Can someone please help me with camera data when tracking?

Hi, some of the camera specs are self explanatory …but i’m not grasping some of them. does anyone mind looking at my camera specs in the link below and teaching me where to stick them in this Thing?


http://reviews.cnet.com/digital-camcorders/sony-handycam-hdr-cx210/4507-6500_7-35122934.html

Well the spec sheet suggests that it is a 35mm equivalent lens at 36 - 1080 mm, but who know what field of view you chose to shoot at? As it is an unmarked zoom lens it could be anywhere. Also check the dimensions recorded are centered correctly. There are other recent threads here to determine the actual length of the lens from the images you record but they use other addon scripts to do that.
Finally successful trackers suggest converting your footage to a frame sequence instead of using natively compressed video (AVCHD). At the least you should generate a proxy timecode in Blender.

It was at 0 zoom if that helps. I will look into everything you said but what is the advantage of breaking it into a frame sequence? I don’t understand what that would change? Thanks

The frame seq suggestion is really related to processor overhead and removing points of failure. Highly compressed video files can decompress incorectly, skipping frames etc. You must access all frames in the right order obviously.

If the zoom was at the widest point that is great, unfortunatley with inexpensive cameras (lenses) you can’t be sure that it will actually be the advertised Field Of View (FOV or focal length). In your case 36mm at Blender’s 35mm-camera preset.

Thanks. Do I leave the optical center as it is?