This is an I2V workflow created for Wan2.2.
I just made something similar. The usage is almost the same.
But, Since the video deteriorates with each loop, I reduced the degree of deterioration by setting “end_image.” in this "FLF2V Version"
As a result, the prompt will be written as follows.
duration (seconds) | character action | camera work | filepath-of-end_image
(separate each item with a “|” symbol)
When using this feature, the end images of each loop that will serve as keyframes must be prepared in advance.
We have confirmed that the quality of the generated video deteriorates when the frame rates of KSampler's input and output are different.
In some cases, there is no problem. Examples are as follows.
1 sec 30 fps -> Input: 31 - Output: 29 (Mismatch)
2 sec 30 fps -> Input: 61 - Output: 61
3 sec 30 fps -> Input: 91 - Output: 89 (Mismatch)
Therefore, we have implemented a mechanism to adjust the output frame count per loop.
Go ahead and upload yours!
Your query returned no results – please try removing some filters or trying a different term.