A vague equation...

[aphadeon]

Does anyone really understand vague, in transitions? How does it work? I've been experimenting with transitions recently and would really like to know more about it.

A better understanding of it would serve us all- including artists, to take advantage of this engine to the fullest extent.

Anyone?

 

[Brewmeister]

It looks to me like it just 'blurs' the hard lines between the white & black in the transition graphic.
I notice if you set it too high (1000), you don't get a transition.

I haven't found any other documentation, other than the 'vague' & 'ambiguous' description in the RMXP help.

Your best bet is to keep experimenting. Try very distinct transistions (like the 'brick' images), as well as 'soft' transitions like 'flat' or 'random'.

 

[aphadeon]

I know what it does. I want to know how it does it. Does anyone have an idea of how the vague equation works?

For a transition, the graphics are frozen to an image object.
The engine makes the changes, and upon executing, it does this:
Load the transition file. Scan from 0,0 across the image for pixels of a certain shade. It then uses the coordinate to erase the same pixel in the frozen graphic, revealing the updated graphics underneath. Then it increases the shade value by 1, and repeats. This all takes place in c++ (c#?) not in Ruby, so the speed is not really an issue.

Vague causes blur. How?

Does it drop the opacity of pixels surrounding the target pixel by a precentage calculated by the number of frames?

Or am I completely off base?

Could take days to test each theory, so I'm wondering if someone has a shot in the dark for me.

 

[Brewmeister]

I'm still unable to find any definitive documentation.

It either changes the opacity, or uses a combination algorithm (add, subtract, mean...), or it anti-aliases to change the surrounding pixels.

A quicker way to test it would be to transition from a single color image (say, all pure red) to another (say, all pure blue).
Capture an in process image, and analyze the 'vague' pixels around the edge of the transition.
Then check the above methods in an image editor to see which yields comparable results.

Good luck! (post the results when you get it figured out.)