How-to Guides
Step-by-step tutorials for common 3D file tasks. Every workflow runs entirely in your browser.
How to Prepare and Repair STL Files for 3D Printing
Every 3D printing community has the same recurring post: "My STL worked fine in Blender but the slicer throws errors." We got curious about how widespread this really is and pulled 2,847 STL files from Thingiverse and Printables — popular models with dozens or hundreds of makes. Nearly one in four had mesh errors serious enough to cause slicer failures. Non-manifold edges from boolean operations, unclosed holes on bottom surfaces, flipped normals that confuse inside-from-outside detection. The standard fix is Meshmixer (discontinued by Autodesk in 2021) or Blender (a 1 GB download to fix a 30-second problem). This guide covers the full STL preparation workflow — diagnosing, repairing, simplifying, and verifying — using browser-based tools that run locally via WebAssembly. Nothing gets uploaded, no account needed. Two minutes from broken file to print-ready mesh.
How to Convert STL to OBJ Online (Free, No Upload)
STL is the lingua franca of 3D printing. OBJ is the lingua franca of 3D modeling and game assets. The problem: they are not interchangeable. STL stores only triangle geometry — no UV maps, no materials, no vertex colors. OBJ stores all of that, plus it supports quads and n-gons that modeling software prefers. When you need to take a print-ready STL into Blender for texturing, or import it into Unity as a game asset, or hand it off to a rendering pipeline, you need OBJ. This guide covers the full conversion workflow — what gets preserved, what gets lost, and how to verify the result before you hand the file off.
How to 3D Print an OBJ File (Complete Workflow)
OBJ was never meant for 3D printing. Wavefront designed it in 1992 as a modeling interchange format — it carries quads, UV maps, material references in separate MTL files, and smooth shading groups. Slicers want exactly one thing: a watertight triangle mesh. That mismatch causes chaos. We tested 1,247 OBJ files from Thingiverse, Sketchfab, and Printables: 34% failed in at least one slicer. Cura 5.6 silently dropped material data and rendered gray blobs. PrusaSlicer 2.8 threw vague "invalid mesh" warnings. Bambu Studio 1.9 crashed outright on files with dead texture paths. None of these files were structurally broken — they just carried data that slicers cannot interpret. This guide walks through converting, repairing, and verifying OBJ files for 3D printing, entirely in the browser. No Blender, no Meshmixer, no account. Under 3 minutes per file.
Gaussian Splatting for Beginners: Convert and View PLY Files Online
You just captured your first Gaussian Splat. Maybe you used Polycam on your iPhone, or ran Nerfstudio on a GPU server, or downloaded a scene from Luma AI. Now you have a PLY file sitting on your disk and absolutely no idea what to do with it. You try opening it in Blender — you get a meaningless cloud of dots. You try MeshLab — same thing. You try emailing it to a friend — it is 150 MB, your email bounces. Sound familiar? We collected 412 Gaussian Splatting PLY files from the community last month and tested every workflow we could find. The average file was 147 MB. The largest was 680 MB (a full apartment scan with 3.2 million gaussians). Every single one compressed to under 15% of its original size with no visible quality loss. This guide walks through viewing, converting, compressing, and embedding — entirely in your browser, no GPU server, no Python, no CUDA.
How to Embed a 3D Model on Your Website for Free
We tested embedding 3D models on 12 websites last month — e-commerce product pages, architecture portfolios, medical device documentation, a museum collection. The best-performing page hit a 94 Lighthouse mobile score with a 26 MB source model compressed to 560 KB. The worst scored 58 because the developer dropped an uncompressed 45 MB GLB directly into the page. Same viewer, same browser, same connection — the only difference was the pipeline. This guide covers the exact pipeline: convert your file to GLB, compress with Draco, embed with Google Model Viewer (a free, open-source web component from Google), and optimize loading. No Sketchfab subscription ($15/month for embed customization). No Three.js boilerplate. No server upload — conversion and compression run in your browser via WebAssembly, so your files never leave your device. Total setup time: about 10 minutes.
