I want to make my models exist in the video I take, no handheld stuff so no motion tracking.
Like lets say I have some text/logo that I want to sit perched on a bench or something, how do bring in the video and allow the text to cast shadows. Also, let’s say that the text/model/object has a reflective texture, how would it be possible to have it reflect what is in the real world? Thanks…
First you need to motion match your shot using a 3d camera tracking software.
Means to compute the camera path in three dimensions from a two dimentions moving video shot.
Give a try to both if you are on Win
Both have a blender script importer/exporter available.
Then you should play with your resulting camera path from blender, try to reconstruct some basic geometry & make them invisible to only recieve shadows from your 3D objects.
You’ll have to use some compositing technique inside Blender using composite nodes or in external commercial compositing package (After Effects, Combustion, Shake, Nuke).
If you want a robust but cheap ($399 for this IS CHEAP) 3d matchmover with great support i suggest you Syntheyes from Andersson Technologies (http://www.ssontech.com/) available on OSX & Win, 32bits & 64bits
Like lets say I have some text/logo that I want to sit perched on a bench or something, how do bring in the video and allow the text to cast shadows.
To cast shadows? make a 3D object at that spot in the video to receve the shadow. To bring in the video, look at the video sequencer in blender, or learn compositing nodes.
Also, let’s say that the text/model/object has a reflective texture, how would it be possible to have it reflect what is in the real world?
Take a second image showing what it would reflect if it were actually there.