<!-- mobian-agent-page publisher="dailydev" canonical="https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw" -->

---
title: The End of Blurry Point Clouds: 3D Gaussian Splatting...
description: A principal engineer from Cesium (Bentley Systems) explains 3D Gaussian Splatting, a radiance-field technique that turns photos into photorealistic 3D scenes...
canonical: https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw
twitter:card: summary_large_image
twitter:site: @dailydotdev
og:type: website
og:site_name: daily.dev
og:title: The End of Blurry Point Clouds: 3D Gaussian Splatting Explained | daily.dev
og:description: A principal engineer from Cesium (Bentley Systems) explains 3D Gaussian Splatting, a radiance-field technique that turns photos into photorealistic 3D scenes...
og:url: https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw
og:image: https://api.daily.dev/og/posts/TlzYYwXcw.png
og:image:alt: The End of Blurry Point Clouds: 3D Gaussian Splatting Explained
og:image:width: 1200
og:image:height: 630
og:locale: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# The End of Blurry Point Clouds: 3D Gaussian Splatting Explained

**[Khronos](https://daily.dev/sources/khronos)** · 35 min read · 1 upvotes · 0 comments

## Summary

A principal engineer from Cesium (Bentley Systems) explains 3D Gaussian Splatting, a radiance-field technique that turns photos into photorealistic 3D scenes using translucent ellipsoid 'blobs' rendered via the painter's algorithm, contrasting it with photogrammetry meshes, point clouds, and NeRFs. The discussion covers how splats capture thin structures, reflections, and lighting better than meshes, the precision trade-offs versus point clouds, and ongoing work at Khronos to standardize a 3D Gaussian Splatting extension for glTF, built with input from Cesium, Esri, Huawei, Niantic Spatial, and Nvidia. Compression formats like SPZ, SOG, and LGSC are discussed as critical to managing massive splat file sizes, alongside a look ahead at 4D and HDR splatting.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://youtu.be/a7bAUMbMEec>

## Questions this post answers

### What is 3D Gaussian splatting and how does it differ from a mesh or point cloud?

3D Gaussian splatting is a radiance-field reconstruction method that turns a set of photos into a photorealistic 3D scene using small translucent ellipsoids called Gaussians, each with position, color, size, and opacity attributes, rendered onto a 2D plane using the painter's algorithm. Unlike photogrammetry meshes or point clouds, it captures thin structures, reflections, and view-dependent lighting more convincingly and photorealistically.

_Developers building 3D visualization pipelines can follow how graphics techniques like this evolve on daily.dev._

### How accurate are measurements taken from a 3D Gaussian splat compared to a point cloud?

Roughly 90% of measurements taken from a 3D Gaussian splat are effectively as accurate as a point cloud, but the remaining 10% can suffer from small precision shifts because points are inferred by blending overlapping ellipsoid blobs rather than measured directly. For applications needing that last bit of precision, pairing the splat with an underlying mesh or point cloud is recommended.

_Teams weighing 3D capture techniques for precision-critical work can track these trade-offs on daily.dev._

### What compression formats exist for reducing the storage size of 3D Gaussian splats?

Several compression formats have emerged to shrink otherwise massive splat datasets, which can reach hundreds of gigabytes uncompressed: SPZ by Niantic Spatial, popular for geospatial use cases; SOG (spatially ordered Gaussians) by PlayCanvas, popular in gaming and visual media; and LGSC by Qualcomm, with additional formats from companies like X Grids also in use.

_Engineers dealing with large 3D asset pipelines can keep tabs on emerging compression standards via daily.dev._

## Similar posts on daily.dev

- [Why an Open Standard For Gaussian Splats Changes 3D Modeling](https://daily.dev/posts/why-an-open-standard-for-gaussian-splats-changes-3d-modeling-we7bf6j5c) · Khronos · 0 upvotes · 0 comments

---

Tags: [#3d](https://daily.dev/tags/3d)

[View this post on daily.dev](https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw)

```json
{"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]}
{"@context":"https://schema.org","@type":"TechArticle","headline":"The End of Blurry Point Clouds: 3D Gaussian Splatting Explained","url":"https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw"},"datePublished":"2026-08-17T13:34:17.344Z","dateModified":"2026-08-17T13:34:42.709Z","description":"A principal engineer from Cesium (Bentley Systems) explains 3D Gaussian Splatting, a radiance-field technique that turns photos into photorealistic 3D scenes...","image":"https://i.ytimg.com/vi/a7bAUMbMEec/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/a7bAUMbMEec/sddefault.jpg","isAccessibleForFree":true,"articleSection":"Khronos","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Khronos","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/792618c65bcf419783f2a62c03d29e1c","url":"https://daily.dev/sources/khronos"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"3d","timeRequired":"PT35M","video":{"@type":"VideoObject","name":"The End of Blurry Point Clouds: 3D Gaussian Splatting Explained","description":"A principal engineer from Cesium (Bentley Systems) explains 3D Gaussian Splatting, a radiance-field technique that turns photos into photorealistic 3D scenes...","thumbnailUrl":"https://i.ytimg.com/vi/a7bAUMbMEec/sddefault.jpg","uploadDate":"2026-08-17T13:34:17.344Z","duration":"PT35M","url":"https://api.daily.dev/r/TlzYYwXcw","embedUrl":"https://www.youtube.com/embed/a7bAUMbMEec"}}
{"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Khronos","item":"https://daily.dev/sources/khronos"},{"@type":"ListItem","position":3,"name":"The End of Blurry Point Clouds: 3D Gaussian Splatting Explained"}]}
{"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-end-of-blurry-point-clouds-3d-gaussian-splatting-explained-tlzyywxcw#faq","mainEntity":[{"@type":"Question","name":"What is 3D Gaussian splatting and how does it differ from a mesh or point cloud?","acceptedAnswer":{"@type":"Answer","text":"3D Gaussian splatting is a radiance-field reconstruction method that turns a set of photos into a photorealistic 3D scene using small translucent ellipsoids called Gaussians, each with position, color, size, and opacity attributes, rendered onto a 2D plane using the painter's algorithm. Unlike photogrammetry meshes or point clouds, it captures thin structures, reflections, and view-dependent lighting more convincingly and photorealistically. Developers building 3D visualization pipelines can follow how graphics techniques like this evolve on daily.dev."}},{"@type":"Question","name":"How accurate are measurements taken from a 3D Gaussian splat compared to a point cloud?","acceptedAnswer":{"@type":"Answer","text":"Roughly 90% of measurements taken from a 3D Gaussian splat are effectively as accurate as a point cloud, but the remaining 10% can suffer from small precision shifts because points are inferred by blending overlapping ellipsoid blobs rather than measured directly. For applications needing that last bit of precision, pairing the splat with an underlying mesh or point cloud is recommended. Teams weighing 3D capture techniques for precision-critical work can track these trade-offs on daily.dev."}},{"@type":"Question","name":"What compression formats exist for reducing the storage size of 3D Gaussian splats?","acceptedAnswer":{"@type":"Answer","text":"Several compression formats have emerged to shrink otherwise massive splat datasets, which can reach hundreds of gigabytes uncompressed: SPZ by Niantic Spatial, popular for geospatial use cases; SOG (spatially ordered Gaussians) by PlayCanvas, popular in gaming and visual media; and LGSC by Qualcomm, with additional formats from companies like X Grids also in use. Engineers dealing with large 3D asset pipelines can keep tabs on emerging compression standards via daily.dev."}}]}
```

