String Resolution Seedance 2.5 Multimodal Video - Seedance 2.5 Video Resolution Selector
Node Purpose: String Resolution Seedance 2.5 Multimodal Video outputs the nearest Seedance 2.5 resolution tier label (480p / 720p / 1080p / 2k / 4k) based on the input video's pixel area. If no video is connected, it passes through the manually selected label. It is designed to pair with RunningHub Seedance 2.5 Multimodal Video in ComfyUI-1hewOffice (nodes/api/runninghub/runninghub_seedance_25_multimodal_video.py).
Inputs
| Name |
Port |
Type |
Default |
Range |
Description |
selection |
- |
COMBO |
480p |
480p / 720p / 1080p / 2k / 4k |
Manual resolution label used when video is not connected. |
video |
optional |
VIDEO |
- |
- |
Input video used to infer the nearest resolution tier. |
Outputs
| Name |
Type |
Description |
string |
STRING |
Resolution label string, ready to connect to an API node's resolution input. |
Tiers and target areas
| Tier |
16:9 standard size |
Target area (pixels) |
480p |
854x480 |
409,920 |
720p |
1280x720 |
921,600 |
1080p |
1920x1080 |
2,073,600 |
2k |
2560x1440 |
3,686,400 |
4k |
3840x2160 |
8,294,400 |
Features
- Tiers:
480p / 720p / 1080p / 2k / 4k, matching the usual resolution tiers of the Seedance 2.5 multimodal video node (native1080p is not part of the inference).
- Area-based matching: compares the video's total pixel area against the tier target areas and picks the nearest in log space, consistent with the image
String Resolution nodes in this family.
- Portrait-safe: matching is by area, not by a single side, so a
1080x1920 video is still classified as 1080p.
- Manual fallback: returns the selected label directly when no video is connected or its size cannot be read.
Typical Usage
- Feed the input video to both this node and the Seedance 2.5 multimodal video node's reference-video input, then connect this node's output to
resolution so the generated resolution tracks the source.
- Leave
video unconnected to use it as a fixed resolution string selector.
Notes & Tips
- Matching is based on total pixel area, not exact width/height ratio.
- Target areas use each tier's 16:9 standard size;
2k is based on 2560x1440.
- This node only outputs the string label: it does not resize the video and performs no network requests.
