Skip to content

Commit 8974ebc

Browse files
Add files via upload
0 parents  commit 8974ebc

File tree

4 files changed

+574
-0
lines changed

4 files changed

+574
-0
lines changed

README.md

Lines changed: 198 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,198 @@
1+
# Image Metadata Editor Forge
2+
3+
A simple & user-friendly extension for Stable Diffusion WebUI Forge that lets you view, edit, add & manage metadata in your AI-generated images.
4+
5+
## Features
6+
7+
![Image Metadata Editor in Forge](assets/ui.png)
8+
9+
This extension helps you:
10+
- **View metadata** stored inside your AI-generated images (like prompts, settings, etc.)
11+
- **Edit existing metadata** or add new info to images
12+
- **Save images** w/ updated metadata
13+
- **Remove all metadata** from images if needed
14+
- **Add custom info** like artist name, notes, or categories
15+
16+
Think of metadata as invisible info stored inside your image files - like the recipe used to create the AI image.
17+
18+
## Installation (Step-by-Step)
19+
20+
### For Complete Beginners:
21+
22+
1. **Find your Forge folder**: `stable-diffusion-webui-forge` on your computer
23+
24+
2. **Navigate to extensions**: Inside that folder, find & open the `extensions` folder
25+
26+
3. **Download this extension**:
27+
- **Easy way**: Download as ZIP from [release page](), extract it, & put the folder in `extensions`
28+
- **Git way** (if you have Git installed): Open command prompt/terminal in the extensions folder & run:
29+
```
30+
git clone https://github.com/MackinationsAi/sd-forge-metadata-editor.git
31+
```
32+
33+
4. **Restart Forge**: Close Stable Diffusion WebUI Forge completely & start it again
34+
35+
5. **Find the extension**: Look for "Image Metadata Editor" section in your txt2img or img2img tabs (it will be collapsed/closed by default - click to open it)
36+
37+
### Basic Workflow
38+
39+
#### 1. **Upload an Image**
40+
- Click the "Source" image upload area
41+
- Select any image file (PNG, JPEG, etc.)
42+
- Metadata is automatically read & displayed
43+
44+
#### 2. **View Existing Metadata**
45+
The extension automatically displays:
46+
- **Generation Info Tab**: Shows parsed metadata in readable format
47+
- **Nothing found in the image** = No metadata detected
48+
- **Parsed parameters** = Shows prompt, negative prompt, steps, etc.
49+
50+
#### 3. **Edit Parameters**
51+
In the "Generation Info" tab, you can edit:
52+
- **Prompt**: Main generation prompt
53+
- **Negative Prompt**: Negative prompt text
54+
- **Steps**: Number of sampling steps (1-150)
55+
- **CFG Scale**: Classifier guidance scale (1.0-30.0)
56+
- **Sampler**: Sampling method (e.g., "DPM++ 2M")
57+
- **Schedule type**: Sampling schedule (e.g., "Karras")
58+
- **Model**: Model name used
59+
- **Seed**: Generation seed (-1 for random)
60+
- **Size-1 & Size-2**: Image dimensions (width x height)
61+
62+
#### 4. **Add Custom Metadata** (Optional)
63+
In the "Custom Fields" tab:
64+
65+
**Method A - Using Add Button:**
66+
1. Enter a key (e.g., "artist")
67+
2. Enter a value (e.g., "MyUsername")
68+
3. Click "Add" button
69+
4. The key-value pair is added to the image’s metadata JSON
70+
71+
**Method B - Direct JSON Editing:**
72+
1. Edit the JSON directly in the text area:
73+
```json
74+
{
75+
"artist": "MyUsername",
76+
"category": "portrait",
77+
"rating": "5",
78+
"notes": "My favorite generation"
79+
}
80+
```
81+
82+
#### 5. **Save Your Image**
83+
- Click **"Save w/ Metadata"**
84+
- The image appears in "Output Image" tab
85+
- Success message shows: "✅ Successfully saved metadata to image!"
86+
- Go to the Output Image tab to download the image
87+
88+
### Advanced Features
89+
90+
#### **Raw Metadata Tab**
91+
- View complete metadata as JSON
92+
- Manually edit any metadata field
93+
- See exactly what's stored in the image
94+
- Useful for debugging or advanced editing
95+
96+
#### **Clear Metadata**
97+
