- Added resolution aspect ratios to SDXL quick latent/quick resolution nodes
- Added fallback comment for empty prompt info when writing image info (to prevent the question of if it was a failed write or if there was just no prompt ever)
- Fixed imports based on changes in comfy
- Added additional resolution to ratio-based nodes.
- Added width and height to outputs of Image Scale by Shortside
- Fixed issue with Image Save with Prompt/Info & File node
- Fixed issue with integer width/height return from quick latent nodes.
- Added Image Scale by Shortside node
- Added unified SD1.5/SDXL CLIP-Text node
- Added Prompt Weight node
- Added SDXL workflow
- Added string combine node
-
Added image preprocessing node to make images grayscale
-
Fixed "Image Upscale by Factor" node logic
-
Added non-INFO variant of file IO nodes to quick-fix a compatibility issue that seems to crop up with the TTN suite
-
Renamed "Read Image with Prompt" to "Image Load with Metadata" and moved it to the 'image' category
-
Re-ordered node loading to make categories of WLSH submenu be in a more reasonable order
-
Redo of README for readability
-
Updated workflow png
-
Added "Empty Latent by Size" node to set latent size directly but return width and height
-
Added "Quick Resolution Multiplier" node to quickly multiply an input height and width (for SDXL)
-
Added 19:9 aspect ratio to empty latent by ratio node (and others)
-
Added width/height output to empty latents (so you don't need a separate node)
-
Added "Empty Latent by Pixels" which lets you set aspect ratio and megapixels total
-
Moved the resolution number nodes to a numbers category
- Added 'Simple Pattern Replace' node
Changed input types for several nodes to default to 'input' instead of 'widget'. For example, the file save nodes won't have prompt text boxes.
-
Updated Read Prompt Data From Image node. Using comfyUI commit d7b3b0f8c11c6261d0d8b859ea98f2d818b7e67d, allows for preview of selected image.
-
SDXL version of CLIP Positive-Negative nodes.
-
Image save nodes preview. Now shows previews of images saved in image output nodes
-
Added "Read Prompt Data from Image" node. Use this to read various metadata from Auto1111 or from images saved with tthe prompt medatada from this node pack.
-
Changed image metadata save format Modified the part of the image the metadata is saved to for png types when using the various image output/prompt output nodes
-
Added INFO as output to custom KSamplerAdvanced INFO is a dictionary containing a bunch of the configuration data for that sampler node.
-
Added INFO input to Image/Prompt save nodes Can be used to append more information to the metadata.
-
Converted SEED input types to INT Fixed need for
SEED
type inputs on sampler and file saves. May require re-routing old workflows.