hard drive or automatically download them using yt-dlp (built in).
Added AnimateDiff and Mediapipe Face ControlNets.
Added the Refiner to upscale your animations.
Added the ability to restrict ContolNets to specific Regions.
Added Context Scheduling options for more stable animations.
Redesigned ControlNet image mapping interface.
Tooltips added to various controls to help newer users.
Regions! You can have up to 3 regions including the background (root prompt). This allows you to separate your animation into different layers allowing you greater control over the output.
The new MaskMaker allows you to automatically generate masks from videos for your regions.
LoRA mapping allows you to isolate LoRAs to a specific region and ramp its strength along the duration of your animation.
You can now paste prompt maps (ex. generated in ChatGPT) directly into the prompt grids.
QR Code Monster v1 and v2 added.
LCM support added.
SDXL support added.
Lots and lots of bug fixes and optimizations.
Automatically imports your Models, Motion Modules, VAEs, LoRAs and Embeddings.
IP Adapter support
Full multi ControlNet support.
Easily import video and images to create your ControlNet image map using drag and drop.
Diffex will automatically copy and sequence all of your images/videos and put them in their necessary folders.
adetailer integration to fix faces
EXIF data reader so you can import images you made in A1111 (and possibly Comfy)