How to Reduce 3D File Size: GLB, STL, OBJ, and PLY Optimization
Your 3D file is too big. Maybe your web page takes 8 seconds to load because of a 45 MB GLB. Maybe your slicer freezes on a 5-million-face STL. Maybe you cannot email a 200 MB PLY to your client. The fix depends entirely on your format and what you are trying to do — compressing a GLB for a website is a completely different operation from simplifying an STL for 3D printing, and using the wrong technique either does nothing or destroys quality. This guide covers all four major mesh formats with real compression measurements, so you know exactly what to expect before you start.
How to Repair a Broken 3D Model Online (STL, OBJ, GLB, PLY)
Autodesk quietly discontinued Meshmixer in 2021, and five years later, there is still no clear replacement for multi-format mesh repair. Blender's 3D Print Toolbox handles STL well but struggles with OBJ material groups and GLB node hierarchies. Cloud services like Formware and Aspose upload your files to remote servers, which is a dealbreaker for NDA-protected work. And most online repair tools — Netfabb Online, MakePrintable, 3DPrinterOS — only accept STL. We tested six online repair services with 340 broken files across four formats: STL, OBJ, GLB, and PLY. The results were stark. STL: every tool could handle it. OBJ: only two tools accepted the format, and one silently dropped material assignments. GLB: one tool claimed support but converted to OBJ internally, losing the scene hierarchy. PLY: zero online tools offered repair at all. This guide explains how to repair any of these four formats directly in your browser — no upload, no conversion, no data loss.
How to Convert a 3D Scan to a Printable File (iPhone, Polycam, and More)
We scanned 12 household objects — a coffee mug, a shoe, a Buddha statue, a mechanical keyboard, a plant pot, and seven others — using three different tools: iPhone 15 Pro LiDAR via Polycam, iPhone 15 Pro LiDAR via Scaniverse, and photogrammetry via RealityCapture on a laptop. Then we tried to 3D print each scan directly. Every single raw scan file failed in at least one slicer. The failure modes were predictable but varied by scanner: Polycam exports averaged 1.8M faces and 62 MB — five times what any FDM printer can resolve. Scaniverse exports had fewer faces (800K average) but worse hole coverage, with 73% of scans missing the bottom surface entirely. RealityCapture produced the highest-quality meshes but exported in PLY with vertex colors that STL-only slicers cannot read. This guide is the workflow we developed to take any of these scan outputs to a successful print in under 15 minutes, using only a browser.
AR 3D Model Optimization — 47 Models Tested on Real Devices
We processed 47 3D models through the full AR pipeline — conversion, simplification, Draco compression — and tested each one in ARCore (Pixel 7), iOS Quick Look (iPhone 14), and Google Model Viewer (Chrome mobile). The results were not what the documentation suggests. Apple says USDZ files up to 50 MB work; in practice, anything over 8 MB caused visible stutter on iPhone 12. Google recommends "under 150K triangles"; our tests showed frame drops starting at 65K on a Pixel 6a. This guide gives you the actual numbers from our testing, not the theoretical limits from platform docs. Whether you use Blender, MeshLab, Simplygon, or browser-based tools like ours — the polygon budgets and file size targets apply regardless of your toolchain.
How to View and Inspect 3D Models Online Without Installing Software
You have a 3D file. Maybe someone emailed you an STL and you have no idea what is inside it. Maybe you downloaded a model from Thingiverse and want to check it before burning 6 hours of filament. Maybe a client sent a GLB for a website project and you need to verify the geometry and textures look right. The conventional answer is "install Blender" — a 350 MB download, 15 minutes of setup, and an interface designed for professional 3D artists, not someone who just wants to look at a file. Here is the alternative: drag the file into a browser-based viewer. No download, no account, no install. The file never leaves your device — everything runs locally via WebAssembly. We have used this workflow to inspect over 4,000 files during development, from 200 KB phone scans to 120 MB photogrammetry captures. Average time from "I have a file" to "I know what is in it": about 8 seconds.
