Cardboard SDK Transformations

[Alfred Hanssen]

I'm building a 360 video viewer for iOS in order to better understand the nuances of monoscopic and stereoscopic 360 video. Effectively, I'm trying to reverse engineer what the Google Cardboard SDK does.

My in-progress code is here: https://github.com/alfiehanssen/ThreeSixtyPlayer

I understand monoscopic video and the basics of stereoscopic video. I would like to better understand the client-side configuration/transformations required to display stereoscopic video.

Specifically, in two parts:

1. 1. Cardboard can display monoscopic video files in stereoscopic mode. Although the same exact video is being presented to each eye, each eye's video view is clearly different. What transformations are being made to each eye to get this to work?

2. 2. For stereoscopic videos (let's assume top/bottom layout), it also appears that transformations are applied to each eye's video view. What are the transformations being applied to each eye's video view?

It looks as though the video is being skewed. There are black masks applied to all sides of each eye's video, where are these coming from / are they the result of transformations?

I asked this question on Stackoverflow: http://stackoverflow.com/questions/40454781/reverse-engineering-google-cardboard