Desktop Trainer

Postshot

Postshot is Jawset’s desktop application for creating radiance fields and Gaussian Splatting scenes from photos or videos with an artist-friendly workflow.

Stage: TrainingArtistsDesktop productionLive training preview

Open official linkPostshot Product Page->

What It Does

Postshot sits between research code and cloud apps. It is built for people who want to train locally without assembling a Python pipeline, while still getting access to production-style controls and live feedback.

It is especially helpful when a visual artist or capture technician needs to iterate on datasets and quality settings without becoming the pipeline engineer for the whole project.

How To Use It In 3DGS

Bring in photos or video from a controlled capture.
Let Postshot process and train while watching the live preview for coverage or quality problems.
Adjust scene settings such as splat scale, opacity, and spherical harmonics where needed.
Export the model and preview it in a browser viewer before sharing.

Things To Watch

It is not the best choice for Linux-first research automation.
GPU requirements matter; weak or unsupported hardware will limit the experience.
A friendly UI does not rescue a poor capture, so still invest in stable footage and overlap.

Official And Related Links

Postshot Product Page Postshot User Guide

Build Channels And Hardware

Jawset separates stable Release Builds from frequently updated Pre-Release Builds; use release builds for client work and pre-release only when you need the newest function.
Official system guidance targets Windows 10+ and NVIDIA GPUs with Compute Capability 7.5 or higher, such as RTX 2060-class hardware or better.
Bug reports should include `Postshot.log`; training bugs are easier to diagnose when the input images or video can be shared.

Postshot User Guide Postshot Builds