Morphic: Generate consistent character scenes

Demo summary
The user selects the 'Jack' character model and generates multiple scenes of him eating, running, and playing volleyball while maintaining visual consistency.
Step-by-step
- Select the desired character model (e.g., 'Jack') from the dropdown menu
- Set the generation mode to the default setting
- Enter a prompt describing the character performing an action
- Click to generate the image
- Review the generated scenes and select the best result
Options
- Generate videos instead of static images
- Input various action prompts like eating, running, or playing sports
Tips
- Use the specific character model selection to ensure visual consistency across different prompts
Highlights
“he does stay consistent across the prompts you do generate with the character model”
All demos from “Morphic | Walkthrough”
2:111:27Generate images in different styles with MorphicThe presenter demonstrates how to use the prompt bar to generate images of a monkey in photorealistic, anime, and 3D kid styles.Morphic· Text to Image
3:561:18Train custom style and character modelsA walkthrough of the Morphic dashboard showing how to upload reference images to train a specific 'pixelated' style and a consistent character named Jack.Morphic· AI Avatar Generator
5:191:09Generate consistent character scenesCurrentThe user selects the 'Jack' character model and generates multiple scenes of him eating, running, and playing volleyball while maintaining visual consistency.Morphic· AI Person Generator
6:391:23Edit image frames using text promptsMorphic's frame editing is shown by prompting changes to an existing image, such as changing camera angles to close-ups or adding sunglasses to a character.Morphic· AI Inpainting
8:081:45Extract and add layers with Object SelectionThe demo shows how to use Object Selection to extract a branch as a separate layer and Region Selection to add a new mosquito element into a frame.Morphic· AI Inpainting
10:220:42Create a character model from a single imageThe presenter demonstrates the 'Create Model' feature which auto-generates profile angles and close-ups from one single reference image of a woman.Morphic· AI Avatar Generator
12:100:57Animate images into video clipsThe user demonstrates converting a static frame of a scuba diver into a video by describing the desired motion in the video prompt bar.Morphic· Image to Video
13:480:48Create transition videos from keyframesThe demo shows selecting two different frames and using the 'Create Video' tool to generate a transition shot where a bike emerges through a window.Morphic· Image to Video
15:040:30Perform video-to-video editsThe presenter applies a text prompt to an existing video to change the environment to a night scene with heavy rainstorms.Morphic· Video to Video
15:350:43Generate multi-cut sequences with textThe user demonstrates prompting a single video generation with 'cut shot' instructions to create a sequence of a boy walking, a close-up, and a jump.Morphic· Text to Video
16:180:33Generate AI video with lip sync and audioThe demo shows how to use the V3 model and quotation marks in a prompt to generate a character speaking specific dialogue with synchronized lip movement.Morphic· AI Lip Sync Generator
16:511:09Direct 3D motion and camera pathsThe presenter uses the 3D motion tool to set keyframes and direct camera movement (zooming and panning) over a static image to create a 3D visual effect.Morphic· AI Animation Generator- Watch “Morphic | Walkthrough” →
AI Person Generator
5:191:09Generate consistent character scenesCurrentThe user selects the 'Jack' character model and generates multiple scenes of him eating, running, and playing volleyball while maintaining visual consistency.Morphic
7:545:01Create consistent AI characters in MorphicThe user demonstrates how to generate a character image and use the 'Character Model' feature to create a 360-degree reference for maintaining visual consistency across shots.Vishnu Vijayan
13:012:53Generate images using character tagsThe creator shows how to use the '@' symbol to tag previously created character models (Devon, Mother, Father) within new image prompts to place them in a carnival scene.Vishnu Vijayan