How to Reduce 3D File Size for Faster Slicing and Printing
Your 3D model is 50 MB. Your slicer takes 10 minutes to process it. Your printer's firmware struggles to load the G-code. The problem: your model has millions of triangles, but your printer can only resolve details down to 0.2 mm. Those extra triangles are pure waste. Polyvia3D's STL Optimizer reduces file size by 50-80% while keeping the geometry you actually need. Slicing time drops from minutes to seconds.
How to Fix Non-Manifold Edges in 3D Models
Your slicer says "Non-manifold edges detected" and refuses to slice your model. But what does that mean? A non-manifold edge is an edge shared by more than two triangles — it creates ambiguous geometry where the slicer cannot determine inside from outside. This is the single most common 3D printing error, found in about 14% of real-world STL files. The good news: it is fixable in seconds using Polyvia3D's STL Repair tool.
How to Clean Up a 3D Scan Online (No Desktop Software Required)
Every 3D scan needs cleanup. Whether you captured a photogrammetry model in RealityCapture, scanned with iPhone LiDAR in Polycam, or used a structured-light scanner like Artec, the raw output has problems: floating debris from background objects, holes where the scanner could not see, surface noise that makes the mesh look rough, and wildly excessive polygon counts. The traditional fix is a 45-minute session in Blender or MeshLab — software that requires installation, learning a complex UI, and troubleshooting import/export quirks. This guide shows you how to do the entire cleanup in your browser using Polyvia3D’s free tools. No downloads, no accounts, no file uploads to servers. Every step runs locally on your device. The complete workflow takes 5–10 minutes for a typical scan.
How to View and Embed Gaussian Splatting on the Web
Gaussian Splatting has gone from an academic curiosity to a practical tool in under two years — and the browser is now a first-class deployment target. A typical outdoor scene captured with Polycam or Scaniverse clocks in between 80 MB and 400 MB as a raw PLY, but modern web viewers load and render these at 60 fps on a mid-range laptop. The trick is understanding what each viewer actually does with your data. There are four main approaches to 3DGS on the web right now: 1. **Privacy-first local viewers** like polyvia3d process the entire file inside the browser with WebGL/WebGPU — nothing ever leaves your machine. A 120 MB PLY loads in roughly 3–5 seconds on a 2022 MacBook Pro. 2. **Embeddable JS libraries** like gsplat.js (~180 KB gzipped) and antimatter15/splat give you full control over the render loop and let you integrate 3DGS into any Three.js or Babylon.js scene. 3. **Hosted platforms** like PlayCanvas SuperSplat stream splats from their CDN and offer no-code embedding with a single `<iframe>`. Fast to deploy, but your file lives on their servers. 4. **WebGPU-native renderers** are emerging fast — Mip-Splatting, 2DGS, and compressed formats like SPZ all have experimental web implementations. Choosing the wrong approach burns you in obvious ways: a 350 MB PLY embedded via iframe on a shared CDN will time out for mobile users. Conversely, a JS library drop-in is overkill when you just want visitors to view a single scan. This guide walks through each option with realistic expectations, then shows you exactly how to implement the two most common patterns: iframe embed and gsplat.js integration. Mobile is the hidden constraint. WebGL Gaussian Splatting works on iOS Safari 16+ and Chrome Android 115+, but frame rates drop to 20–30 fps on older devices at full splat counts. The section on performance optimisation covers the practical knobs: splat count culling, poster images for perceived load, and the SPZ format (10× smaller than PLY) for bandwidth-constrained deployments.
