Sample work · Video & Animation
Animating existing character: from layers to speaking character
A finished character in layers is automatically rigged with joints and animated: breathing, blinking, waving, walking, and speaking. The drawing itself remains unchanged.
Sample work (demo), not a client job. Built as a trial task with invented example data; the check figures below come from the automatic test run and are not polished.
What was built
- Joints from layer geometry: shoulder, elbow, hip, knee, and neck
- Waving, walking, blinking, and breathing, parts stay connected
- The mouth follows the volume of the voice track
2D skeletal animation from layers (SVG groups or PNG per part), no frame-by-frame cartoon, no 3D, AI voice.
How it was checked
Automatic test run on 2026-11-01, taken from the log (test data invented, figures measured):
- SVG groups + PNG parts: fidelity SSIM 0.9939, gap 1.14px, mouth r=0.978
- ONE flat PNG each -> Nico 14/Robbi 14 parts, rest pose SSIM 0.9972/0.998, gap 1.57/1.8px, hole 0.101/0.0%
- Client audio .mp3 9.557s r=0.999
- 0 EUR
Limits
What this sample does not cover, so nobody expects more:
- Inputs: SVG with named groups, one PNG per part OR (new) ONE flat image that is first automatically disassembled; please export PSD/AI as SVG/PNG; part names head, torso, arm_l_upper ... mouth
- FLAT IMAGE: only frontal characters in rest pose (arms hanging, legs side by side), visible neck, recognizable edge between arm and torso; only tested on 2 characters (Nico, Robbi), unfamiliar styles ...
- FLAT IMAGE: Visual inspection 01.11.2026 - small contour remnants/fringes remain on the raised arm and notches on the torso edge (below measurement threshold); rework by hand before delivery or the ...
- FLAT IMAGE: Hidden parts are REPLACED, not drawn (color continuation under arm/eyes), no real inner life; hair/cap/ears rigid in the head; there is no finger movement
- Eyes/mouth: separated from the image if there are two spots of equal height + one lower central spot in the face; otherwise simple line replacement layer (skin tone/line color, not the painted mouth) ...
- Background removal: background color or alpha; cluttered background via rembg (measured only with Nico, IoU in zahlen.rembg_iou), shadows/floor in the image become part of the character
- CLIENT AUDIO: Stand-in was a synthetic Kokoro voice (English, am_michael) as MP3; real client recordings (room, noise, pauses, dialect) not tested; mouth follows the volume, not the ...
- 2D skeletal animation (cut-out), no frame-by-frame, no 3D, no perspective changes; joint gap and movement measured from browser matrices, holes from the images, not by eye
- Hole check: spot checks at the inner seam points of each joint (distance between parent and child position, no background allowed); gaps at the outer edge (elbow bend, leg gap during ...)
Something like this for your business?
Describe briefly what you need. You get a fixed price before I start; nothing happens without your yes.