Today at our beer and pizza, Max Chan mentioned about a new concept and workflow in compositing: Deep Image Compositing.
It was a new concept that he heard or read about, but he could not be sure exactly what it was. So I did a search for it on the internet and this is what I found.
It is essentially a method for storing information about an image that goes beyond the basic pixel colour and intensity. It is similar to multi channel image formats like the EXR developed by Industrial Light and Magic, RPF, RLA and the PSD format used by Adobe Photoshop.
However, it is different from these formats. It attempts to solve edge aliasing in these extra channels when used in compositing. To this end, it also implements a different way to store and compress image data.
For a quick video demonstration and overview (if you decide not to visit the Deep Image website), here are 2 videos from www.vimeo.com, posted by Johannes Saam, who presented an overview of deep image compositing at the Foundry user group meeting.