ComfyUI-anima-Resolutions Introduction
<p align="center">
<img src="assets/icon.png" alt="Anima Resolutions icon" width="200">
ComfyUI-anima-Resolutions is a small collection of ComfyUI nodes that helps you manage image dimensions in an Anima workflow. You can use it to choose a starting canvas size, keep detailer crops aligned to suitable dimensions, and resize an image while making sure its final width and height are divisible by a chosen number.
These nodes solve different parts of the same practical problem: image sizes often need to fit a workflow’s requirements, but calculating and adjusting them by hand can interrupt your creative process. The extension does not load a model, change sampler settings, or generate images on its own.
How ComfyUI-anima-Resolutions Works
Think of the extension as three tools for three stages of a workflow:
- Choose a canvas size. Anima Resolutions provides preset widths and heights. Connect its outputs to nodes that accept dimensions for your first generation pass. It supplies numbers; it does not resize an existing image.
- Align detailer crops. If you use a compatible ComfyUI Impact Pack detailer, Anima Detailer Align Hook rounds the crop’s sampling width and height up to a selected multiple. For example, with an alignment of 32, a dimension of 1000 would round up to 1024.
- Scale a generated image. Anima Image Scale By Multiple resizes an image by your chosen factor, then adjusts its dimensions to be divisible by 8, 16, 32, or 64. This means the final size may differ slightly from simply multiplying the original width and height.
You can use the resolution and scaling nodes without Impact Pack. Only the detailer hook depends on it.
ComfyUI-anima-Resolutions Features
Anima Resolutions: choose a first-pass size
Select a 1024, 1280, or 1536 resolution preset, then choose a shape suited to your composition: square, portrait, landscape, or cinematic wide. The node outputs resolution, width, and height as integers for other nodes to use.
For most Anima workflows, the author recommends starting at roughly 1 to 1.6 megapixels for the first pass. Larger images can work, but they take longer and may require more VRAM (graphics-card memory).
| Starting size | Example use | Approximate pixels |
|---|---|---|
| 1024×1024 | Square or general compositions | 1.05 MP |
| 896×1344 | Portrait | 1.20 MP |
| 1024×1536 | Larger portrait | 1.57 MP |
| 1344×896 | Landscape | 1.20 MP |
| 1536×1024 | Wider landscape | 1.57 MP |
| 1536×864 | Cinematic wide composition | 1.33 MP |
Available presets
| Resolution group | Sizes |
|---|---|
| 1024 | 1024×1024 (1:1); 1152×896 (9:7); 896×1152 (7:9); 1152×864 (4:3); 864×1152 (3:4); 1344×896 (3:2); 1248×832 (3:2); 896×1344 (2:3); 832×1248 (2:3); 1280×720 (16:9); 720×1280 (9:16); 1344×576 (21:9); 576×1344 (9:21) |
| 1280 | 1280×1280 (1:1); 1440×1120 (9:7); 1120×1440 (7:9); 1472×1104 (4:3); 1104×1472 (3:4); 1536×1024 (3:2); 1024×1536 (2:3); 1536×864 (16:9); 864×1536 (9:16); 1680×720 (21:9); 720×1680 (9:21) |
| 1536 | 1536×1536 (1:1); 1728×1344 (9:7); 1344×1728 (7:9); 1728×1296 (4:3); 1296×1728 (3:4); 1872×1248 (3:2); 1248×1872 (2:3); 2048×1152 (16:9); 1152×2048 (9:16); 2016×864 (21:9); 864×2016 (9:21) |
The group name identifies a set of presets; the exact width and height depend on the shape you select.
Anima Detailer Align Hook: align crop sampling sizes
A detailer works on selected parts of an image, such as a crop around a subject. Connect this node’s detailer_hook output to the matching input on an Impact Pack detailer node, such as Detailer For Each. The hook rounds the detailer’s crop sampling width and height up to your selected alignment.
Use 32 for Anima, as recommended by the author. This feature requires ComfyUI Impact Pack; it is not needed for the other two nodes.
Anima Image Scale By Multiple: resize and align an image
Connect an IMAGE, choose a scale factor, and select a dimension multiple of 8, 16, 32, or 64. The node resizes the image and ensures that both final dimensions are divisible by that multiple.
For example, scaling a 1024×1536 image by 1.5× produces 1536×2304. If a calculated size does not meet the selected multiple, the node snaps the dimensions to fit. It also outputs the final width, height, and applied_scale—the actual average scale after that adjustment.
Set max_long_edge to 0 if you do not want a longest-edge limit. The node uses ComfyUI resize methods: nearest-exact, bilinear, area, bicubic, and true Lanczos. You can choose a method to compare how resized edges and details look in your images.
Troubleshooting ComfyUI-anima-Resolutions
The detailer hook does not connect to my detailer. What should I check?
- Confirm that you are using a ComfyUI Impact Pack detailer with a matching
detailer_hookinput, such as Detailer For Each. - Connect the Anima Detailer Align Hook’s
detailer_hookoutput to that input. - Make sure ComfyUI Impact Pack is available in your ComfyUI setup. The hook requires it; the other nodes do not.
Why did Anima Resolutions not change my existing image?
It only outputs dimension values. Use those values with nodes that accept a width and height for your first pass. To resize an existing IMAGE, use Anima Image Scale By Multiple.
Why is my resized image slightly different from the size I calculated?
The scale node adjusts the result so that both dimensions are divisible by your selected multiple. Check its final width and height outputs for the exact size, and applied_scale for the actual average scaling.
Why is my workflow slower or using more VRAM?
Larger images contain more pixels. Try a first-pass preset closer to the author’s recommended 1–1.6 MP range, or use a smaller scaling factor. Remember that an optional 1.5× scale makes the final image larger than its first-pass size.
Learn More about ComfyUI-anima-Resolutions
- ComfyUI Impact Pack: reference for the optional detailer nodes used with Anima Detailer Align Hook.
- Project test status: view the repository’s automated test results.
