Hi,
I used for a long term many different “classical” photogrammetry software and one of the key steps to improve the result is to use a camera calibration (when applicable from the camera - fixed focal lenght, focus etc). This could be done either in an external software or library (Matlab, OpenCV) or in the software itself.
Now in RealityCapture i see that the camera calibration is estimated, but also can be inserted using prior calibration fields or the XML files.
Is there anyone who use this kind of approach? And if yes, what is the experience? Does it still improve the resulting models or is it more obsolete with the superior alignment routine?