> For the complete documentation index, see [llms.txt](https://docs.openbrush.app/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.openbrush.app/user-guide/exporting-videos/gaussian-capture.md).

# Gaussian Capture

Generate images, camera poses and sparse point data for Gaussian splat workflows.

Gaussian Capture generates a set of images and camera data for training a Gaussian splat in another application. Shape widgets define where Open Brush places the virtual cameras and what part of the sketch they face.

{% hint style="warning" %}
Gaussian splat training is not performed by Open Brush. The output is an input dataset for a separate training tool. Large capture sets take time and use substantial storage.
{% endhint %}

## Add capture areas

1. Switch to Advanced mode.
2. Open **More Options** > **Labs**.
3. Open **Gaussian Splat Pose Capture**.
4. Add one or more capture areas.
5. Move, rotate and scale each widget to cover the part of the sketch you want to capture.

The available shapes distribute cameras differently:

| Shape      | Camera placement                                                          |
| ---------- | ------------------------------------------------------------------------- |
| Sphere     | Cameras surround the widget and face its centre.                          |
| Hemisphere | Cameras cover one half of a sphere and face its centre.                   |
| Ellipsoid  | Cameras surround a stretched or flattened volume and face its centre.     |
| Box        | Cameras are distributed through a volume using subdivisions on each axis. |

You can use several capture widgets in one sketch. The export includes the views generated by every active capture widget.

## Adjust the number of cameras

Grab a capture widget with both hands and scale it normally to change the size of the capture area.

Hold **X/A** while scaling to change its capture settings instead:

1. Sphere, hemisphere and ellipsoid widgets adjust the number of rings and views per ring.
2. Box widgets adjust the subdivisions on the active axis.

The widget displays the resulting camera count. More cameras provide more viewpoints but take longer to capture and train.

## Export the dataset

1. Save the sketch first if you want the output folder to use its name.
2. Return to the Gaussian capture tray.
3. Select **Export Gaussian Poses**.
4. Wait for the capture progress to finish before moving the scene or closing Open Brush.

Each run creates a new numbered folder in:

```
Documents/Open Brush/SplatPoses
```

The folder contains PNG images plus COLMAP-compatible `cameras.txt`, `images.txt` and `points3D.txt` files. An unsaved sketch uses `Untitled` as the folder-name prefix.

## Capture tips

1. Keep the subject inside every capture area and avoid placing cameras inside solid geometry.
2. Start with fewer cameras while testing placement.
3. Animated and audio-reactive content is paused during capture so the exported views remain consistent.
4. Use overlapping capture areas when a single shape cannot cover the subject with useful viewpoints.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.openbrush.app/user-guide/exporting-videos/gaussian-capture.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.