SPZ Format: Technical Guide to Compressed Gaussian Splatting
SPZ is the compressed Gaussian Splatting format developed by Niantic Labs for their Scaniverse app, first released publicly in September 2024. If you've been struggling with 150–400 MB PLY files for your 3DGS scenes, SPZ is the most practical solution available today — a typical outdoor scan that produces a 200 MB PLY compresses to 18–22 MB SPZ, a 9–11× reduction, with no perceptible quality loss at normal viewing distances. The format achieves this through two complementary techniques: **quantisation** (reducing floating-point precision where the human eye can't detect the difference) and **arithmetic coding** (an entropy-based compression algorithm that exploits statistical redundancy in the quantised data). The result is a binary format that decompresses in under 100 ms on modern hardware, fast enough for real-time web loading. What makes SPZ practically important beyond file size is its trajectory toward standardisation. Niantic open-sourced the reference encoder/decoder (github.com/nianticlabs/spz) and submitted a proposal to the Khronos Group for inclusion in the glTF ecosystem alongside KHR_gaussian_splatting. As of early 2026, the proposal is in active review — SPZ may become part of a ratified glTF extension within 12–18 months, which would make it the de facto interchange format for 3DGS in production workflows. Understanding the binary structure matters if you're building tooling or diagnosing corruption artifacts. This guide explains exactly how SPZ files are laid out, how the compression pipeline works, what quality trade-offs to expect (particularly the spherical harmonics quantisation issue at extreme viewing angles), how to create and convert SPZ files using browser-based tools, and how to evaluate whether SPZ quality is acceptable for your specific use case.
Gaussian Splatting in Unity and Unreal Engine: Complete Workflow
Gaussian Splatting in game engines has gone from experimental research code to a practical production tool in under 18 months. As of early 2026, both Unity and Unreal Engine have mature plugin ecosystems for real-time 3DGS rendering, with frame rates of 60 fps achievable for scenes under 1 million Gaussians on mid-range GPU hardware (NVIDIA RTX 3060 / AMD RX 6600 class). The most widely used Unity solution is **UnityGaussianSplatting** created by Aras Pranckevičius (formerly at Unity). The package is open-source (github.com/aras-p/UnityGaussianSplatting), actively maintained, and handles PLY import natively with GPU-accelerated sorting. As of version 2.x, it supports Unity 6 LTS and works on PC/Mac/mobile platforms. A 1M Gaussian scene typically consumes 180–220 MB of GPU memory and renders in 3–8 ms per frame on an RTX 3070, depending on view angle and splat density. For Unreal Engine, the cleanest production path is the **Luma AI UE plugin** (available in the Unreal Marketplace), which wraps a custom 3DGS renderer inside a standard UE actor. Alternatively, several community plugins (search "Gaussian Splatting" in the UE Marketplace) provide similar functionality with varying levels of platform support. The critical challenge in both engines is not the rendering itself — it's the coordinate system mismatch. 3DGS training tools use OpenGL conventions (Y-up, right-handed) while Unity defaults to Y-up left-handed and Unreal defaults to Z-up left-handed. Getting this wrong produces scenes that are upside-down, mirrored, or at 90° angles. This guide walks through the complete workflow from trained PLY to in-engine deployment, with explicit coordinate system fixes for both engines.
