[Walter Soyka] “That’s funny; a dissolve is just “over””
I hope you won’t mind me pointing out that this is a bit of an over-simplification, even as regards the most common implementations of the dissolve in the applications we are familiar with. Certainly Over is one way of doing it and is probably what most people think of as what’s happening, but it’s far from being the only way.
The way I see it, a dissolve requires three things: a) a blending operation to determine how the pixels of the two overlapping layers interact; b) a means of handling the alpha channels of the two images, if present; and c) one or two sets of animation curves that determine the rate of the transition.
This being the case, it’s not difficult to see that there are a huge number of ways in which a dissolve could be implemented. For example, any of the commonly used blend modes will yield workable results and allow for many different looks but clearly one could devise custom blend modes for even more complex results. FCP X offers a pretty wide range of these kinds of options, some better-looking than others. Obviously how the pixels are blended will greatly impact on which appear more persistent than others – in fact the blending operation will have a greater impact than the animation curve a lot of the time.
Simon Ubsdell
tokyo-uk.com