I’m curious, i often film with my canon SLR, and Blender can read the .mov files it generates.
And i would like to completely do the film making workflows with Blender.
For several other reasons i already prefer Blender above video software, because of the high control i have in Blender, you can do any kind of animation just as you want it; not because its a standard title box or so…
The one thing that is bugging me is “deshaking” with hand shot movies; as i not always travel with my tripod. I have done some tests now with blender, but i am used to do this with virtualdub plus the deshaker plugin. Virtual dub cant read my .mov files, so i always need to convert first to avi, then deshake 2phases-tuning, then save.
The virtualdub deshaker and Blender work each quite different
Virtual dub uses a fine grid of markers and from which an x% needs to be OK or else deshaking for that frame is skipped, also you can configure some center area, it works pretty fast (well it still takes time), based on the grid of markers the deshaker can tell how the camera has moved (basedupon the fine grid of point data).
Blender uses only a few markers to follow, but a lot more computation seams required and is slower. Although blender can do other 3D tracking things too, as it uses different math. Blender has lots of controlable options, to improve a live video, about borders, or markers ; where deshaker seams more end result focused how smooth should the rotation be or XY translation, how smooth do you want adaptive borders to zoom etc.
Blender has the big plus of reading directly .mov files
Having one workflow tool is well awesome… but not if it takes to much time
So then for deshaking when would you prefer to use Blender over Deshaker ?
Are there situations in which Blender should give better results ?
If i asked this question on a virtualdub forum, anyone would point me to use virtualdub, but i like to understand more reasons for prefering blender for this, if there are.