Introducing Synth4D: omnipercipient spatial intelligence Read the note →

Spatial intelligence

Everywhere
at once.

One take. Every angle. Synth4D is an omnipercipient model for spatial intelligence.

Original · source

Evolution of Reality Capture

1826

Film Photography

1890

Motion Pictures

1957

Digital Photography

1986

Digital Video

2020

Spatial Photography

Now

Spatial Video

Moving through
space and time.

An omnipercipient model for spatial intelligence, adding another dimension to video.

Research

Synth4D: An Omnipercipient Model for Spatial Intelligence

Synth4D estimates a 4D Gaussian splat from synchronized multi-view capture, adding another dimension to video. We describe the model, the rigs it is trained against, and the claim: spatial intelligence over a take the cameras actually saw.

Synth4D research note →
ViewSynth multi-camera capture rig

How

Rigs. Model. Media.

  1. 01 Capture on our rigs
  2. 02 Spatial video with Synth4D
  3. 03 Deliver spatial video

Demo

Original vs Synth4D

Original Source capture
Reshoot Synth4D · 4DGS

Applications

Marketing, Entertainment, Education.

Adding another dimension to video. New cameras after wrap.

Marketing

On-brand spatial assets. Fashion, ads, activations, reframed after the shoot.

Entertainment

Music videos, film, live events. Bullet time and free-viewpoint from the take you already have.

Education

Lectures, training, museums. A demonstration you can walk around after it ends.

Gallery

Selected work

Build with us.

Capture, production, and Synth4D.

Contact ViewSynth →