I'm writing my dissertation on video assist and onion skinning, and whether or not they're vital in producing a strong animated performance in stop motion animation.
After talking to Tim Allen, link to website here, and Reuben Loane, link to website here, although i'm sure you're well aware of them! I discovered something new about the way i have been animating for the last 3 years.
Well wrong, or at least inaccurate. In short, when animating, you can use a tool called 'onion skinning' to view half of the previous frame, and half of your current live view in a kind of ghosting of the two being viewed simultaneously.
After talking to Tim and Reuben though i've found out that this is only useful for very basic animation, and becomes more of a hindrance than anything else when it comes to actual complex character work.
Yesterday, as a separate test, i animated a simple wave completely blind (with the computer screen turned off). and it was difficult to say the very least. But today, i've animated the same movement with only video assist (flicking back between live view and captured frames, not overlaying multiple frames at the same time) and got some results i am quite happy with considering its a completely different way of animating. I'm hopefully going to do more tests before i start filming for my final film to further practice this method.