21–26 of 26 posts
I find everything I read about OTOY's stuff (seemingly purposefully) confusing. They seem to have a render farm, light field format and renderer, and a streaming video format that they will always mix together in their demos. "Look at these amazing renders. It runs on a phone!"* *its actually just streaming video to the phone. I'm pretty excited about this stuff but I keep finding myself frustrated trying to crack th…
At 13:45 he talks about: 1: streaming a Unity/Unreal game into a surface texture 2: packaging an entire Unity project into an ORBX file So... am I understanding this right: that ORBX can contain not just a light field, but all the assets and logic for a game, compiled to LuaJIT, which the ORBX player (or orbx.js) will then play? And Unity can target this for output?
2) yes, it works as shown in the video using the standard Unity samples unmodified to prove viability of the system on GearVR and Samsung Internet. We also have this working non PC/rift and daydream now
ORBX.js is ORBX is a container with the render graph of the content in xml/Json + assets. Just like a web page, it can be cached to archive (.orbx file) or streamed from a URL or URI over raw UDP/tcp or web wss or https
What about real video recording to generate real VIDEO light fields? Also: how much processing is it required to get video light fields from video recording (with an array of cameras I suppose)? I mean: does it scale?
Earlier quoted context omitted.
The minute they mention MPEG, I already know whats going to happen - Licensing. Spoken by a dev that didn't incorporate video, due to licensing, not technical reasons. The open bit is just the draw.
See my post above. ORBX schema and formats submitted to mpeg must be usable by anyone and everywhere with only an MIT license (same as lua which we use script nodes in the graph). I was told this is fine, but that they need it as a tier 1 MPEG license (their equivalent as I understand it).
If not now, then later. In my opinion, this would be MPEG getting their teeth into a new tech, in order to either license it now, or wait until it has traction and then license it.
It is absolutely amazing tech though. Shame.
What about real video recording to generate real VIDEO light fields? Also: how much processing is it required to get video light fields from video recording (with an array of cameras I suppose)? I mean: does it scale?
This is the basis of our work with Facebook which can capture a light field using 6 or 24 cameras. We run a Job on cloud and it comes back as a 3D scene ORBX asset you can load into Octane for Nuke, Unity etc. https://twitter.com/otoy/status/894972675679965184
Earlier quoted context omitted.
This is the basis of our work with Facebook which can capture a light field using 6 or 24 cameras. We run a Job on cloud and it comes back as a 3D scene ORBX asset you can load into Octane for Nuke, Unity etc. https://twitter.com/otoy/status/894972675679965184
Can anyone do it? Is the software to do that open source?