guys I can't tell you how much time is wasted trying to get the T pose "just right" it would be MUCH better if we could just go from camera A, B, C, D and draw an outline around the actor from each camera viewpoint. this would allow for a much faster way to capture that T pose.....or just some easy way to select and drag..right now it's terrible and could be so much better....or "color in" the space that is the actor with the corresponding colors from a color picker tool...or make "marks" for hands, head, torso, legs, feet etc with dots

_________________
Obin Olson
Director/Editor/DP
www.dv3productions.com
www.nevermorethefilm.com
www.mobijokes.tv

I must say that the time needed from capture to Track Forward is such that I am having a very hard time thinking about buying this software.

Here is how you can fix it:

RECORD VIDEO TAKES

IPI STUDIO looks at "watch folder" waiting for NEW recordings
IPI STUDIO uses a que system to load, find the T pose and start a forward track automatically, Studio would take the CLEAR frame from the FIRST VIDEO and then use that for the rest of the vidoes.
IPI STUDIO "renders" the videos in that folder and automatically saves the BVH files in a new sub folder as they are done
IPI STUDIO alerts user to errors that need correction as it goes

capturing data for using in PREVIZ or a feature film requires LOTS of sessions, this software does not work for that due to the user time needed to setup each shot and "render" the bvh data. If you can atleast take out the steps from capture to BVH it would be a big help.

I would only use this software for a very small job needing specific mocap data for shortform work but this can be changed if you make the WORKFLOW better.

_________________
Obin Olson
Director/Editor/DP
www.dv3productions.com
www.nevermorethefilm.com
www.mobijokes.tv

I'm a bit surprised that you have a problem with time from capture to Track Forward. Most of our users complain about time of tracking (currently, it takes 1-2 seconds per frame even on fast computers with fast video cards).

You can reduce setup time by saving actor parameters to file ("Save Actor" button on "Actor" tab) and then loading from file for subsequent video clips. If your video clips are part of the same capture session, you should not need any changes to actor parameters.

If your video clips were recorded in different video sessions with the same actor, you might have to fine-tune actor clothes colors, because lighting or camera settings could be a bit different. But still you can use saved actor parameters as a starting point for adjustments, this should make setup process faster.

Please note that you don't have to be accurate when you adjust initial actor pose. You can adjust it very roughly and then you can refine it using "Refit Pose" button. For the "Analyze Actor Appearance" function, you only need to align actor model with actor image in one view (not all views), and you also don't have to align actor model with actor image very accurately.

We are working on a better interface for actor model setup, it will be included in one of upcoming versions.



For large amounts of motion capture, more expensive traditional mocap system like Vicon can be more efficient. Our system has other advantages: you can be more flexible. For example, you don't have to plan your motion capture session months in advance to capture all motions in one day and then wait another month to re-capture some motions if something goes wrong. With our system, you can plan shorter motion capture sessions but more often; it can be a more iterative process.