What is Volumetric Capture and How It's Done?

I bet you already heard about volumetric capture a lot. There's many buzzwords flying around, but it's easy to get lost in this field. This happens because it's a new, very fast growing industry and you might not have so much information about it.

Some people think it's some kind of filter for a regular video, some think it's a 3D scan. Well, it's definitely not a filter, it's a completely new technology. Volumetric capture produces volumetric video which is close to a 3D scan, but the technology is different and much more advanced. In simple words, it's a type of 3D model, but it's not just scanned, it's recorded. It is a moving 3D video which on top of that is interactive - you can change the perspective, the angle, zoom in and zoom out. You can interact with the video as you could with a real life object or person.

There are a few different ways to use volumetric capture to record - some companies like Microsoft, use photogrammetry for image conversion to 3D objects. After that there are different algorithms to generate a pointcloud or a mesh of the object. Then they extract very realistic 3D models and create animations out of them.

This is a very expensive method, because they use a really big rig of hundreds of cameras to take regular and infrared captures. Then the very difficult algorithms combine those pictures, remove the unnecessary parts by keeping only the main details from the pointcloud.

This is possible to invest a great amount of money, and have a few seconds of volumetric video. But this type of volumetric capture is very difficult and takes a lot of work and computing power so it's very expensive.

There are also other ways. Since Kinect Azure or Intel Realsense was released, it's possible to create good volumetric videos on your own. Some companies, like EF EVE have created software which allows to have just a few Kinect Azure or similar depth cameras and capture your own volumetric video.

How you ask? It's simple - you need to put 2 or more depth cameras in different angles. The more cameras the better - with two you can only capture a part of the object, with 4 you can already get a full-body scan. So you set up 4 cameras, and with EF EVE auto calibration tool you save a lot of time and energy for the cameras calibration. After you

do that, boom - you are capturing a volumetric video of yourself, your dog or maybe toy which is moving.

Later you can edit the video in the creator tool also provided by EF EVE. Remove noise, apply filters or edit the materials. And that's it! By the way, after you capture your volumetric video, you can upload it in their platform to share it with others.

Volumetric capture is now that simple - you don't need to invest a pile of money or have any programming skills. It's the time for a new generation of video and everyone can make it. It's the future right in your hands.

EF EVE™ is the World's most loved volumetric capture software, editing and distribution tools for volumetric video. Use 2, 4 or 8 Azure Kinect, Kinect V2 or Intel Realsense and start making the most immersive 3D content for Web, Vr and AR.