We now support the following models for video generation in Morph Studio:
Model Name | Description |
---|---|
Text-to-Video-Morph-0.1 | Our in-house model for text-to-video generation, our text-to-video model by default has a realistic tone, feel free to add descriptive adjectives to modify the style and looks of your shot. |
Image-to-Video-Morph-0.1 | Our in-house model for image-to-video generation, the generation result is better if you guide it with a text prompt, but text entry when using image-to-video is not required. Our model performs better compared to SVD when it comes to close up of characters and animation of objects. |
Video-to-Video-Morph-0.1 | Our in-house style transfer models support video-to-video renders guided by a text prompt. The primary function of this model is to change the style of your videos to a preset of our own, while keeping the characters and layout of the original video. |
Image-to-Video-SVD-1.1 | This is an image-to-video model powered by Stability.ai, and is best used for shots where there are clear layers within the composition (scenery shots, B-rolls, etc) |
Changing the model:
Text-to-Video-Morph-0.1 is set to be the default model of any newly created shot card. Your choice of model will be changed depending on if you have uploaded an image or video. Each model also has its own set of parameters available for adjustments as shown below:
Model Settings
Model | Text Prompt | Camera | Duration | Motion | Image Attachment | Video Attachment | Motion Brush |
---|---|---|---|---|---|---|---|
Text-to-Video-Morph-0.1 | ✔ | ✔ | ✔ | ✔ | - | - | - |
Image-to-Video-Morph-0.1 | ✔ | ✔ | ✔ | ✔ | ✔ | - | ✔ |
Video-to-Video_Morph-0.1 | ✔ | - | - | - | - | ✔ | - |
Image-to-Video-SVD-1.1 | ✔ | - | ✔ | ✔ | - | - | ✔ |
Setting Description
Camera:
right now we support static, zoom in, zoom out, pan left, pan right, pan up, pan down, rotate clockwise, and rotate counterclockwise. The option to choose multiple camera movements at once will be available in the future.
Duration:
We support up to 10s of video generation. The default duration is set to 3s.
Motion:
We support a range from 1-10 for motion values; the minimum and default is 1, maximum is 10; the higher the number, the more dramatic the motion.
Image Attachment (Image):
You can upload your images by clicking on the paperclip button. Please note that the highest resolution of videos generated as of now is 1080p. A slightly more high-res image is recommended to retain more details from the image, but please keep it below 4k so that the server does not get overloaded.
Video Attachment (Video):
You can upload your video also using the paperclip button for Style Transfer. The limit on the size of your attachment is 50M.