Hello Scanners,
The time has arrived! A huge THANK YOU to everyone who participated in our first-ever Scaniverse AMA! We were absolutely thrilled to see your questions flow in, and even more excited to share our answers with you. We’ve compiled the responses to the 13 questions submitted by the community. We also want to let you know that we took everything into consideration – including any follow-up questions or extra context you provided in the comments. So dive in and enjoy! We’re so grateful for your engagement and support!
1. Can you explain the long term plan for the two formats: Mesh vs Splat?
Thanks for the question! Meshes and splats are fundamentally different representations of 3D data, each with its strengths. Meshes, made up of interconnected polygons, excel at capturing precise geometric detail and are ideal for applications like measurement, 3D printing, and CAD. Splats, on the other hand, use a collection of tiny, textured points (Gaussian splats) to represent a 3D scene. This makes them exceptionally well-suited for photorealistic visualizations, handling complex lighting and reflections with ease. Think of it this way: meshes are like the blueprints of an object, while splats are like a photograph capturing its appearance.
As user @sjcbz pointed out, splats are a newer technology in the 3D scanning space, and our different feature and export options reflect these different strengths, our own development timelines, and the wider ecosystem support. In general, we recommend using:
- Meshes: for geometric fidelity and objects where accurate dimensions are crucial, or if you need to export it in popular file formats supported by other engines
- Splats: when visual realism is paramount, especially for larger scenes or those with intricate surface details
At Scaniverse, our vision is to “scan the universe,” and we’re particularly excited about the potential of splats to capture the world in stunning photorealism. That’s why we’re investing heavily in their development. This includes initiatives like creating a map of splats which users can easily share to and browse from, open-sourcing our proprietary SPZ file format, and enabling splat import in Niantic Studio, empowering developers to integrate real-world environments into their applications. We believe splats represent the future of 3D capture, and we’re committed to this journey.
2. What does “Enhance” do?
Enhance is a step function we added for creators who are willing to go an extra mile to maximize the quality of their splat scans. Essentially, it fine-tunes and retrains your splat representation with more Gaussians, iteratively re-optimizing their relative positions, sizes, and colors, leading to a more detailed and accurate 3D model. Each Enhance pass (up to a maximum of 5) takes about a minute on latest iOS devices, and you can progressively refine your scan for the desired level of detail.
It’s like adding more pixels to a digital image – the more splats, the sharper and more defined the details become. You’ll typically see subtle but noticeable improvements in the sharpness of details, the smoothness of surfaces, and the overall fidelity of the scan.
The “30 more seconds” microwave analogy is a fun way to think about it! We’re always exploring new ideas, and a “Max Enhance” option is something we might consider in the future if there’s enough community interest
3. Any plans to work with Niantic Wayfarer?
We share the Wayfarer community’s passion for exploring and documenting the world! They’re a vital part of our mission to “scan the universe”. While we’ve considered closer integration between Scaniverse and Wayfarer, we currently believe that maintaining separate apps allows us to tailor the experience to the specific needs of each community. However, we’re always looking for ways to connect the two platforms. For example, you can already include Scaniverse splat links within Wayfarer reports and appeals. The idea of anonymous contributions is something we’ve heard from the community, and we’ll definitely keep it in mind as we plan future features.
4. Auto-correct drifting when overlap is detected?
We’re always looking at ways to improve our scan algorithm, so thanks for bringing this to our attention! Scan reconstruction in Scaniverse relies on pose data from your perspective and movement, and drift occurs when errors in the tracking of the device’s motion accumulate over time, leading to distortions in the final model. This is typically an issue with larger area meshes. We’re constantly improving the process, and are exploring how to enable scanning of larger areas in smaller increments. In the meantime, here are some best practices you can follow to improve your scans. Maintaining a consistent camera angle, ensuring sufficient overlap between scans (including 360-degree captures), and moving the device smoothly can all help reduce drift.
5. What are your plans for gamification?
Thank you for your enthusiasm across our apps, and your creative gamification ideas! We’re always on the lookout for ways to make scanning and sharing more engaging and fun. We love the idea of recognizing and rewarding dedicated contributors, perhaps with badges like “First Scanner” or “Top Scanner”, or granting exclusive access after certain milestones. We’ll definitely think about this in our next feature session (:
6. What about a mechanism to report errors?
We understand how frustrating it can be when a scan doesn’t turn out as expected, especially when you only have one chance to capture a particular location. The best way to report issues is to share your scan(s) and describe the problem in the Scaniverse Community Forum or the Bug Reports channel. This allows other expert scanners to share similar experiences, offer advice and potentially identify common issues our team can use to investigate and provide a fix. You can also contact us directly at support@scaniverse.com, providing as much detail as possible about your device, the scanning conditions, and the specific problems you encountered. This will help us diagnose the issue more effectively.
