After Effects Plugin for 3D Portrait
StyleFrame Team8 min read
Face 3D from aescripts is the most direct After Effects plugin for 3D portraits. It uses AI depth maps to turn a single photo into an animatable 3D model inside After Effects, and it needs After Effects 24.2 or later with the Advanced 3D Renderer. If you prefer a template-driven parallax approach, Volumax builds a face-matched depth map from a photo. If you want an actual generated mesh, ObjPop turns an image into a 3D object using a machine learning model.
What "3D Portrait" Means in After Effects
The phrase covers three different results, and the right plugin depends on which one you need.
- 2.5D parallax: a flat photo is displaced by a depth map, so a virtual camera can move slightly around the subject. This is the classic "3D photo" look and works well for documentaries, social edits, and title sequences.
- Depth-displaced mesh: a photo is projected onto a mesh built from a depth estimate, which allows more camera travel, relighting, and an OBJ export.
- Generated 3D object: a machine learning model infers a whole 3D shape from an image. This gives you the most geometry but also the most variability.
None of these gives you a true scan of a person. They all infer geometry from a single view, so the hidden side of the head, the area behind hair, and fine facial detail are guesses. Plan your camera moves accordingly.
Face 3D (aescripts)
Face 3D is built specifically for portrait and photo animation. It runs as a docked extension panel (Window > Extensions) and uses a free Blender install for background processing. It is listed at $49.99, with a sale price of $37.00 at the time of the research.
What it does well
- Built-in depth map management: the Depth Map Manager can generate depth maps through several open-source models, including Marigold, Distill-Any-Depth, and Depth-Anything-v2-small. It can also use Depth Scanner 2 and Quick Depth 3 from aescripts, or import depth maps from a local folder.
- Editable depth: the depth map arrives in a comp with effects already applied, so you can adjust contrast with curves or soften it with a blur.
- Relighting and OBJ output: later versions added scene relighting and smoother, lighter OBJ models.
- Project organization: the plugin creates a working folder for models and textures, and you can change its location.
Limits to know
- It requires After Effects 24.2+ and a GPU that supports the Advanced 3D Renderer. The plugin warns you if either condition isn't met.
- Blender must be installed alongside it.
- The result is only as good as the depth estimate, so expect to tune the map for each portrait.
Pick it if you want the most portrait-focused workflow, a real mesh in After Effects, and the ability to swap depth models until one suits your image.
Volumax
Volumax is an After Effects template rather than a conventional plugin. That makes it approachable. You drop your photo into a photo comp, click Setup, and align a premade male or female head mesh to the face in the photo. You can change the 3D camera angle to match the direction the subject is facing, then use the adjustment tools to line up the face and neck.
Once the mesh matches, you turn off X-ray mode to reveal the full depth map. A controller null then drives the 3D camera movement. The standard move is to keyframe the null from one side of the comp to the other for a smooth parallax push. A reviewer demonstrated this at quarter resolution to keep playback quick.
Limits to know: the face-shaped proxy mesh is a shortcut. It is fast and predictable for typical faces, but it is not a measured depth estimate of your specific subject, so unusual poses, accessories, or very stylized portraits need more manual adjustment.
Pick it if you need fast, clean parallax moves, you value a visual setup you can control by hand, and you don't need to export a mesh.
ObjPop
ObjPop takes a different route. You upload an image and a machine learning model returns a 3D object that lands directly in After Effects. It offers two modes:
- API mode: generation runs remotely and consumes credits. In the demo, new sign-ups received a small number of starting credits, and each generation deducted one.
- Local mode: you start a local server, which downloads a Hugging Face model on first run and can take several minutes. Later starts still take about a minute. Controls include mesh resolution, background removal, and a foreground ratio that affects how much depth is baked in. Generation time depends on your hardware.
The presenter of the demo acknowledged some imperfections in the models. For portraits, treat it as a way to get a rough 3D volume quickly, not a replacement for a careful depth-map workflow.
Pick it if you want a full generated 3D object rather than a displaced photo, and you can tolerate variability in the results and either credit use or local setup time.
The No-Plugin Depth Map Workflow
You can build the same effect with native After Effects tools, which is useful when you can't install third-party scripts or want to understand what the plugins automate.
- Generate a depth map for the portrait with any depth estimation tool. Brighter pixels are closer and darker pixels are farther.
- Bring both layers into After Effects: the photo and the depth map.
- Edit the depth map. Use curves to increase the separation between near and far, then add a blur to soften the transitions. Sharp depth edges cause stretched textures at the silhouette.
- Displace the image using the depth map, either with a displacement effect or with the Advanced 3D Renderer and a mesh.
- Add a camera and a null to drive a subtle orbit or push-in.
- Fix edges. Extend or repaint the background behind the subject so movement doesn't reveal empty areas.
This approach is slower, but it gives you full control and works with any depth source.
Choosing Between Them
Tips for Better 3D Portrait Results
Start with a good source image. Sharp focus, clean edges, and even lighting produce better depth estimates. Heavy motion blur or very busy backgrounds confuse most models.
Fix the depth map before blaming the plugin. Most "bad 3D" is a flat or noisy depth map. Increase contrast with curves, then blur slightly to eliminate stair-stepping and stretched textures.
Keep camera moves small. A few degrees of orbit sells the effect. Large moves reveal the limits of single-image inference, especially around ears, hair, and shoulders.
Separate the subject from the background. Treating them as different layers lets you push the background back, fill the gap behind the subject, and avoid smearing.
Scale depth deliberately. If a model looks too shallow or too deep, adjust the Z scale independently rather than scaling uniformly.
Use relighting sparingly. If your tool supports it, add a gentle light to reinforce depth, but remember the texture still contains the original photo's baked-in lighting.
Work at preview quality first. Draft at lower resolution, lock the move, then render at full quality.
Where Styleframe Fits
These plugins all work inside After Effects and use a single photo as their base. is useful when a portrait project needs something they don't cover. It is an AI video and motion tool built around artistic control, not one-click black boxes. Every generation is driven by your own reference images, video clips, and keyframes.
Practical ways it can sit beside a 3D portrait workflow:
- 2D-to-3D and Turntable: generate a 3D interpretation of a character or portrait and a turntable view, giving you additional angles of a subject to reference or composite.
- Remove Background: isolate the subject before you build a depth map or composite it over a new plate.
- Restyle and Ink-and-Paint: apply a reference-driven style to a portrait or a rendered move, so the final look matches a style frame instead of the raw photo.
- Frames-to-Video: use keyframes to control how a shot develops between start and end frames.
- 4K Upscale: bring finished output up to production-ready 4K.
Iteration is non-destructive, so you can try variations without overwriting earlier results. Output exports as image sequences, which slot straight into After Effects or Nuke, and an After Effects plugin is on the way. Your compositing, camera work, and color pipeline stay where they are. Styleframe supplies extra source material and style control around them.
FAQ
Do I need Blender for 3D portrait plugins in After Effects? Face 3D does, since it uses Blender for background processing. Volumax and the manual workflow do not.
Can these tools create a fully rotatable head? Not reliably. Single-image methods infer depth for the visible surface, so wide rotations expose missing information. Keep moves modest or supplement with extra views.
Which is better for client work? That depends on the shot. For subtle parallax on a still, Face 3D or Volumax is typically enough. If a client needs a stylized or re-lit result, plan time for depth map cleanup, background repair, and compositing in any of them.
Is the Advanced 3D Renderer required? For Face 3D, yes, along with After Effects 24.2 or above. Its documentation also includes a tutorial for working when the renderer isn't supported.