Lets Talk About Digital Humans
In the bustle of the new year setting in, CES and updates in the space almost daily, we thought to take a brief few minutes to share some thoughts on what a digital human is, how they work and why they are important. With the dedication to education and that very discussion, we wanted to have a healthy start here.
What is a Digital Human?
A Digital Human is a volumed or 3D asset built from human performance using various data streams (RGB/Motion Capture, point cloud & polygons) that create a human performance represented virtually or physically. This digital human is brought to life by sequential data streams. There are variations like digital twins, digital doubles and holograms that all speak the same language.
Investments are starting to take place in this space. All kinds of aspects surrounding the digital human experience are beginning to take form as the technology and its creators start finding solutions to the problems of the industry. Mostly revolving around the tools and the massive amounts of data.
What is Volumetric Video?
Let's take a closer look at the volumetric capture process. You have an array of cameras, all needing precisely positioned, pointed in the right direction and stable with no vibrations. These cameras are all synchronized and captureing still frames from these precisely calculated positions. https://modtechlabs.com/volumetric-video-introduction/
Each of these cameras are capturing 2 dimensional images. That image data is then manipulated and organized so that when processed, it removes the unneeded data and takes the volume data with grouped camera datasets to provide individual partial 3D models.Those partial 3D models are then "stitched" together to create a fully volumed 3D asset. Then that process is repeated at whatever the frame rate is to create a sequenced data stream of point clouds and polygon mesh (soup) that can then be comported to whatever the use desired spatial media output.
We have always been looking for ways to make believe. And what better way to make believe than to interact with other human beings. Or in the modern day to fool the brain into thinking its interacting with another human being. This is why theater has been so powerful for thousands of years. And in the history of cinema and digital graphics for gaming, VFX and projection mapping, we continue the same trend. The term uncanny valley is now one of the key metrics that decides success or failure of an effect or a project as a whole. Now, we reach an era where digital humans are in fact here and the believability of these digital humans is the key to sparking imagination for the collective audience. And thats what we all want as creators, to spark imagination.
How can you use this Asset?
This is the question being explored and the answered now in 2022 - The First Year of the Digital Human. How do we create new ways for digital humans to be consumed? The answer is that with very few limitations beyond budget and time. New and creative ways to implement digital humans in the real and virtual worlds are coming to life. The use of XR and live event applications are becoming possible. Holographic displays large enough to project a human body realistically are becoming possible and high quality assets can be pipelined not only to the high end applications, but be broken down for web and streamlined purposes, all in 2022.
Live Displays vs Virtual Presentations
Simply put, it's important to understand that it is all the same data. Doing this uses the same data types as doing this. That being said, comporting the data is the key. But remember, virtual platforms can be multi-purposed with the same data. To say that when you use volumetric video to create a 3D performance, that performance can be used in gaming pipelines and then immediately reformatted as an NFT or other digital messaging format. It is an exciting time to be in this space.
Compression vs. Quality
How do you have high quality assets, that can maintain as much of that quality given the reduced data rates of some of the end displays? The answer is a simple adage that has prevailed through all kinds of compressions over the years, Garbage In, Garbage Out. In order to preserve a digital human to be used in a multitude of high end application, you have to achieve high quality acquisition and compress from there.
By starting with a high quality capture, you can have data pipelines that maintain quality as you compress the files for different outputs. We are getting to the point where the data can be wrangled in a way that allows us to preserve high quality for whatever the activation. The question is who will put the tools into creators' hands first?
What is the future?
As data streams become less restrictive and as the tools for working with digital humans becomes more common for creators and technology interests, the use of digital humans for messaging, make believe and so much more will become common in the world around us. And its happening quicker than most people think.
Stay tuned everyone. the first year of the digital human really is here.
Published January 7th 2022