Remember that once the raw scan data is saved, you can reprocess it multiple times in either Mesh or Splat mode to experiment with different settings and potentially achieve better results.
7. Coordinating with Open Street Map?
We share your passion for the vision! We’re excited for a future within Scaniverse where we can show you a richer, high fidelity pedestrian view of the world, and are working hard behind the scenes to enable this with both our user scans, and other available geodata. Please continue to support us in this mission.
Fun fact: We partnered with Open Street Maps last year on a Missing Maps campaign, where each splat shared contributed a dollar to the cause of mapping vulnerable areas and accelerating humanitarian relief efforts.
8. Downloading open formats and the rights to scans?
We’re committed to making your scan data as accessible as possible. We recently open-sourced the SPZ format for splats, which allows you to export and use your 3D assets in various development environments, including Niantic Studio. For 3D printing, which typically requires mesh data, you can reprocess your scans in Mesh mode and export them in standard formats like OBJ.
Regardless of whether you choose mesh or splat, you retain ownership over your scans. All reconstruction happens locally, and your scan data remains on device unless you choose to share it. When you do share a scan (whether unlisted or publicly on the map), you grant Niantic certain usage rights to that content, as listed in our Terms of Service. Our goal is to create a shared, living representation of the world and allow users to freely contribute to that, and these rights are necessary to facilitate that vision.
9. What about graffiti and other transient 3D environments?
This is a fascinating topic as our world is dynamic and ever-changing. We envision Scaniverse as a tool for capturing the latest view of the world, but also documenting these transient moments, from changing weather conditions to seasonal decorations. Each scan includes timestamp data, which opens up exciting possibilities for features like time sliders, allowing you to revisit locations at different points in time. Imagine being able to explore Beco do Batman and see how the art has evolved over the years! While we can’t commit to a specific timeline, it’s definitely a future we look forward to.
10. How will this app embrace artists?
We deeply value the artistic community and understand that artists and makers have unique motivations. We love helping our fellow scanners and creators beautifully capture the world around them, and share their work with like-minded enthusiasts. We’ve already taken steps in this direction by adding on-device Gaussian splatting for quick photorealistic captures, and a world map and discovery feed to showcase exceptional scans. We’re actively exploring ways to enhance scan discovery, streamline the scanning process, provide more creative control over outputs, and foster a stronger sense of community among artists. Stay tuned for exciting new features designed specifically for creators and the community coming your way!
11. Very nice 3D print tool! Any plans to address 3D printer limitations?
Those are some great scans - we’re thrilled that you find Scaniverse useful for 3D printing! We recognize that creating 3D data can be a barrier for enthusiasts, and we’re proud that Scaniverse can be a powerful but easy-to-use solution.
While 3D printing isn’t a use case we’ve actively optimized for, we’re glad you brought it to our attention. We’re always trying to give users more creative control over their scans, and will be looking to provide more basic editing capabilities in-app. If you have a specific feature or idea in mind, please feel free to write to us at hi@scaniverse.com!
12. How to delete scans in Into the Scaniverse app?
Thank you for using our Quest experience! We hope you enjoy viewing your scans in VR. The scans that show up under your Into the Scaniverse profile are scans that have been previously uploaded to our servers, which happens when you have either created a link or posted to the map. To remove them, you would need to go to your Profile within the Scaniverse mobile app, visit the Unlisted tab to select the scan, and click Delete shared scan. Please reach out to us directly at support@scaniverse.com if you experience any issues with data deletion, and we would be happy to help.
13. How do you figure the scale of a scan?
Thank you! Scaniverse uses a combination of techniques to determine the scale of your scans. Primarily, it leverages the motion data captured by your device’s sensors along with computer vision algorithms. By tracking the motion of your phone as you scan, our software can estimate the relative size and proportions of the scene. To further refine the scale, Scaniverse sometimes uses “scene understanding” techniques to match the scale of the reconstructed model to its real-world dimensions. We’re always looking at ways to improve the scaling process to make it even more accurate and reliable.
That’s all for now folks! Thank you to all who submitted a question and we hope you’re satisfied with our team’s response. Feel free to ask any follow up questions below.
Until Next Time,
The Scaniverse Team