So i have this footage and for a few frames during hte footage, i simply do not have enough features to track. A perfectly lit green screen with no markers on it. MAybe 2 or 3 are visible.
 
So what will happen when trying to solve for hte camera? Usually you want at least 7-8 markers in each and every frame.
 
In my case a few will have just 2 or 3. Will mmp be able to interpolate the camera parameters for these frames ? (like mmp interpolate between 2D keyframes).
 
OR do i have to default to matchmation?
 
thanks.

tried to learn more about this interpolation thing and here my new question related to the 1st one.
 
Apparently i could choose to exlcude the difficult frames  from the camera solve (the ones with too few features to track).
 
Do my solve on the good frames first.
 
Then decide to extend the camera solve by re inserting those low quality frames to the solving sequence and rerun the camera solve.
 
However i would have to use the icon 'camera fixed' because the newframes to track are too difficult for the engine to find a valid solution.
 
So far so good?
 
My question now is: by using the 'camera fixed' icon, mmp will use the camera data from the last frame correctly solved and fix those values through out the entire sequence of low quality frames?
But isn't camera position one of the features of the camera that are kept constant? In this case this extension of hte camera technique wold not work because i'm not tracking a locked shot.
 
Is this correct?
 
Thanks.