I was wondering how the two Kinect video and data streams when you take video with the dual Kinect is saved as a single file output. The reason I ask this question is because I am carrying out an experiment where I will place 2 dual Kinect systems with the cameras in 180 degree angles. They will be positioned N, S, E, and W of a location. I was wondering if there was anyway to pick and choose the depth and video streams of the Kinects. so I could pair the N and E Kinect cameras and put it into iPi Studio or the S and E Kinect cameras.This is why I wanted to know the format of the output. I was wondering if there was anyway I could cut of the video and patch them together myself. Thanks in advance for all your help.

As you know, currently we employ our custom video format iPiVideo. By design, it allows for storing streaming data from multiple sources. Currently we are not planning to implement advanced edition capabilities for iPiVideo, such as extracting data of specific sources or combining several records into one.

As I understand, you are planning to record the same action simultaneously with 2 instances of iPi Recorder on different computer. First instance will be recording from (N, S) pair, and second - from (E, W) pair. Next, you want to do tracking in Studio with different pairs - (N, S) and (W, E) as recorded, and also (N, E) and (S, W) combined from different records.
For that, you can try to go with iPi Recorder 1.x which employs standard video format AVI. To extract video of specific Kinect from dual-Kinect video, you can use a cropping tool of some video editor, setting crop area to frame of Kinect in interest. After extracting two single-Kinect videos, you can combine it to dual-Kinect with video editor. The main thing here is that when cropping and combining, you need assure that video editor maintains raw video data unchanged (i.e. output video format matches input format). The other main thing is that while you are combining data from different records you need to synchronize them in time, so that images of 2 Kinects in one frame of combined video were related to the same moment of time. Otherwise, video will be useless for tracking. To help with synchronizing, you may turn on some flash light visible to all cameras for a short period in the beginning of recording (you must enable RGB recording in Recorder to see this light in output).
AVI files, recorded by iPi Recorder 1.x, and those you have assembled using video editor, can be converted to iPiVideo format using the Converter tool of iPi Recorder 2. Then, you will be able to process them in iPi Mocap Studio 2.

If the videos were stiched together but were off 1 or 2 frames in time and I converted the video back together using iPi Recorder 2.0 would iPi Studio still be able to process the video? If it would be able to, how accurate could I expect the motion and joint information calculated to be? Thanks in advance.

Studio will process video if Recorder is able to convert it, considering notes about maintaining raw video format while editing the video.
However, you must understand that there is no any magic. If video for different cameras are shifted this means at the same frame different cameras see actor in different poses. So there is no one "true" pose that should be the result of tracking at this frame.
1-2 frames shift may be more or less OK if actor motions are not quick. Anyway, it's up to you to decide whether you are satisfied with tracking result or not.