hi.
 
i really don't understand what is the point of camera extensions beside saving time when extending the work area (when you extend your work are but the new frames you are now including have a continuation of the 2D tracks already solved in 3D and there are still plenty of 2D markers for the engine to compute a solution on these newly included frames).
 
 
I've read the help and there isn't much information about camera extensions.
 
I don't get the point of this feature. Niko can you give me a concrete example where camera extensions would be used besides the one i mentionned above?
 
In hte help it says that you can use this feature to solve sections where it is hard to get a solution. But how can a camear extension help in solving frames that are hard to solve on the first place. Unless the markers a fixed in 3D space and the camera is locked off (and therefore just extends the solution of the well tracked frames to the new frames by keeping hte camera parameters constant == extend fixed???) but in this case, if nothing is moving, you don;t need to matchmove those frames....
 
i don't get it..
 
thanks.

Hi,
 
Sometimes it is hard to globally solve a shot and to compute a good camera
but a part of the sequence seems easier to process and more robust.
When you solve this part and then extend your solution, MM uses the first computation as a starting point to be refined and do not try to solve all the frames at once.  
The first estimation of the camera parameters is then more accurate.
I don't have a concrete example to give you but you can consider this option as a workaround to compute shots that failed to solve in a normal way.
 
regards,
 
Niko

thanks.