- Click "Clear Metadata" to remove ALL metadata
98+
- Useful for creating clean images for sharing
99+
- Cannot be undone, so save original first
100+
101+
## Troubleshooting
102+
103+
### "Nothing found in the image"
104+
- This means your image has no metadata (completely normal for photos or edited images)
105+
- You can still add metadata & save it w/ that new info
106+
107+
### Extension Not Showing Up
108+
1. Make sure the files are in the right place: `extensions/sd-forge-metadata-editor/scripts/metadata_editor.py`
109+
2. Restart Forge completely (close & reopen)
110+
3. Check the console for error messages
111+
112+
### Can't Save Images
113+
- Make sure you have permission to write files where Forge is installed
114+
- Try w/ a simple prompt first
115+
- Check if your antivirus is blocking file creation
116+
117+
### Custom Fields Not Working
118+
- Make sure your JSON format is correct (the extension will show errors)
119+
- Use simple text w/o special characters when starting out
120+
121+
## FAQ
122+
123+
**Q: Will this work w/ images from other AI tools?**
124+
A: Yes, if they save metadata in standard formats. You can also add metadata to any image.
125+
126+
**Q: Can I use this to organize my image collection?**
127+
A: Yes! Add custom fields like "category", "rating", or "notes" to help organize your images.
128+
129+
**Q: Does this change my original image?**
130+
A: No, it creates a new copy w/ the metadata. Your original stays unchanged. So you have to download the new image w/ metadata to see the changes.
131+
132+
**Q: Why save as PNG instead of JPEG?**
133+
A: PNG preserves metadata much better than JPEG, ensuring your info doesn't get lost.
134+
135+
## How It Works
136+
137+
### Compatibility
138+
-**Stable Diffusion WebUI Forge**: Primary target platform
139+
-**AUTOMATIC1111**: Can read A1111 metadata format
140+
-**Other SD Tools**: Standard metadata format ensures compatibility
141+
142+
### File Structure
143+
```
144+
sd-forge-metadata-editor/
145+
├── README.md # This documentation
146+
├── assets/
147+
| ├── ui.png # UI Example
148+
| └── input_test_image.png # Input Image Example
149+
└── scripts/
150+
└── metadata_editor.py # Main extension code
151+
```
152+
153+
1. **Reading**: Uses `images.read_info_from_image()` (same as Forge's PNG Info tab)
154+
2. **Parsing**: Extracts individual parameters from metadata strings
155+
3. **Editing**: Provides user-friendly form inputs for common parameters
156+
4. **Custom Fields**: JSON-based storage for additional metadata
157+
5. **Saving**: Embeds metadata as PNG text chunks for maximum compatibility
158+
159+
### Adding Metadata to Raw Image
160+
1. Upload a photo or image w/ or w/o metadata
161+
2. Fill in Generation Info fields:
162+
- Prompt: "beautiful landscape, mountains"
163+
- Steps: 20
164+
- CFG Scale: 7.0
165+
- Sampler: "DPM++ 2M"
166+
3. Add custom fields:
167+
```json
168+
{"source": "original photo", "edited": "2024"}
169+
```
170+
4. Click "Save w/ Metadata"
171+
172+
### Can Read Metadata From:
173+
- PNG files w/ embedded metadata
174+
- JPEG files w/ EXIF data
175+
- Images created by Stable Diffusion, AUTOMATIC1111, Forge & ComfyUI
176+
177+
### Supported Formats
178+
- Can read metadata from: PNG (text), JPEG (EXIF), SD formats (Forge, A1111, ComfyUI)
179+
- Saves as: PNG (best for preserving metadata)
180+
181+
## Contributing
182+
183+
This extension is open source. Feel free to:
184+
- Report bugs or issues
185+
- Suggest new features
186+
- Submit pull requests
187+
- Share usage examples
188+
189+
## License
190+
191+
This extension is available under the MIT License.
192+
193+
---
194+
195+
**Version**: 1.0
196+
**Author**: MackinationsAi
197+
**Compatibility**: Stable Diffusion WebUI Forge
198+
**Last Updated**: May 2025

assets/input_test_image.png

7.2 MB
Loading

assets/ui.png

508 KB
Loading

0 commit comments

Comments
 (0)