Animation of a dialogue between two characters based on a ready image
Important note for the performer: This video is a pilot. In case of successful execution and long-term alignment on quality/deadlines, we are looking for an author for ongoing monthly collaboration. The project is long-term, with plans for regular release of similar videos.
1. General description of the task
It is necessary to "bring to life" two animated characters from a static image.
The characters should engage in a dialogue based on the provided audio track/text.
The animation style should be restrained, natural, with an emphasis on precise lip synchronization (Lip Sync).
The use of AI tools (LivePortrait, Hedra, SadTalker, etc.) is encouraged.
2. Source materials
Visual reference: Image image.png / image.psd
Audio materials: Voiceover of the dialogue
Text materials: Dialogue script with time codes
3. Technical requirements for the finished video
File format: MP4 (codec H.264).
Resolution: 1920×1080 (Full HD, 16:9)
Frame rate: 24 or 25 FPS.
Timing: Corresponds to the duration of the audio track [2 min].
4. Animation requirements (Setup and Dynamics)
The animation should be minimalist, without sharp and exaggerated movements.
Facial animation (Facial expressions and Lip Sync):
The main priority is the accurate matching of lip movements with the spoken text (articulation should look natural).
Minimal eyebrow movement to convey basic emotions (surprise, agreement, emphasis on an important word).
Natural blinking of the eyes (not too frequent, random).
Body movements and gestures:
Minimal head nodding in sync with speech or as a sign of agreement/disagreement with the interlocutor.
Micro-movements of the shoulders during inhalation/position change.
Hand gestures should be rare and justified (for example, a light hand gesture to emphasize a thought). In static moments, the hands rest.
Interaction:
The characters' gazes during speech should be directed at each other or slightly at the viewer (depending on the composition of the original frame).
The listening character should not "freeze": a reaction to the speaker's words is necessary (a slight nod, facial response).
5. Approval and revision process
Stage 1: Approval of the layer cuts (if separation of characters from the background is required) and animatic (rough sketch of movements).
Stage 2: Demonstration of the final rendering.
Appendix: Dialogue script (Example of filling)
0:00 — 0:05 | Character 1 (left): "Hi! Have you seen the new graphics?" (Slight nod, neutral-friendly facial expression).
0:05 — 0:12 | Character 2 (right): "Yes, and to be honest, they scare me a little." (Raises eyebrows, slightly tilts head to the side).
Mandatory use of specialized tools such as Runway Gen-2, Luma Dream Machine, or Pika,
and for precise matching of speech and lips (Lip Sync) use tools like SadTalker, LivePortrait, or HeyGen.
Google Drive