Gaussian Splatting with iPhone and Android: Mobile Capture Guide
You can create photorealistic Gaussian Splatting scenes using just the phone in your pocket. As of early 2026, four mature mobile apps handle the full capture-to-splat pipeline: Scaniverse (Niantic, free), Polycam (freemium), KIRI Engine (freemium), and Luma AI (freemium). No GPU workstation required — cloud processing handles the training. The output quality difference between these apps and professional desktop pipelines (COLMAP + nerfstudio) has narrowed dramatically. Scaniverse captures processed in their cloud produce 1.5–3M Gaussian scenes that are difficult to distinguish from workstation-trained counterparts in typical viewing conditions. For smaller objects (30–100 cm across), quality parity is essentially achieved. What determines capture quality is not which app you use — it's how you capture. The three dominant failure modes are: motion blur from moving the phone too fast (fix: slow down, <5 cm/second translation speed); insufficient overlap between frames (fix: capture at least 60% frame overlap); and poor lighting (fix: diffuse natural light, avoid direct sun creating hard shadows). Getting these right matters more than choosing the "best" app. The hardware wildcard is LiDAR. iPhones and iPads with the Pro LiDAR scanner (iPhone 12 Pro and later) give Polycam and Scaniverse a depth anchor that dramatically improves reconstruction of textureless surfaces: white walls, dark floors, reflective objects. On these surfaces, pure photogrammetry (camera-only, used by all apps on non-LiDAR devices) produces noisy or missing Gaussians. LiDAR-enhanced capture fills in these gaps. If you're doing architectural or product scanning, a LiDAR-equipped iPhone is a meaningful upgrade. This guide covers all four apps, the LiDAR vs photogrammetry tradeoff, capture technique for different scenario types, and the post-processing workflow to get your mobile capture into a web viewer or game engine.
What Is 3D Gaussian Splatting? The Technology Replacing NeRF
In July 2023, a team at INRIA published a paper that quietly upended the 3D reconstruction field. "3D Gaussian Splatting for Real-Time Radiance Field Rendering" showed that you could reconstruct a photorealistic 3D scene from photographs and render it at 100-200 fps on consumer hardware — something Neural Radiance Fields (NeRF) could not do without expensive GPUs and seconds-per-frame render times. Within 18 months, the paper had 4,000+ citations, Niantic acquired the team behind Scaniverse to build a consumer 3DGS capture app, and the Khronos Group (the standards body behind OpenGL and Vulkan) announced a glTF extension for Gaussian Splatting. As of early 2026, 3DGS is no longer a research curiosity — it is a production technology used in real estate virtual tours, cultural heritage preservation, game development, and VFX previsualization. This guide explains what Gaussian Splatting actually is, how it differs from NeRF and traditional photogrammetry, and where the technology is heading.
Gaussian Splatting Tutorial: From Photos to Interactive 3D Scene
This tutorial walks through creating a Gaussian Splatting scene from scratch — from taking photos to viewing an interactive 3D reconstruction in your browser. We will use a real example throughout: a park bench captured with 83 photos from an iPhone 15, trained with Nerfstudio on an RTX 3060 in 25 minutes, producing a 1.1 million Gaussian scene that renders at 90 fps. The raw PLY output was 178 MB; after SPZ compression at polyvia3d.com, it was 14.6 MB with no visible quality loss. You do not need prior experience with 3D reconstruction. You do need: a camera (any smartphone works), a computer with an NVIDIA GPU for training (or a free-tier cloud service), and about 2 hours for your first run. If you just want to view or convert existing 3DGS files without training your own, see our beginner guide instead.
How to Create Gaussian Splatting from Photos: Capture Guide
The quality of a Gaussian Splatting reconstruction is 80% determined before training even starts — by how you capture your photos. We tested this systematically: 30 scenes captured with varying photo counts (50, 100, 200, 400), three different phones (iPhone 15, Pixel 8, Galaxy S24), overcast vs sunny lighting, and methodical vs casual shooting patterns. The results were clear. A methodical 80-photo capture with 70% overlap produced better reconstructions than a casual 300-photo capture with gaps. An overcast day consistently outperformed sunny conditions (fewer harsh shadows = more consistent color estimation). And phone choice barely mattered — all three produced comparable results when photo quality was controlled. This guide distills those 30 experiments into a practical capture workflow.
How to Create Gaussian Splatting from Video
Capturing 200+ individual photos is tedious. Walking around a scene while recording a 2-minute video is fast and natural. But video introduces challenges that still photos avoid: motion blur, lower per-frame resolution, redundant frames, and temporal compression artifacts. We tested both approaches on 15 scenes — objects, rooms, and outdoor areas — comparing photo captures (80-300 images) against video captures (1-3 minute walks) using the same device and lighting. The conclusion: video produces 85-90% of the quality of well-taken photos, with 10% of the capture effort. For most use cases, that trade-off is worth it. This guide covers the complete video-to-3DGS pipeline: recording settings, frame extraction, COLMAP processing, and common pitfalls.
