Inside or outside of Blender, I need to look at an overlay of two videos and quickly animate (paint) a mask for their proper combination. What do I use to do it? In Blender, I have the sequence stips overlaid and can paint on any given frame but don’t know how to specify that the paint data (for each frame) be saved and reapplied as a mask. Outside of Blender I have no thoughts, except that I want to use all free open source stuff. Help?
I have no experience painting on image sequences, but there is a sequence wipe tutorial in my sig that uses nodes, I don’t know how helpful it will be to your particular situation, but maybe it can be of some use. On occasion, I have made masks, by changing my background to solid black, and certain materials to shadeless white, and removing lamps and AO. Again, not sure if this is helpful to your particular project or not.
Sorry for taking so long to reply. The problem with trying to model a black and white image is that I cannot see the two images superimposed on which I need to paint. Also, this method requires manually dragging verticies and maintaining an acceptable form of geometry for the mask. Perhaps I could do some form of vertex paint on a mesh and get the camera lined up just perfectly but in truth that sounds like more trouble than its worth. I really need a mechanism that is designed for the task. Thoughts?
What you can do is use shape keys to drive your mask animation. Or you can add hooks to a shape and animate the hooks.
What Modron suggests will work. Black background and white shape that is animated. Then render that out to a sequence and you can bring it back into the VSE and use Alpha Over to maks your source.
Thanks for the reply, but you don’t see my point at all. I know technically how to render black and white shapes to a mask, but its a silly method. I need to do a lot of frames, so I need an EFFICIENT method. lookatimagesdecide where to mask paintpaintpaint setasmaskforthisframe nextframe paintpaintpaint etc… What you’re suggesting involves a lot of dragging verts to make a shape, making sure the shape keys all have key influence 1 on their frames and 0 around their frames (w/ making the shape key by dragging verts thats 4 keys total, and the shape keys aren’t really linked to the frames for editing purposes, plus in order to see if the mask you made is really what you want you have to repeatedly render the frame because you wouldn’t be drawing on you final images. NOW, can somebody give me more helpful advice or tell me why my perception of the described methods being ridiculously complex for what needs to be a quick process is incorrect?