Andrii Shramko

Andrii Shramko

Engineer · Inventor · 4DGS Researcher

Preparing to release the ultimate 4DGS module for ComfyUI. Take your seats.

True 360°. Volumetric video generation without flicker or artifacts.

Join the closed WaitlistOnly 197 slots left

Hey friends! I'm Andrii Shramko. Maybe we've already crossed paths on LinkedIn where I share my R&D projects, or you've used my open scans scan 1 · scan 2. Like you and my clients, I'm critically short on decent volumetric generation tools for commercial work. I keep hunting for better solutions, but the market right now is a dead end. So I put together a team and we're building our own ultimate 4DGS module for ComfyUI — true 360° without gaussian flicker. We're heading into the final acceleration phase.

We're all tired of noisy, heavy 4DGS and non-commercial licenses.

You've probably already tried tools like Apple SHARP. Let's be honest — it's a cool toy, but it's not fit for real work. They give you a measly 10–20° parallax, completely ignore the back of the subject, and "boil" with artifacts whenever you try animations. And yes, commercial use of SHARP is strictly forbidden by license. How are we supposed to sell such VFX shots to clients? We can't.

We need a tool that delivers stable, great results and is fully legal for commercial work. That's exactly what we're building into one pipeline.

Apple SHARP

  • Limited 10–20° parallax only
  • Ignores subject's back side
  • Temporal "boiling" in animations
  • Commercial use strictly forbidden
  • ~100 GB per minute of video

Shramko-Human-4D

  • Full 360° volumetric view
  • Accurate back-side reconstruction
  • Absolute temporal stability
  • 100% legal for commercial use
  • Ultra-optimized lightweight files — less than 1GB per minute

Why the Waitlist?

Let's be straight: right now we're actively putting this puzzle together. We're one step away from turning this tech into a real tool for everyone. So why this page and a waitlist right now? Because building such a module and processing petabytes of volumetric data is a huge investment of time and resources. Before we make the final push, I need to see real community response. Your signup is a signal to me and my partners that we're building what the market actually needs. I need to know: will you run nodes locally on your hardware, or do you need a cloud API? Leave your details — when we're ready, you'll get access first.

Reserve early access. Launch is around the corner!

Your waitlist signup is a vote FOR this product and a ticket to the front row. Fill in the short form so I know what you need the tool for. Everyone who supports us at this stage gets exclusive terms right after launch — enough credits for plenty of tests.

Choose your main scenario

Technology: capabilities, benefits & use cases

What will our module give content creators? The new architecture fundamentally solves the main problems of 4D generation: temporal inconsistency, geometry "hallucinations," and huge file sizes. The generative node uses diffusion networks trained not just on flat video but on "video → ready volumetric video" pairs from the Shramko-Human-4D dataset. From day one the model outputs optimized 4DGS, physically correct fabric simulation, micro-expressions, and bullet-time slow-motion on 4DGS video.

Key commercial use cases

Direct streaming to mobile XR apps

You get not just new flat frames but ready, super-optimized volumetric 4DGS video. These assets are light enough to stream natively on Apple Vision Pro, Meta Quest 3, or AR apps on smartphones — no lag, no overheating.

Interactive 4DGS films

Cinema moves into spatial dimension. The viewer can not only watch the video but physically move (walk, lean) inside the generated scene while it plays.

VFX & Virtual Production

Fast capture and digitization of real actors, then seamless transfer into Unreal Engine / Unity as digital twins for photorealistic LED volumes — without expensive mocap suits.

Shramko-Volumetric-Bridge for ComfyUI

Our tool ships as a custom node package called "Shramko-Volumetric-Bridge." The goal: make 4DGS creation as intuitive as regular video generation — just connect nodes in a graph.

ComfyUI 4DGS workflow — Shramko-Volumetric-Bridge

Pipeline & nodes

Ingest Node

Ingest Node

Accepts your media, auto-analyzes video, and extracts camera metadata (focal length from EXIF) — critical for correct 3D coordinates and zero distortion.

Processing Node

Processing Node

Main magic happens here. The network analyzes frames, predicts hidden geometry, and computes optical flow for smooth animation. Runs locally on your GPU or as a thin API client to cloud.

Visualization Node

Visualization Node

No need to export to external tools. A WebGL/WebGPU player appears right in ComfyUI — rotate and zoom the 4D scene with your mouse while it plays.

Export Node

Export Node

Optimizes gaussians and packs data for Unreal/Unity, VFX, or ultra-light files for streaming to Meta Quest and Apple Vision Pro.

Your main advantages

You get a studio-grade tool running on your own hardware inside ComfyUI. That means 100% privacy of your source assets and independence from cloud render farms. Thanks to the legally clean training data — every 4D scene you create is fully legal for commercial client work.