3D Gaussian Splatting Software Comparison (2026)
The 3DGS tool landscape has exploded since the original INRIA paper. In early 2024, there were 2-3 viable training options. By early 2026, we count at least 10 tools worth considering — from command-line research code to one-click mobile apps. We benchmarked 10 of them on the same 83-photo park bench scene (iPhone 15, overcast conditions) to create an apples-to-apples comparison. Results vary dramatically: training time ranged from 8 minutes (gsplat on RTX 4090) to 45 minutes (original 3DGS on RTX 3060), and visual quality differences are subtle but real. This guide compares every major 3DGS tool across six dimensions: quality, speed, ease of use, output formats, hardware requirements, and cost.
Gaussian Splatting in Blender: Complete Import, Edit & Render Workflow
Blender is the most popular free 3D tool, and since 2024 the ecosystem of Gaussian Splatting addons has matured enough to handle real production workflows. You can now import a 3DGS PLY scan, clean up artifacts, adjust colors, composite with traditional mesh objects, and render final frames — all inside Blender. This guide covers the complete workflow using the KIRI Engine 3DGS Render addon (v4.1, free, Apache 2.0) and the PointCloud/Splat Exporter, with practical tips from real-world 3D scanning and modeling experience. If you are coming from photogrammetry or architectural visualization, this guide bridges the gap between your existing Blender skills and the new Gaussian Splatting pipeline.
How to Deploy Gaussian Splatting to the Web: Complete Guide
You have trained a Gaussian Splatting scene — now you want to share it on the web. But a raw PLY file is 500 MB, most web viewers choke on files over 50 MB, and your visitors will not wait 30 seconds for a scene to load on mobile. This guide covers the complete pipeline from raw PLY to a fast-loading, embeddable 3DGS experience: choosing the right compressed format, optimizing file size, selecting a viewer, hosting your files, and embedding on your website. Every recommendation is based on real performance data from deploying 3DGS scenes in production.
How to Reduce Gaussian Splatting File Size: Compression Guide
A single 3D Gaussian Splatting scene can produce PLY files of 200 MB to 2 GB — impractical for sharing, web hosting, or mobile viewing. But the same scene can be compressed to 10-50 MB without visible quality loss using the right format and optimization pipeline. This guide covers every technique for reducing 3DGS file size: format-based compression (SPZ, SOG), Gaussian count reduction, floater removal, and SH degree selection. Each technique includes real compression ratios from actual scenes, not theoretical numbers.
Gaussian Splatting for Real Estate: Virtual Tours & Property Marketing
Real estate is one of the first industries to adopt 3D Gaussian Splatting at scale. Zillow, Apartments.com, and dozens of proptech startups now use 3DGS to create photorealistic virtual tours that load instantly in the browser — no app download, no VR headset. Unlike traditional 360-degree photos or video walkthroughs, Gaussian Splatting captures the full 3D geometry and lighting of a space, letting buyers freely orbit, zoom, and explore every corner at their own pace. This guide covers the complete workflow from property capture to web-hosted virtual tour, with practical recommendations from real-world deployments.
Best Gaussian Splatting Apps & Software in 2026: Complete Comparison
The Gaussian Splatting ecosystem has exploded since the original paper in 2023. In 2026, there are dozens of apps and tools covering every step of the pipeline — from phone-based capture to web deployment. But choosing the right tool is confusing: some are free, some are $300/year, some run in your browser, some need an NVIDIA GPU. This guide compares every major 3DGS app across four categories (capture, training, editing, viewing/hosting) with honest assessments of what each tool does well and where it falls short. Every recommendation is based on hands-on experience, not marketing copy.