Upload 2088559/2363046/README.md with huggingface_hub
Browse files- 2088559/2363046/README.md +54 -2
2088559/2363046/README.md
CHANGED
|
@@ -2,9 +2,61 @@
|
|
| 2 |
license: other
|
| 3 |
tags:
|
| 4 |
- civitai
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 6 |
Author: [neph1](https://civitai.com/user/neph1)
|
| 7 |
|
| 8 |
-
Model: [
|
|
|
|
|
|
|
| 9 |
|
| 10 |
-
|
|
|
|
| 2 |
license: other
|
| 3 |
tags:
|
| 4 |
- civitai
|
| 5 |
+
- concept
|
| 6 |
+
- cinematic
|
| 7 |
+
- scene
|
| 8 |
+
- cut
|
| 9 |
+
- transition
|
| 10 |
---
|
| 11 |
+
|
| 12 |
+
# Cinematic Hard Cut - 2363046
|
| 13 |
+
|
| 14 |
+
**Model Type**: LORA
|
| 15 |
+
|
| 16 |
+
**Base Model**: Wan Video 2.2 I2V-A14B
|
| 17 |
+
|
| 18 |
+
**Trigger Words**: None
|
| 19 |
+
|
| 20 |
+
**Tags**: concept, cinematic, scene, cut, transition
|
| 21 |
+
|
| 22 |
+
## Description
|
| 23 |
+
|
| 24 |
+
V3 update: Included smash cut type transitions. "The camera makes a smash cut..."
|
| 25 |
+
|
| 26 |
+
As much as I love Wan, if it's one thing it's not good at, it's cuts.
|
| 27 |
+
|
| 28 |
+
This lora is an experiment in greater shot control, by being trained on hard cuts. Most of the clips are also cut on action, fwiw, to allow for a continuous flow.
|
| 29 |
+
|
| 30 |
+
It's captioned with "wide-angle", "mid-shot" and "close-up". But I think Wan already handles them well.
|
| 31 |
+
|
| 32 |
+
Prompt format:
|
| 33 |
+
|
| 34 |
+
[a brief description of your initial shot.]
|
| 35 |
+
|
| 36 |
+
the camera makes a hard cut to [resulting shot]
|
| 37 |
+
|
| 38 |
+
[what happens after]
|
| 39 |
+
|
| 40 |
+
You can string several cuts. If you do, it works best if you also change the type of shot.
|
| 41 |
+
|
| 42 |
+
Where's the low noise model? There isn't one. It seems to work well with only the high noise model. I'll still train and test a low noise model, to see if it will improve consistency during cuts.
|
| 43 |
+
|
| 44 |
+
Play with strength if the scene changes too much. Too low strength will make it a dissolve transition (at least testing seems to imply that).
|
| 45 |
+
|
| 46 |
+
Using lora's (for the target cut) you can get pretty far.
|
| 47 |
+
|
| 48 |
+
If the cut doesn't 'take':
|
| 49 |
+
|
| 50 |
+
* Try using less other lora's.
|
| 51 |
+
* If you use a high resolution, lower it.
|
| 52 |
+
* Try generating a shorter clip.
|
| 53 |
+
|
| 54 |
+
---
|
| 55 |
+
|
| 56 |
Author: [neph1](https://civitai.com/user/neph1)
|
| 57 |
|
| 58 |
+
Model: [CivitAI Model Page](https://civitai.com/models/2088559?modelVersionId=2363046)
|
| 59 |
+
|
| 60 |
+
Archive: [CivArchive Page](https://civarchive.com/models/2088559?modelVersionId=2363046)
|
| 61 |
|
| 62 |
+
<!-- Version: 20260422_v4_update -->
|